United States
Environmental Protection
Agency
Office of
Emergency and
Remedial Response
EPAIRODIR05-91/175
September 1991
C 0(1 I
'P'B~z- r6 tf /3 {,
oEPA
Superfund
Record of Decision:
Sturgis Municipal Wells, MI
u . S. Environmental Protection Agency
Region III Hazardous Waste
Technical Information Center
841 Chestnut Street. 9th Floor
Philadelphia. PA 19107
HazardoUs Waste Collection
Information Resource Center
US EPA RegIon 3 '
Phllodelphlal PA19107
EPA Report Collection
Information Resource Center
US EPA Region 3
Philadelphia. PA 19107
-------
.
REPORTDOCUMENTA~ON I 1. REPORT NO.. I ~ 3. A8cIpI8nI'8 Acc:8uIon No.
PAGE EPA/ROD/R05-91/175
4. 1118 and !kMI8 5. A8port D818
SUPERFUND RECORD OF DECISION 09/30/91
Sturgis Municipal Wells, MI B.
First Remedial Action - Final
7. AuIhor(8) .. PiIt10nnlng Org8nIza1Ion AapI. No'
1. PwtonnIng 0rg8InID1I0n Nama and AcId- 10. ProjactIT uklWoril UnIt No.
11. Contract(C) 01' Gr8nt(G) No.
(C)
(G)
1~ 8poMorIng 0rganz8d0n Nama and Addr-. 13. Type of Repor1& Period Covered
U.S. Environmental Protection Agency 800/000
401 M Street, S.W.
Washington, D.C. 20460 14.
15. fk..' m81118IJ No...
111. Ab8Ir8ct (limit: 2110 wotda)
The 5-square-mile Sturgis Municipal Wells site is an active municipal well field in
Sturgis, St. Joseph County, Michigan. Land use in the area is predominantly mixed
industrial and residential, with several wetlands areas located near the site. The
estimated 10,000 residents of Sturgis use the intermediate and deep aquifers as their
drinking water supply. The site includes the former Wade Electric facility, which
closed in 1966 and burned down in 1974, and the Kirsch Company, which operated as a
manufacturing facility until 1980. These two properties have been identified as two
source areas responsible for the ground water contamination in the aquifer used for
the municipal water supply. State investigations starting in 1982 identified
contaminated soil and VOC-contaminated ground water onsite and extensive ground water
contamination in various wells throughout the city. This Record of Decision (ROD)
addresses final remediation of soil and ground water. The primary contaminants of
concern affecting the soil and ground water are VOCs including benzene, PCE, and TCE;
and other organics including PARs.
(See Attached Page)
17. o..a-t An8Iy8I8 .. 088crIpI0ra
Record of Decision - Sturgis Municipal Wells, MI
First Remedial Action - Final
Contaminated Media: soil, gw
Key Contaminants: VOCs (benzene, PCE, and TCE), other organics (PAHs)
b.~T--
c. C08A 111'1811Wnq1
18. A¥IIII8bII1y --- 11. S8cwtty CI888 (1hI8 R8port) 21. No. of Pall"
None 165
20. S8cwtty CI888 (1hI8 Page) 22. PrIce
}(Inn".
~1' n'" I :zn (4-171
50272 101
(8aa
s.. ltulnJCfi- "
-
(FomwrIy NT1So35)
Depat1rn8nt of Co.......ce
-------
EPA/ROD/R05-91/175
Sturgis Municipal Wells, MI
First Remedial Action - Final
Abstract (Continued)
The selected remedial action for this site includes treating VOC-contaminated soil using
vapor extraction, followed by activated carbon adsorption to capture off-gases;
regenerating carbon offsite; excavating and disposing of offsite 10,890 cubic yards of
low level PAH-contaminated soil (10,000 cubic yards from the Kirsch property-and 890
cubic yards from the Wade property); onsite pumping and treatment of ground water using
air stripping, followed by vapor phase granular activated carbon to treat off-gases;
discharging the treated water onsite either to surface water via a storm sewer or made
available for use by the municipal water system; ground water monitoring; and
implementing institutional controls, and site access restrictions including fencing. A
contingency plan would be developed to allow for the rapid installation of a mobile water
supply treatment system, should it be required at any of the existing municipal supply
wells. The estimated present worth cost for this remedial action is $13,810,000 for
discharge of treated ground water to a storm sewer, which includes an annual O&M cost of
$644,600 for years 0-3 and $598,000 for years 4-30.
PERFORMANCE STANDARDS OR GOALS: Chemical-specific soil clean-up goals for VOCs are based
on State standards, and include PCE 14 ug/kg and TCE 60 ug/kg. Soil containing PAHs will
be excavated to 330 mg/kg, based on State standards. Chemical-specific ground water
clean-up goals are based on State standards, and include benzene 1 ug/l, PCE 1 ug/l, and
TCE 3 ug/l.
-------
DECLARATION
SELECTED REMEDIAL ALTERNATIVE
FOR THE
STURGIS MUNICIPAL WELL FIELD SITE
STURGIS, MICHIGAN
Statement of Basis and Purcose
This decision document presents the selected remedial ac~ion for
the sturgis Municipal Well Field Site (Site), Sturgis, Michigan,
which was chosen in accordance with the Ccmprehensive Environmental
Response, Compensation, and Liability Act (CERClA) of 1980, as
amended by the Superfund Amendments and Reauthorization Act (SARA)
of 1986, and, to the extent practicable, the National oil and.
Hazardous Substances Pollution co~ingency Plan (NCP). This
decision is based on the administrative record for this site.
The State of Michigan concurs with the selected remedy.
of Concurrence is attached to this Record Of Decision.
The Letter
Assessment of the site
Actual or threatened releases of hazardous substances from this
Site, if not addressed by implementing the response action in this
Record of Decision (ROD), may present an imminent and substantial
endangerment to pUblic health, .we1fare, or the environment.
DescriDtion of the Selected Remedv
The selected remedial action is a final remedy for the Site. The
purpose of this remedy is to eliminate the sources of groundwater
contamination and restore the aquifer to its beneficial use.
The major components of the selected remedy include:
Soil vapor extraction of Volatile Organic Compounds
(the principal threat) in the source area soils to
Michigan Act 307 Type B Cleanup Levels;
Excavation of soils contaminated with low-level
Polynuclear Aromatic Hydrocarbons (PARs) to Michigan
Act 307 Type B Cleanup Levels for indicator
contaminants and disposal of soils in a Michigan Type 2
Solid Waste Landfill;
Extraction and treatment of groundwater using air
stripping, with vapor phase granular activated carbon
to be used to treat the off-gasses;
-------
2
Discharge of treated water to surface water (via storm
sewer) or to the municipal system;
A minimum of a thirty-year groundwater monitoring
program to assure the effectiveness of remedial action
and the quality of the municipal water supply.
3tatutorv Determinations
~ ._,;]~ :.:.~~
~~e selec~ed =emedy is protective of human health and the
environmen-c, complies witn .f"eQ61:al and state.. requirement&;.~ha't. ar~
legally applicable or relevant and appropriate to the remedial
action, and is cost effective. This remedy utilizes permanent
solutions and alternative treatment or resource recovery
technologies to the maximum extent practicable, and satisfies the
statutory preference for remedies that employ treatment that
reduces the toxicity, mobility, or volume as a principal element.
Because this remedy will not result in hazardous substances on site
above health based levels, the five ye~r review will not apply to
this action.
tf:/J
te
-------
STATE OF MICHIGAN
NATURAL RESOURCES COMMISSION
MARLENE J. FLUHARTY
GORDON E. GUYER
O. STEWART MYERS
RAYMOND POUPORE
IJ
JOHN ENGLER. Govemor
DEPARTMENT OF NATURAL RESOURCES
STEVENS T. MASON BUILDING
P.O. BOX 30028
lANSING, 1.41 48909
~
RC:UW ~, DIREX:ltR
September 20, 1991
Mr. Valdas Adamkus, Regional Administrator
U.S. Environmental Protection Agency
Region S, SRA-14 .
230 South Dearborn Street
Chicago, Illinois 60604
Dear Mr. Adamkus:
The Michigan Department of Natural Resources (MDNR), on behalf of the State of
Michigan, has reviewed the September 17, 1991 draft Record of Decision (ROD)
for the Sturgis Municipal Well Field Superfund Site in St. Joseph County. The
MDHR concurs with the proposed remedies in the ROO. We also concur with the
determination that cleanup levels for both soil and groundwater will be
dictated by the Type B cleanup rules promulgated under the Michigan
Environmental Response Act (MERA) , 1982 PA 307, as amended. We agree that
specific remediation goals are to include:
* Soil vapor extraction of Volatile Organic Compounds (VOCs) in the
source area soils to MERA Type B Cleanup Levels;
* Excavation of soils contaminated with low-level Polynuclear Aromatic
Hydrocarbons (PAHs) to MERA Type B Cleanup Levels for indicator
contaminants and disposal of soil in a Michigan Type II Solid Waste
Landf ill;
* Aquifer restoration by extraction and treatment of groundwater using
air stripping, with vapor phase granular activated carbon to be used
to treat the off-gases. Extraction and treatment would be conducted
until MERA Type B Cleanup Levels are reached for VOCs;
* Discharge of treated water to surface water via storm sewer, or to
municipal system. Discharge will meet the substantive requirements of
an NPDES permit or pretreatment requirements, whichever is
appropriate;
* A minimum of thirty years of groundwater monitoring to assure the
effectiveness of the remedial action and quality of municipal water.
-------
Mr. Valdas Adamkus
-2-
September 20, 1991
It remains our position that the Michigan Water Resources Commission Act
(WRC), (1929 P.A. 245, as amended) MCl 323.6(1) and the associated Part 22
Administrative Rules MAC R.323.2201 et.sea. are applicable or relevant and
appropriate requirements (ARARs) for the remedial action for this site because
hazardous substances in the aquifer beneath the site are migrating to degrade
previously uncontaminated groundwater. However, it is the MOHR's judgement
that the selected remedial actions for the Sturgis MuniciDal Well Field Site
will provide for attainment of all (ARARs) including the WRC and the
w.:GC:ltea :~~: 22 Rules. The se~ected remedial action will preventftJrther
discharges vi' injurious S~Dstances into the groundwater outside" of the
containment area and properly remediate existing groundwater contamination. "
We are pleased to concur with the selected re8edies for the Sturgis Municipal
Well Field Site and look forward to working with you on their i8plelentation.
~~7tp-1}
Delbert Rector
t,:;r:NG Deputy Director
517-373-7917
cc:
Mr. Jon Dikinis, EPA
Ms. Mary Pat Tyson, EPA
Ms. Terese Van Donsel, EPA v---
Mr. William Bradford, MOHR
Mr. Scott Cornelius, MOHR
Mr. Robert Franks, MOHR
-------
TABU 01' ool1'1'BII'1'8
SITE NAME, LOCATION, AND DESCRIPTION
SITE HISTORY AND ENFORCEMENT ACTIVITIES
HIGHLIGHTS OF COMMUNITY PARTICIPATION.
SCOPE &~D ROLE OF RESPONSE ACTION
S~~Y OF SITE CHARACTERISTICS
. . . . .
SUMMARY OF SITE RISKS. . .
. . . .
DESCRIPTION OF ALTERNATIVES
. . . . . .
. . . '.
. . . .
. . . . .
. . . . .
. . .
. . . . .
. . . . .
. . . . .
.......
. . . . .
. . . . .
. . . .
COMPARATIVE ANALYSIS OF ALTERNATIVES: THE NINE CRITERIA
THE SELECTED REMEDY. . .
. . . . .
STATUTORY DETERMINATIONS
. . . .
DOCUMENTATION OF SIGNIFICANT CHANGES
RESPONSIVENESS. SUMMARY
.......
. . . .
. . . . .
. . . . . .
. . . .
. . . .
. . . .
. . . . ..
FOLLOWING
1
4
5
6
7
9
19 .
27
38
40
45
-------
RECORD OF DECISION SUMMARY
Sturgis Municipal Well Field
Sturgis, Michigan
SZTB ...., LOCATZOB, ARD DBSCR%PTZOB
1.0
The Sturgis Municipal Well Field Site is located in the City
of Sturgis, Michigan. The city is located in south central
Michigan in st. Joseph County approximately two miles north of the
Indiana state line, halfway between the chicago and Detroit areas.
The majority of the city lies within the political confines of
Sturgis Township, while northern, north-eastern, and eastern
~ortions of the ~unicipal area lie within sher.Q~n, Burr Oak and
Fawn Ri.",rer Townships, respectively. . :The ::ity encompasses
approximately 5 square miles and approximately 10,000 people reside
in the city. The city's economic base is largely industrial.
The City of sturgis lies on an outwash plain at the foot of
three recessional moraines. Topographic relief within the confines.
of the city is very small, with maximum relief of approximately 40
feet. However, relief increases considerably as one moves to the
north of the city into the morainal feature. Topography to the
northeast and northwest of the city is somewhat hWIDDocky with
several lakes and wetlands occupying the low lying areas. A
surface water divide to the west and northwest of the city
separates the prairie River and Fawn River watersheds. Because of
the relatively high permeability of soils in this area, rapid
infiltration of precipitation can be expected.
Several surface water features are present within the vicinity
of Sturgis. Approximately two miles to the south and east of the
city lies the Fawn River which flows out of its headwaters,
approximately 8 miles east of Sturgis, to its confluence with the
st. Joseph River approximately 70 miles away. Few minor surface
water features are present within city limits, including absorption
ponds at Ross Laboratories, Kirsch Co. Plant NO.2, and Sturgis
Foundry Corp.; and the Nye Drain which drains the city's sewerage
treatment plant outfall south to the Fawn River. Absorption ponds
at Sturgis Foundry Corp. and Kirsch Co. Plant No.2 allow
non-contact cooling water to infiltrate into the ground. The
absorption pond at Ross Laboratories allows storm water run-off to
infiltrate into the ground.
Several large kettle lakes are present within the Sturgis
Moraine, including Minnewaukan, Omena, Grey and Stewart Lakes. The
margin of the Sturgis Moraine also contains significant wetland
areas, typical of glaciated, deranged drainage systems. The
wetland associated wi th Baker Lake, east of the ci ty , is the
largest wetland in the area, covering approximately 0.5 square
miles. Each of the lakes along the margin of the sturgis Moraine
contains relatively large wetland areas surrounding and adjacent
to the lake. Other wetlands of note are located southeast of the
city along the Fawn River.
-------
2
The geology beneath the city consists of layered deposits of
high permeability (sand and gravel) aquifers and low permeability
(silt and clay) till units. soil borings were used to interpret
the aquifers' stratigraphy. The till units are discontinuous and
provide windows connecting the aquifers. OUtwash thickness
encountered between the deep and the intermediate till units varied
from between S2 feet to 14S feet. Most of the water supply wells
remove water from the deepest aquifer.
other than the results of the Remedial Investigation (RI) ,
there are no published, detailed s~udies of the hydrogeology or
groundwater flow conditions in the Site area. However, based on
gr~undwa~er levels in the city, the t~pography and locations of the
surface water bodies described above, the city appears to be
located on top of a groundwater high between the surface water
bodies. In the absence of pumping, groundwater flow may be
radially outward from the city. Municipal and industrial pumping
has created a cone of depression, which is superimposed on the
water table and piezometric flow system. Therefore, there does not.
appear to be a strong influence of a regional gradient .on the
groundwater flow system.
The aquifer beneath the city supplies the city's 10,000
residents and numerous industries with water. Several of the wells
have been contaminated with Trichloroethene (TCE) or
Tetrachloroethene (PCE) and have either been shut down or used for
non-consumptive purposes. The RI identified two source areas
responsible for the aquifer contamination: the Kirsch. Company Plant
No.1 property (Kirsch) and the former Wade Electric Property
(Wade). See Figure 1.
The Kirsch property currently consists of offices for the
Kirsch Co. and a vacant lot. The vacant lot and existing buildings
had previously been a large manufacturing facility. It is believed
that solvent handling practices during manufacturing and
maintenance have led to the TCE and PCE contamination documented in
the soil and groundwater by the RI. The vacant lot also contained
a cyanide based metal plating operation. Manufacturing at this
location ceased in 1980. The manufacturing facilities previously
located on the vacant lot have been demolished.
The Wade property is currently occupied by the sturgis Archery
Center. The Wade Electric facility closed in 1966 and burned down
in 1974. The current property owner removed six underground tanks
from the rear of the Wade facility. According to the owner, these
tanks contained a thick oily substance and were not leaking.
-------
. J1.~I" - ~--- t.
- ::--::J ~ ~ =-"
" I
l ,~, ,
- - - --_J ~CL:L~C":=:. /1r".. - , ..~
;,~- ._IOu 'j--'oIl[l1 :"1,1, I~S ~' 'I '~~ "'" T-- :-=.--"C-=
,~---, J ; :.~~t--= ~l~lJ i --., -- u.--;
:~~:~~..'~~,rl' ~:~lDl~ -)i, Jr, -,I'f~', f;IL-"':-~ ~~'--~~-.'~, ",~*
]IL"'~ IL.Jw L, "-~~, . J 11... ~- . j
,", !..J:;.;rrr::,~~~~~' ~'I
S' --. lllio~"-;';1 I '."'. r. -.-:---- J~\--'--....J
. .'11 ...~~ -:':-:--:::-~- 'F4--'--
, ;.,':.~ ' :-,,~~~ ~0;... q:: _.:::i'S!~2.::'~:I': It!a U ',Lr (. jar ~ :~I~':
: /~ . :', =~:I,~,~ ~!u,:_:'",-:~~~-~t=r: ~~ ,~UJe['.J! "
1 ---: '. .,.so.. i." ,:; ,,' ..I ...~ I ' ~i--;:::' -,~ ---looe I',
~"~.. . :: . :::: "', ;::.., :r:~-~-'-.i _.~:: 5i~.:\~' ~.'\' .~~.t~:\-. iI' ~....: .=--~:. l-:-'Il~]'i 'I' .:
~::"_~/ .::,.. ;' ~ j J. L.::i:! ~~~~'.!f2' )r~Fi~~ :1- [iUI _1n[~i 1 ....'
, "",10,' -SW"D "'.',~ ".-'UJL. I --'r-'r'r'III'
. . ,1' QW-, "~"2 JI,W-t- . ~ \ ~ .\ '\ ,. "'- _. I I It,
. i: ! il '; ," ,,5~""-":'~";G~':~iP"" ~,)~,~:,:; ~i\1 j\J'!.-, i' "~; ~ ' :1,1
' ". .-, - >' ~--- -~ L-' ~"DE 1, , :1\ ,,\ :- -"-. . LJ' I JI ~I 'I'
::', _,1-"'--':---'--"- ,--= J-'." ELECTR L 'I. "---==I"R ] ~! ,uLL~,~
. ~ ---- I . \ ~ ------IL_._~---SL... - '----i
~- " - ..,.,- ... ........:::. ------,----;- r--I-Cli----j-
"--~-;I-' , "',,,; .< ;,' -. i:-'i:c-' =-:~,-."J-~'t" '
""d r ~, . Lao ... ;; ,':':"8'=:=- ~--.= --:-= .l-lj "-1 =,Ji:
o I ";",: --~....... -- ...~..,.' ---- : I -.1' I . + i I'"
g 9, 'J'-~/:: , .... ..,. ",,-~1:[" ---- ~:~-~, _i=-~ :r-"i__-J::
~ 0 I ~i ~~:::;' ,'__'L ---"~-'--_~~II, f,-"
~ - 'J l ~ " ,-' ,w.o..; l~ i---J~u ,
-" C L" -' - '-""'-'---'» r II -- -- ,---, ----r- ,
~ '1 r~ ~ ...,
"? \ ~i fl' , ,", .;, I ~:J~jlll
".-IA
...-,.
".-tC
5
..--I
It.-,.
".-IC "If
"
I Ii
I , I ~
I .;/;~:..
~~ "
1-.-.
.-"I~'
"; &::::~
'--_._~
- I..!'I1I!!I- ~"2'.s
';=:-'--:C.=."COc:.' " -r --.
..S.'~-_\1
t., !:.l'U~'~_~O~'
w.,.
..'0
L~'I"iI:,.,
_..
+
~~""_-..b
.- '!.'!~.u"!-
, ,
,.
~
j: i~
. , - .-' ----'-'----1'
1~' --;,.-:.-.1
~I
.LfWIl
...,
10-1.
WIOJIMf( SIIU'AC( -n. AID UOUI[Wf W81\1.
lOCA'IQI .. "'(1
_.tA
$
ow.,.
(lIU,." IOIIIc.,., WIll lDCAflca .uo ""1
"OWClic. WilL lOCATlCII MO ....1
UILItCMD TUCkS
c::::J
'''WIO IOAO
AI'f'ttOIIMI( IM:ILlO MU ICUCIM.
MQIfS
I. lAS' IW Cl:n:UftD ,......, ,..1.." - tin 0&..[1$
ClIIU. Sn8CIS. a.ouCM.
I. UlStIIC IIlllCItATHJIS AlII UnAJUIIS tI tv. l1li. N. I, ,
- 1f.1 t8IUQI ""I SOlD MIlLS .. MIOD 'I""""
.. .. 0ICD8(1. .., II "ID - -. ... ..,1..
a.onliM. .U lDUHCId .. WYAnl8l ,. ~UI ...-
..m .. II-ns tI8I8 ~J1 Ulla mu ... SI8InO II
IlII08O, 1- .. Wll lOCl'.- - noan- " ~IDD. 1f..1lD
MIl If-JtS h8IQ 1f-4tS ShIU "'U to( SIII'I(tU II CICTI8I1
... If Anu. LEVIS. ".S" "". IIC.," -. IIIOI.CU.
IUJ. ~ is """.Mm' UX'AtD.
J. 1..1 II8GIiII I-S .. '-1 All 18111".. -YO .'1'" IOU.
4. N-I TIII:U:It N-' All "'.'IM. *1'0 ...., cu.
S. IV SOlD "':us.. IIStAUD It .. cln tI S1\I8U RII
fIIIIX1'l. WU SITI run.. "..1. ""'1. l1li..4""""
IUUlUD AS """8.. 118:lU ,. S18IJI ,...., CWo "'-I.
l1li..1 .. ., ..U .StS 18ULLa AS _I_I. WiU r.
MMI' UlCaA18IU IDSI 01'111.. 8-.. ".. ., - Qt-,
.. IIIStAUD AS 01.'. IIIUS nil .. (In" '...IS.
... r--. "'.., SOlO IIO.U IBIAUD " '''''...fI. 1101.
...acr ... ,. MIIl1'II 05118118 IIC. S01U8IIA8 etrn8(I,
..,. "'110 fIIICUiiIIl-JIS ""US .... ...., SOlD mu
10( IIStAUIO n .. MNI - ISSGCIA1U .. 8IlII (..1-
(01. ....-. ""'. MlillSI - KPfD8O. I.. "'..II..UO
.. .. I18UII ..,," SOIU tnU.. I8ITALUI" ... .hlS
.. AlSatIAnl,. UllII 0111011. ... .r. .... - ""r. ...
.. "I nIIID 11-111 SOUl CLU .. IISI.tUD 1181.
fIMSI I " .. IDCDIAL 'lIWUTIunca (II) 'I ItIJ. ""UD
r-.. ..us ,. "'US l18l8I 11-" suln tIIUS ...
IIDALlO DIll.. II8W: II C'I - I.. "'.. .... - II-IIS
....... II-QS sriiO wru.s .. IIISTIUD -I. NUl ..1
.,. I..
~
north,
o ... 1208
l..J"1.1"l___J
SCALI . FUT
-------
4
2.0
SITB HISTORY ABD BBPORCBKBBT ACTIVITIBS
In 1982, during routine chemical testing of the municipal
water supply, the Michigan Department of Public Health (MDPH)
identified the presence of two volatile organic compounds (VOCs) in
the municipal water supply. The identified VOCs were
trichloroethene (TCE) and tetrachloroethene (PCE). TCE and PCE are
organic solvents commonly used-as degreasing or cleaning agents in
dry cleaning, metal fabrication and other industrial or commercial
applications. The two affected wells were wells PWl and PW2 (see
Figure 1). Water samples had been collected from the city
production wells since 1955 for analysis of i~organic constituents
by the MDPH on an irregular basis. However, VOC analysis was first
performed in June 1982, and TCE was detected in samples from wells
PWl and PW2. since then, the production wells remaining in use
have been sampled by MDPH and/or MDNR at least on a yearly basis
for VOCs and indicator parameters.
In August and september 1982, MDPH suggested that the city
discontinue using the two contaminated municipal wells, undertake
an investigation in an attempt to locate the source of VOC
contamination, and explore the possibility of locating alternate
well field sites. In October 1982, the city initiated attempts to
identify the VOC source area and commissioned Gave Associates to
identify the source of contamination. In December 1982, the city
began the process of identifying additional well field sites. A
rise in TCE and PCE was noted in samples collected by MDPH from
wells PWl and PW2 between 1982 and 1983. Between June 1982 and May
1983, TCE concentrations in well PW1 increased. Pumping at these
wells was stopped in 1983. By May 1983, the VOC source area
investigation by Gove Associates had ended with the city unable to
locate the contamination source or plume. The city then decided to
increase the pumping capacity of its remaining two wells, wells PW3
and PW4. No VOCs were detected in samples from well PW3 between
May and December 1983. During this period, the pumping of well PW3
- was increased to replace the lost production from wells PW1 and
PW2. Well PW3 provided approximately sot of the water supply by
1984 (approximately one million gallons per day).
In November 1983, an industrial process well (R4) at Ross
Laboratories on the north side of the city was found to contain
detectable concentrations of TCE. By 1985, three of the five wells
owned by Ross Laboratories (R1, R3 and R4) were contaminated with
VOCs. This problem area has been referred to as the W. Lafayette
street Area in MDNR records. The industry responded by installing
a new well north of the plant (well - R5). Although the new well
has no detectable VOCs, the industry is using carbon adsorption
treatment for water which is used for consumption and as ingredient
water in its manufacturing process.
In April 1984, the city began using groundwater pumped from
-------
5
the newly constructed Oaklawn well (PW5). Pumping had previously
been discontinued at wells PW1 and PW2, except for emergency needs,
and increased at wells PW3 and PW4 to compensate for the loss of
PW1 and PW2.
In June 1984, U.S. EPA Region V issued a confidential report
detailing research into the identity of potentially responsible
parties (PRPs) for the Site contamination. Included in the report
was a-'description-- of source of infor.ation, -. site backqround
information, ::cations of former landfill sites, possible waste
generators anr: a chronology of dry. cleaner - and manufacturing
f=.ciliti ~s. l=:-:::.:1ted L, Stu::,gis since approximately ~soo. The report
was the basis for conducting the inaus~rial survey and Phase I of
the RI.
In January 1985, TCE was detected at or above proposed
drinking water standards in well PW3. Groundwater flow and
contaminants may have moved from one well to another, as the center
of pumping was shifted from wells PW1 and PW2 to well PW3.
Therefore, use of this well was decreased, such that PW3 would be
used only in cases of emergency.
In Auqust 1986, Warzyn Engineering, Inc., began the RIfFS of
the site, under the auspices of MDNR. The information provided
herein results fro. a field investigation conducted as part of the
RI between Auqust 1987 and January 1990.
until recently, the city relied on wells PW4 and PW5 to supply
water needs to local consumers. In February and March 1989, the
city drilled and tested a new municipal well location. The
Thurston Woods well (PW6) became operational in June 1989.
Because no PRPs were identified at the commencement of the
RIfFS, no parties were sent Special Notice Letters. The U.S. EPA
anticipates sending Special Notice Letters to currently identified
PRPs in fiscal year 1992.
3.0
BIGBLIGBTS O. COIIIIUIII'l'Y PU'fICIPA'fI08
A community relations plan was developed in 1987 to document
communi ty concerns and to plan an information strategy. MDNR beld
four public meetings to keep the public informed about the
activities at the Site. MDNR also sent out fact sheets and letters
at various times during the RIfFS process.
As part of its community relations program, MDNR maintained an
information repository at the Sturgis Public Library. The library
is located at 130 N. Nottawa. All formal reports developed during
the RIfFS are available at this location. The repository also
contains documents prepared by MDNR, such as fact sheets and the
Proposed Plan.
-------
6
MDHR and U. s. EPA notified the local community, by way of the
Proposed Plan, of the recommendation of a remedial alternative for
the sturgis Municipal Well Field site. To encourage public
participation in the selection of a remedial alternative, MOHR and
u.s. EPA scheduled a public comment period froa June 10, 1991, to
July 10, 1991. At the request of Cooper Industries (the parent
company of Kirsch Co.), the comment period was extended by 30 days
to August 10, 1991.
Additionally, on June 20, 1991, MDNR and u.s. EPA held a
public meeting to discuss the recommended remedial alternatives and
the ot!'ler alternatives identified and evaluated in the FS. A
t::-anscript of this meeting is included as part of the
Administrative Record for the sturgis Municipal Well Field site.
u. s. EPA I S responses to comments received during this public
meeting and to written comments received during the public comment
period are included in the Responsiveness Summary which is attached
to this ROD.
Press releases, announcing the public comment period, the
public meeting and the availability of the Proposed plan, were sent
to the sturais Journal, the sole local radio station, and the local
cable television company. Press releases announcing the extension
of the public comment period were sent to the above media contacts,
the Kalamazoo Gazette and the South Bend Tribune.
Advertisements were placed in the sturais Journal concerning
the extension of the public comment period.
4.0
scon UD IIOLB 01' DSPOBS. AC'lXOB
The selected remedy for the sturgis Municipal Well Field site
is intended to be the final response action at the site. The
remedy will combine treatment of groundwater, treatment of VOCs in
the source area soils, excavation and disposal of residual PAR
contaminated soils, and long-term groundwater monitoring. The
principal threat (VOC contaminated soils) at this site will be
.treated using soil vapor extraction and the low level threat (PAR
contamination) will be contained off-site.
-------
7
5.0
81J111m8Y O. 8%9 CDDCTBRJ:8'1'%C8
Groundwater Contamination
The primary aquifer contaminants are TCE and PCE. TCE'
concentrations range up to 17,000 ppb and PCE concentrations range
up to 150 ppb. The primary source of this contamination is the
Kirsch property. The contamination from Kirsch enters the shallow
aquifer, follows the groundwater flow to the southwest, and flows
through a window in the till units to the deep aquifer. TCE and
PCE contamination'in the groundwater is so extreme'in some areas
that the groundwater is a characteristic hazardous waste under the
Resource Conservation and Recovery Act (RCRA). No information is
available to indicate that RCRA listed wastes are present.
The Wade property is also contaminated with TCE and PCE and
has been found to contribute to contamination in the aquifer. As
with the Kirsch property, no information is available to indicate
that RCRA listed wastes are present. The RI has shown that
contamination from the Wade property enters the shallow aquifer,
moves east/southeast, and flows through a window in the till units
to the deep aquifer. This contamination is also pulled toward the
active water supply wells.
contamination has been detected in the public water supply
wells PW1, PW2, PW3 and PW4 (TCE was found during 1 sampling round
in PW4, yet subsequent sampling by Michigan Department of Public
Health did not detect the contaminant). contamination has also
been detected in the sturgis Foundry Well (F1) and Ross Wells R1,
R3 and R4. The volWle of contaminated groundwater at the site has
been estimated to be 4.79 trillion gallons. Based on a 10 ppb
contour, the area underlain by contaminated groundwater is
estimated to be 496 acres.
Compounds of the chlorinated ethane class (1,1;1-
trichloroethane, 1,1-dichloroethane, and 1,2-dicbloroethane) were
less frequently, but consistently, identified in groundwater
samples from some locations. The most frequently detected of
these, 1,1,1-trichloroethane, ranged in concentration fro. 0.3 to
9.0 ppb. A number of additional volatile organic chemicals were
infrequently identified in groundwater samples. Compounds of the
trihalomethane class (chloroform, bromodichloromethane, and
dibromochloromethane) were identified at low concentrations (2.0
ppb or less) in municipal, industrial and monitoring wells.
Identification of these compounds in municipal wells occurred only
in samples collected from Round 3 and may be representative of
residual chlorinated water.
Groundwater samples from selected locations at potential
source areas and from municipal water supply wells were analyzed
for Target Analyte List (TAL) metals and cyanide to evaluate the
inorganic quality of the groundwater in these areas. Levels of
-------
8
barium, chromium, iron, manganese and nickel were slightly elevated
from site-specific background levels.
Levels of cyanide from groundwater samples at the Kirsch
source area were also elevated (maximum 384 ppb). The property'was
used for industrial processes since before 1920. Drawings of the
former buildings on the property show a cyanide/metal reprocessing
building. Therefore, the presence of cyanide at these wells is
probably due to releases from points of use on the property or at
the reprocessing building.
soil contami~ation
The primary contaminants in the Kirsch property soils are TCE,
PCE and PARs. Concentrations in so.e of the shallow soils (ranqinq.
up to 260,000 ppb PCE and 99,000 ppb TCE) are high enough that the
soils may be considered a RCRA characteristic hazardous waste if
excavated. The PAR contamination is widespread through the shallow
soils on the property. Total PAR concentrations ranqe up to 61,200
ppb. The source of the PARs is unknown (typically PARs are found
in products produced from coal tar or can be formed by incomplete
combustion in burning). Approximately 10,000 cubic yards of soil
are thought to be contaminated with PARs in excess of 330 ppb.
PARs were not detected in groundwater.
The composition of VOC contamination in the soils at Kirsch
varies across the property. For example, at location 58-06, 97.2'
of the contamination in the 2.5-ft depth sample is from TCE, while
at Wl15, 96.9' of the contamination in the 1-ft depth sample is
from PCE. Two VOCs were detected in surface soils (0.5 ft) at very
low concentrations: 1,1,1- trichloroethane (2 to 4 ppb) was
detected in four of eight samples and TCE (2 ppb) was detected in
three of the eight samples collected.
The primary contaminants in Wade soils are also TCE, PCE and
PARs. contamination is primarily limited to the southeastern
portion of the property. The highest concentrations (630 ppb PCE
and 160 ppb TCE) were observed in the area beneath where the
underground tanks had been removed. The PAR contamination is
primarily limited to the eastern portion of the property. Total
PAR concentrations range up to 5,590 ppb. Approximately 890 cubic
yards of soil are contaminated with PARs in excess of 330 ppb.
-------
9
Surface Water/Sediment contamination
Surface water and sediment samples were collected from gravel
pits, used as storm runoff or effluent discharge points. Samples
collected were at Ross Labs (1 sediaent and 1 vater sample), Kirsch
Co. Plant No.2 (2 sediment and 2 vater saaples), and the Sturgis
Foundry Corporation (1 sediment and 1 water sample). Surface water
and sediment samples were collected to assess the potential impact
of industrial processes on surface water and. subsequently on
grcundwater by recharge of the aquifer by surface water. CLP
analysis detected methylene chloride in three of the four sedim~nt
sa:nnles and i::. the laboratory ~lank. This conta~ination is
bel.leved- -;t.crt-be""8 laboratory contaminant. Sediment and surface
water samples from Ross Labs and the Kirsch Company Plant No.2 did
not have other detectable levels of VOCs. Based on this surface
water and sediment saapling, these vater sources probably do not
contribute to the aquifer contamination. TCE vas detected in the
surface water sample collected from the Sturgis Foundry Corporation
seepage lagoon (5.0 ppb). However, the TCE in the Sturgis Foundry
Corporation surface water sample is probably due to the presence of
TCE in the foundry vater supply (95.1 ppb). As indicated by
dovngradient well results, this lagoon is not a source of
significant VOC contamination.
'.0
S1JJIIIaJt1' O. SI'III aISO
The Baseline Risk Assessment in the RI Report (chapter 10),
followed the guidance providec:l in u. S. EPA' s Risk Assessment
Guidance for. SUDer fund (RAGs): Volume I. Human Health Evaluati~n
Manual. Risk assessment quidelines developed by the state of
. Michigan were also appliec:l.
Chemicals considered in the Baseline Risk Assessment are those
which are present as a result of chemical releases which have
occurred at the Si te and are termec:l -cheaicals of potential
concern-. To identify these, cheaicals present in soil and
groundwater samples are distinguishec:l from those which. may
naturally be present (Site backg'round) and those which can be
unintentionally introduced into samples through sample collection
or laboratory analysis. Further, consideration is given to the
frequency of occurrence of the chemical at the Site. Those
infrequently identified may not be si9ftificant in view of overall
Site contamination. Chemicals considered to be of potential
concern are evaluated further in the risk assessment.
An exposure assessment is performed to identify actual and
potential pathways by which human exposure to contaminated Site
media may occur. The assessment considers factors such as the
physical location of contaminated areas, the type of contamination
and the populations which may come into contact with these areas.
Exposure pathways are identified for two Site land use scenarios,
-------
10
pathways based on land use practices as they currently exist, and
potential pathways based on land use changes which may occur in the
future and result in additional types of exposure. Both current
and future pathways which represent possible exposures are then
quantified to estimate the magnitude of daily contaminant exposure
a population may incur. To accomplish this, assumptions pertaining
to the exposed population are made, such as the nature of the
individuals (e.g., child vs. adult), the rate of contact with the
contaminated medium (e.g., adult consumes 2 liters water daily)
and the length of time the exposure is likely to occur (e.g., years
V5. lifetime). These population variables are then combined with
chemical concentration data to calculate a level of exposure.
The approximately five square mile area where contamination of
the groundwater by hazardous substances has been detected
encompasses both industrial and residential areas. The groundwater
is used for the sturgis City drinking water supply: therefore,
exposures based on drinking and dermal absorption were used to
estimate the risks posed by the groundwater. The sources of
hazardous substance contamination of the groundwater are the soils
at the Kirsch and Wade properties. The Kirsch soils source area is
primarily a vacant lot that is near the industrial Plant NO.1, but
residences are also located in the immediate vicinity of the
property. The Wade property is currently being used - for a
commercial purpose, but is also bounded by residences. Therefore,
exposures based on reasonable future residential land use were used
to estimate the risks posed by these source areas.
Pathways considered to be most significant at the Site include
exposure through groundwater use and direct contact with soils,
summarized as follows:
CUrrent Land Use Conditions:
Exposure of city residents to contaminated municipal
water by drinking and through dermal absorption while
bathing.
.
.
Exposure of children. to contaminated surface soils
through incidental ingestion and dermal absorption while
playing in source areas.
Exposure of city residents to volatile contaminants in
ambient air released from soils.
.
-------
11
Potential Future Land Use Conditions:
.
Exposure of future residents to contaminated groundwater
resulting from either installation of a well within the
contaminant plume or by miqration of qroundwater
contaminants to existing wells. Exposure may occur
through drinking and dermal absorption. Contaminant
concentrations are assumed to exist in the future as
under current conditions.
"Exposure "of individuals to contaminabid" soils at' a""~fu'ture
residence developed at the source areas. " Exposures may
occur through incidental ingestion -' of soil and derma-l
absorption. It is ass\Dted contaainants in either surface
or subsurface soils at current concentrations are made
available for exposure as a result of Site development.
using these scenarios, risk numbers are calculated for each
contaminant. These calculations factor in the amount of exposure
assumed, the dose of the chemical received (based on the
concentrations found during the RI), and a toxicity esttaator for
each individual chemical which quantifies the toxicity of that
chemical. Different constants and equations are used based on"
whether or not the chemical is carcinogenic. The constant for a
carcinogenic chemical is called a slope factor, and the constant
for a noncarcinogen is called a reference dose.
The results of these calculations are estimates of cancer risk
for carcinogenic risks and estimates of Hazard Indices for
noncarcinogenic risks. The cancer ri8k number is expres8ed in
scientific notation and represents an estiaate of an individual's
increased risk of getting cancer over a lifetime. The carcinogenic
risk estimate is generally a conservative estimate, i.e., the risk
may be less than predicted. For example, 1.0 x 10-brepresents an
increase in an individual's risk of cancer by 1 chance in a
million, under,.,the exposure conditions assumed. u.s. EPA considers
this 1.0 x 10--number as a point of departure when determining risk
at a Site. Risks calculated to be less than this value are
considered protective qf human hea1~ and the environment, while
risks between 1.0 x 10-"'and 1.0 x 10-Octre within a range acceptable
to u.s. EPA but may not be considered protective due to site-
specific conditions. Risks greater than 1.0 x 10-4are generally
unacceptable.
The Hazard Index (HI) represents the risk of adverse non-
cancer effects occurring due to exposure to the Site. The HI
number generated is interpreted differently from the cancer risk
number. To evaluate risk at a site due to noncarcinogenic
contaminants, u.S. EPA has determined that an HI less than or equal
to 1 estimates that no adverse effects are likely to occur due to
the hypothetical exposure, while a Hazard Index greater than 1
-------
12
estimates that adverse effects due to site exposure may occur and
signals that potential risks to human health must be carefully
evaluated. .
Tables 1,2 and 3 summarize the cancer risk numbers and HI
values calculated for each chemical under the current land-use
scenario. Tables 4 and 5 summarize future residential risk and
hazard index values. The numbers listed in these tables represent
the reasonable maximum exposure conditions by using the greatest
concentration of a chemical found in each medium or the 95 percent
upper-bound confidence limit of the arithmetic mean. The
cumulative risk for each scenario is included beneath each table.
In summary, the risk assessment hignlights the following risk
estimates at the site (refer to Tables 1 through 5):
.
The risks to childre% playing a, the Kirsch and Wade
properties are 2 x 10- and 2 x 10- ,respectively.
The current risk to residents from exposure to volatile
coniaminants in ambient air released from1F.0ils is 6.9 x
10- near the Kirsch property and 1.lx10- near the Wade
property.
.
.
The future risk from the residential use of con~aainated
groundwater from the Kirsch property is 6 x 10- . The HI
from residential use of contaminated groundwater is 1.
The risk from fut~re residential exposure to contamina!ed
soils is 5 x 10- at the Kirsch property and 7 x 10- at
the Wade property.
.
Actual or threatened releases of hazardous substances from
this Site, if not addressed by implementing the response action
selected in this ROD, may present an imminent and substantial
endangerment to public health, welfare, or the environment.
-------
13
-Table 1
CHEMICAL HAZARD INDEX ESTIMATES FOR
CURRENT LAND USE EXPOSURE SCENARIOS
CHILDREN PLAYIIiG 011 CONPAIIY PROPBltTY - COJr.rACT WI'rII SURPACB BOILS
Ch..ical. Potentially
C8u.ing lIoncancer Health
Hazard.
1, 1'~ :'-'!'~ehl~oethl!l.ne - ~
AGenaph:. :lene
~ .0
-
Anathraeene
Fluorene
Fluoranthene
Naphthalene
Pyrene
Bis(2-ethvlhexvl)Dhthalate
Butvlbenzvlphthalate
Di-n-butvlphthalate
Cyanide
~l
(UUCII) '.
1, 1, l-Trichloroethane
Tetrachloroethene
Anthracene
Fluoranthene
Naphthalene
Pvrene
.
Bis(2-ethvlhexvl)Dhthalate
ButvlbenzvlDhthalate
Di-n-butylphthalate
Total
(WADE)
Hazard OUotient
Incidental
Ingestion
Hazard OUotient
Dermal
Absorption
KIRSCH
_.~-
,
T
I
CO. PLAIIT NO.1
"""1. 7 e-OS'"r- --
!
'? Sa-Do!
. .-
~'-
-------
14
Table 2
CANCER RISK ESTIMATES FOR
CURRENT LAND USE EXPOSURE SCENARIOS
CHILDREN PLAYING ON COMPAHY PROPERTY - CO~C'r WITH SUR.PACB SOILS
Potential Carcinoaens Incidental Dermal Total Exposure
Ingestion Risk Absorption Risk
KIRSCH CO. PLAN'l' NO. 1
i 1.2e-12 !
T=i-::hloroethene I 3.8e-13 i
3is(2-ethylhexyl)phthal~te 1.Oe-lO 1. Se-ll
Total CarcinoQenic PAHS 1. 7e-05 2.4e-06 i
PCB 6.3e-07 9.2e-08
.. . .. ... ..I.:i...."'H
Total (1tIUCK) 1.88-05 .....2i5.~O' . .. .. .H.:" 28-06
WADE ELECTRIC CO.
Tetrachloroethene 1.7e-11 1. 2e-11
Bis(2-ethylhexyl)Dhthalate 3.1e-10 4. 5e-11
Total carcinogenic PAHS 1. 7e-06 2.5e-07
Total (WAD) 1.7.-06 2.5.-07 2.-06
-------
15
. Table 3
CANCER AND NONCANCER HEALTH RISKS ASSOCIATED WITH CONTAMINATED AIR
CURRENT LAND USE EXPOSURE SCENARIO
USIDBIITIAL BDOSURB m AllBIBIIT AIR
Ch_ical of Inhalation Cancer Risk
Potential Concern Hazard Quotient
~.i.rsch Co. Plant Ne. 1
I I
~ T~~r~c~.croe~hene -- 5.7e-07
fr--o -. -. ' ~-_.., -~..'. j
l,l,l-Trichloroethane 1.2e-09 --
Trichloroethene -- 6.3e-06
Total (KIRSCH) 1.2e-09 6.ge-06
Wade Blectric Co.
Tetrachloroethene -- 6.6e-08
Trichloroethene -- 4.3e-08
Total (WADB) -- 1.1e-07
-------
16
Ta))18 ..
CHEMICAL HAZARD INDEX ESTIMATES
FOR POTENTIAL FUTURE LAND USE EXPOSURE SCENARIOS
RESIDENTS UTILIZIHG COH'rAIIIHATSD GROUIIDWATBR PROII A PRIVATS WELL
Cheaicala Pot.ntially Hazard OUotient Hazard OUotient Total Exposure
Cauaing Honcancer Health Incidental Dermal Pathway Hazard
Haaarda Ingestion Absorption Index
Bromodichloromethane 2.ge-03 4.1e-06
Chloroform 4.6e-02 6.6e-05
Dicr~mo=hlcrome~hane 1.4e-03 2.1e-06
,I l,l-C~ch!oroethane 5.7e-04 8 .'e-07
1,2-Dichloroethene(total) 2.4e-02 3.5e-05
Tetrachloroethene 4.3e-01 6.2e-04
1,1,1-Trichloroethane 2.ge-03 4.1e-06
1,1,2-Trichloroethane 5.7e-02 8.3e-05
Bis(2-ethylhexyl)Dhthalate 7.3e-02 2.1e-04
Barium 1.0e-01 2.ge-03
Chromium 5.4e-01 4.1e-05
Cyanide 3.5e-01 7.3e-04
.. . .
Ta.rAL 1.1.+00 4.7.-03 1.+00'
-------
17
Table 4 (continued)
LIPBTIII8 SXPOSURB OP P'UTURB ~8IDmr.rs TO (:OII'DIII~ SOILS
Che.ical. potentially Hazard OUotient Hazard OUotient Total Exposure
cau.ing Boncancer a.alth Incidental Dermal Pathway Hazard
aazard. Ingestion Absorption Index
KIRSCH co. I'I.oAft 110. 1
Total 1,2-Dichloroethene 1.5e-04 1.5e-04
2-Eutanone 3.3e-06 3.3e-06
i l,:,l-Trichlcroethene 1. 4e-07 1. 4e-07 I I
i
I !
- ~ -
etrClc:-.loroet:.ene
3 -. 3e-O" I
3 . ...-=-02 j
e- . -
Naphthalene 4.6e-04 9.2e-05
Fluorene 4.2e-05 8.3e-06
Anthracene 1.2e-05 2.4e-06
Di-n-butylphthalate 2.4e-06 4.ge-07
Fluoranthene 3.0e-04 5.8e-05
Pyrene 3.6e-04 7.2e-05
Butylbenzylphthalate 2.ge-06 5.7e-07
Acenaphthene 3.2e-05 6.5e-06
Bis(2-ethylhexyl)phthalate 8.4e-06 1. 7e-06
Chromium 1. 6e-0 1 3.3e-02
Cyanide 8.68-03 1. 78-03
Zinc 1.3e-01 2.6e-02
m'%AL (URaCIl) 3.4.-01- --f ~4e-02 -'-- ,..01
WADS BLBCTRIC co.
Chloroform 5.4e-06 5.4e-06
2-Butanone 4.4e-07 4.48-07
Tetrachloroethene 8.1e-05 8.1e-05
Chlorobenzene 9.6e-07 9.6e-07
Naphthalene 8.2e-05 1. 7e-05
Anthracene 1.0e-06 2.1e-07
Fluoranthene 3.ge-05 7.7e-06
pyrene 4.2e-05 8.4e-06
Bis(2-ethylhexyl)Ohthalate 2.6e-05 5.2e-06
Butylbenzylphthalate 4.7e-07 9.4e-08
Di-n-butylphthatate 9.ge-07 1.ge-07
TCnAL (WADS) 2.88-04 1.38-04 4e-04
-------
18
Table 5
CANCER RISK ESTIMATES
FOR POTENTIAL FUTURE LAND USE EXPOSURE SCENARIOS
RBSIDDTS UTILIZIIfG CONTAIIIIfADD GR01JllDWAmR P'ROII A PRIVAD WBLL
Potential C8rcinocren8 Incidental Dermal Total Exposure
Ingestion Risk Absorption Risk
Benzene 1. 7e-06 1.2e-06
Bromodichloromethane i.4e-06 1.1e-08 ,
I I
Chloroform 2 .8e-06 i 4.02-09 .
;
D~bromochloromethane 2.4e-06 3.6e-09
1,2-Dich1oroethane 2.6e-06 3.8e-09
Tetrach1oroethene 2.2e-04 3.2e-07
1,1,2,2-Tetrachloroethane 1.7e-06 2.5e-09
1,1,2-Trich1oroethane 1. 3e-05 1.ge-08
Trichloroethene 5.3e-03 7.7e-06
Bis(2-ethylhexyl)phthalate 2.0e-05 5.ge-08
~ 5.6.-03. ..9. Ie-CM .:).... H(HH H '.";03 .
Lifetime Exposure of Future Residents to Contaminated Soils
KIRSCH CO. PLAIr.r 1fO. 1
Trichloroethene 1.4e-06 1. 4e-06
Tetrachloroethene 1.7e-05 1.7e-05
Bis(2-ethylhexyl)phthalate 2.4e-09 4.7e-10
Total carcinogenic PAHs 4.1e-04 8.3e-05
PCBs 1. 5e-05 3.0e-06
~ (ItIRSCB:) ,.4.-0,-1. 1.0.-04H ..2::.:-.....- $.;'0' ...
WADE BLBCTRIC CO.
Vinyl Chloride 1.8e-08 1.8e-08
Chloroform 3.3e-10 3.3e-10
Trichloroethene 2.3e-09 2.3e-09
Tetrachloroethene 4.1e-08 4.1e-08
Bis(2-ethylhexyl)phthalate 7.3e-09 1.5e-09
Total carcinogenic PAHs 5.ge-05 1.2e-05
~ (WADI!:) 5.ge-05 1.2.-05. 7.-05
-------
19
6.1 UBCBRTAX8TZBS
The Risk Assessment could not quantify risks from those
contaminants without known slope factors or reference factors.
Methods to quantify risks and possible synergistic effects due to
exposure to multiple contaminants or multiple pathways are very
limited. The use of risk additivity helps prevent the
underestimation of cancer risks or potential noncancer health
effects. - -
_.- _.
.-
.-
.."
- .. .--
. . ~ . ~
...,; .~_.....- -r =-
, . 2 DlVIROHKBJi'1'AL RISD
Although no formal quantitative ecological assessment was
believed to be necessary, the potential effects to the environment
were qualitatively evaluated.
The identified sources of contamination at the Site are within
the City of Sturgis. These areas are located adjacent to
industrial or cOJlDDercial enterprises and do not constitute valuable
ecological habitats. At present, the plume of contaJlinated
groundwater is confined to the city as a result of current water
use practices. Discharge of contaminated groundwater to surface
water bodies is also not occurring. If a change in water
utilization practices would occur in the future, such that the
natural flow gradient would control groundwater movement,
contaminated groundwater could be discharged to surface water
bodies which are located outside of the city. If this were to
occur, ecological habitats in these areas may be at some risk.
However, this scenario appears unlikely, because the City of
Sturgis relies on groundwater as its source of water. In addition,
the adverse impact of groundwater contaminants on wetland or
riverine habitats would be dependent on other factors, including
toxicity of the chemical on wildlife, the ultimate chemical
concentration in these areas, and the chemical's persistence in the
environment.
7.0
DB8CRIPfI08 01' ALDltDTIVB8
Based on the resul ts of the RI, the U. S. EPA and MOHR
conducted a Feasibility Study (FS) to identify and evaluate
remedial alternatives that would minimize or eliminate the health
risks caused by site-related contaminants. The goals for remedy
selection at the Sturgis Municipal Well Field site include:
.
Selection of a remedial action that would contain, purge, or
otherwise minimize the health threat posed by contaminated
groundwater:
.
Selection of a remedial action that would minimize exposure to
-------
20
contaminated surface soil; and
.
Selection of remedial action that is protective of human
health and the environment.
Three alternatives were developed in the FS to address soil
contamination; four alternatives were developed to address
groundwater contamination. These alternatives are described in the
text which follows. For more detailed information, see the FS.
The major applicable or relevant and appropriate requirements
(ARARs) which were identified for these alternatives will be
discussed in Section 8.0 - Comparative Analysis of Alternatives,
under ComDliance with ARARs.
Alternatives For Contaminated Groundwater
Alternative awl - 80 Action
As required by the National Contingency Plan, the No Action
Alternative is evaluated. The only action that would be taken
is water quality monitoring of the public water supply wells
and the existing groundwater monitor wells. .
The groundwater monitoring activities would continue for at
least thirty years. Since there is no new construction, there
would be no initial cost to implement this alternative.
Alternative awl "80 Action"
C08t.
Capital Costs
0 , M per year $ 200,000
Net Present Worth $ 2,085,000
Alternative aw2 - Water Supply Tr.ataent
This alternative utilizes treatment of the city water supply
prior to water entering the distribution system. Treatment
could include Air Stripping with Vapor Phase Granular
Activated Carbon Adsorption ("Air Stripping"), Liquid Phase
Granular Activated Carbon Adsorption ("Liquid Phase Carbon
Adsorption"), or Chemical Oxidation. Residual carbon from Air
Stripping and Liquid Phase Carbon Adsorption or sludge from
chemical oxidation may need to be treated to meet RCRA Land
Disposal Restrictions ("LDRs").
-------
21
This alternative would also include a 8Onitoring proqraa and
institutional controls. The monitoring program would track
the movement of contaminants in the groundwater to assure that
the treatment system would be effective. Institutional
controls would consist of st. Joseph County Health DepartJlent
restrictions on locating private water supply wells in the
city.
The on-line treatment system for this alternative is
relatively simple to construct and could be completed in six
months. However, since ~here is no treatment of groundwater
except as it :..s being drawn in for municipal use, the on-line
system would have to remain in op~ration .fpr as long ~ as
extracted groundwater remains contaainated. Costs for this
alternative depend on the treatment technology used and are
based on the assumption that the syetea would be in use for
thirty years.
Although it is expected that carbon residuals fro. Air
stripping and Liquid Phase Carbon Adsorption can meet LDRs
through regeneration, a Treatability Variance for contaminated
carbon residuals would be established.
Al ternati ve GW2 Air Liqui4 Pha8e Cheaiaal
Coata stripping Carbon OZi4ation
&48orption
Capital Costs $ 3,150,000 $ 3,930,000 $ 8,050,000
o , M per year $ 445,000 $ 529,000 $ 827,000
Net Present Worth $ 7,450,000 $ 8,940,000 $ 16,000,000
Alternative GW3 - Pluae KaDag..ent
The obj ecti ve of Al ternati ve GW3 is to prevent the moveaent of
the contaminant plume into the city public water supply wells.
Although this alternative would be designed to contain the
groundwater contamination and not restore the aquifer, it is
possible that Alternative GW3 would eventually allow
groundwater to meet ARARs throughout most of the plume because
of the large volume of groundwater that would need to be
extracted and treated. This alternative will include the use
of both existing and new wells for extracting groundwater from
the contaminated aquifers. The extraction process would be
carefully designed to control the spread of contaminated
groundwater. Approximately seven extraction wells would be
needed to withdraw contaminated groundwater from the aquifers.
Treatability tests would be conducted during design to develop
the treatment technology to a point where the design can be
-------
22
developed.
Possible treatment technologies considered in the FS include
Air stripping, Liquid Phase Carbon Adsorption, and Chemical
oxidation. Once the groundwater i. treated, it would be
either discharged to a storm sewer in c08pliance with NPDES
requirements or returned to the ground through re-injection in
a closed loop system in compliance with Michigan Act 307.
Municipal use of treated water will also be considered and
would be investigated further during Remedial Design. The
groundwater treatment would continue !~til Michigan Act 301
Type B levels are achieved and mai ntained throughout the
plume. Treatment residuals may De RCRA characterist~c
hazardous wastes and require treatment under Land Disposal
Restrictions prior to disposal.
This alternative would also include a groundwater monitoring
program which would continue for at least thirty years and the
implementation of institutional controls. The monitoring
program would be designed to assure that contaminated
groundwater is being controlled and that the quality of the
public water supply is being maintained. Institutional
controls would consist of st. Joseph County Health Department
restrictions on locating wells in the city. Although this
alternative does not include on-line treatment of the water
supply, a contingency plan would be developed to allow for the
rapid installation of a mobile water supply treatment system,
should it be needed at any of the existing municipal supply
wells. The construction of the "plume management" system
would take approximately one construction season (spring,
SUIIIJDer, fall).
Costs for this alternative depend on the treatment technology
used and the location chosen for discharge of treated water.
The advective flushing time for the existing groundwater
contamination is between 5 and 15 years. However, actual
clean-up time would be substantially longer due to the effects
of retardation and dispersion. The retardation effects are
, especially troublesome for this alternative, and could cause
the cleanup time to double to 10 to 30 years. For costing
purposes, it is assumed that the system would be operational
for thirty years.
Although it is expected that carbon residuals from Air
stripping and Liquid Phase Carbon Adsorption can meet LORs
through regeneration, a Treatability Variance for contaminated
carbon residuals would be established.
-------
23
Alt Air stripping Liquid Ph... Cb88ioal
GW3 carbon ozidation
Adaorption
Coata StOl'll .elnject StOl"8 .elnject ItOl"8 ..tnject
Sewr Ian ..... Ian ..... tan
C8plt8l S 4.56 .. S 6.06 " S 4.95 .. S 6.45 .. . 10.6 " . 12.1 ..
o & ..
per yur 1503,000 S503, 000 S590, 000 S590,000 . 1.13 .. . 1.13 ..
MPW S 9.90 .. . 11.1, II . 10.5 .. . 12.0 " . 21.5 .. I 22.9 ..
~~
- .,.~..-.
. .~. -
~:"i
-' .... '-'" .~;:
- --
-
--~~~:,:.~.-' .6
Alt.rnativ. GW4 - Aquifer a..'o;.oratioD . - - 4.-- - ...... -:i~.~
(U.S. BPA aDd lIDD-r.OO88.nd.d groUD41rat.r ol.aDup alt.rD8tiv.
uainC) Air stripping .i th diaoharg. to . aton a...r)
The objective of this alternative is to achieve the efficient
cleanup of the aquifer supplying the public water supply
wells. This restoration would be achieved through withdrawal
of qroundwater from the new and/or existing wells, centered in
the .ost contaminated portions of the aquifer. It is
currently estimated that approximately seven extraction wells
would be needed to withdraw contaminated groundwater fro. the
aquifers. This alternative will also achieve so.e level of
control over .ovement of the contaminated groundwater,
especially if treated groundwater injection is utilized.
Treatability tests would be conducted during Reaedial Design
to develop the treatment technology to a point where the
design can be developed.
As withAl ternati ves GW2 and GW3, the extracted groundwater
would be treated with one of the following treataent
technologies: Air stripping, Liquid Phase carbon Adsorption,
or Chemical oxidation. Treated groundwater would be
discharged to a storm sewer in compliance with NPDES
requirements or returned to the ground through re-inj ection in
a closed loop system in compliance with Michigan Act 245.
Municipal use of treated water will also be considered and
would be investigated further during Remedial Design.
Groundwater treatment would continue until Michigan Act 307
Type B levels are met and maintained throughout the plume.
Treatment residuals may be RCRA characteristic hazardous.
wastes and require treatment under Land Disposal Restrictions
prior to disposal.
Alternative GW4 also includes a groundwater monitoring program
lasting for at least thirty years and implementation of
institutional controls. The monitoring program would be
designed to assure that contaminated groundwater is being
controlled and that the quality of the public water supply is
being maintained. Institutional controls would consist of st.
Joseph County Health Department restrictions on locating wells
in the city. As with Alternative GW3, this alternative does
not include on-line treatment of the water supply. However,
-------
24
a contingency plan would be developed to allow for the rapid
installation of a mobile water supply treatment system, should
it be needed at any of the existing municipal supply wells.
The construction of the "Aquifer Restoration" system would
take approximately one construction season (spring, summer,
fall).
As with Alternative GW3, costs for Alternative GW4 depend on
the treatment technology used and the location chosen for
discharge of treated water. The advective flushing time for
the existinq contamination in the aquifer is estimated to be
between 10 and 20 years. This estimate of advective flushing
time is p~obably an underestimate of actual aquifer clean-up
time, because of dispersion and retardation. Although
retardation would be present, its effects should be ainimal
and the time to cleanup the aquifer should not be
significantly compromised. Additionally, because this
alternative' places wells in the most contaminated areas, the
majority of the contamination would be removed within the
first five years. For costing purposes, it is assumed that
the system would be operational for thirty years.
Although it is expected that carbon residuals from Air
Stripping and Liquid Phase Carbon Adsorption can meet LDRs
throuqh reqeneration, a Treatability Variance for carbon
contaminated residuals would be established.
Al~ ft4 Air 8~rippiD9 Liquid Pha.. Chea10al
carboD OZida~ioD
Co.~. Ad.orp~ioD
Sto... .einject StOnl .elnject StOnl .einject
S-r ion S-r Ion s.er Ion
Cllpitel S 5.78 III S 8.95 III S 4.95 III S 8.14 III 110.59 III 113.80 III
o & III per
ye.r S598,000 S598,000 1998,000 1998,000 . 1 .13 III . 1 . 13 III
M"" ." .50 III 114.7'0 " 114.40 " 117.60 " 821.50 III 824.60 III
Alternatives For contaminated Soil
Al~.rDativ. 81 - .0 Ac~ioD
As required by the National contingency Plan, the "No Action"
alternative is evaluated. Under the "No Action" alternative,
Site actions would be limited to deed restrictions for future
development of the properties, and fencinq to control access
to source areas. This alternative could be implemented in one
-------
25
to two months, not including the time required to negotiate
deed and possible zoning restrictions. Cost estiaates are
presented below.
Alternative 81 "110 &otioD"
co.t.
capital Costs $ 73,000
o & M per year $ 3,300
Net Present Worth $' 0104/000
- _---,-07':'...J
...i- - -
-:-. .....:~ ~
- ,
- .
Alternative 82 - SOil Vapor BKtraotioD/ CODtai..8Dt
Volatile contaminants in the source areas at Kirsch and Wade
Electric would be remediated using soil Vapor Extraction
(SVE). SVE uses a partial vacuWl in wells to withdraw
volatile contaminants from soil. Activated carbon would be
used to capture the off-gases. Cleanup levels for volatile
contaminants would be determined using Michigan Act 307 Type
B cleanup levels. The remaining contaminants in the source
areas which cannot be removed through SVE would be covered
with a multi-layer cap. Two different caps were analyzed in
the Feasibility study.
.
cap A complies with Michigan solid waste regulation.. It
includes:
A 2-foot clay layer
A l-foot drainage layer
A 1.5 foot soil layer
A 6-inch vegetated topsoil
layer
.
Cap B complies with Michigan hazardous waste regulations.
It includes:
A 2-foot clay layer
A synthetic liner
A I-foot drainage layer
A 2-foot vegetated layer
The purpose of a cap is to reduce the amount of water
filtering through the waste, thereby reducing the amount of
contaminants getting into the groundwater. The primary
difference between Cap A and cap B is the addition of a
synthetic liner in Cap B. The liner is designed to provide an
additional barrier to keep water from entering the waste.
This alternative would include construction of a fence around
-------
26
the site to prevent site access during and after cap
installation. Restrictions would also be placed on the
property deed to control future use of the Site. A thirty
year inspection and maintenance program would also be
implemented to ensure that the cap i8 effective.
The construction of the SVE system and the multi-layer cap is
estimated to take one construction season. SVE remediation at
the Site is likely to take approximately 3 years. Costs for
achieving Michigan Act 307 Type B cleanup levels for volatiles
(using SVE) are presented below. Because this alternative
also utilizes a multi-layer cap to contain remaining
contaminants, costs are given for bo~~ Cap A and Cap B.
Although it is eXPected that carbon residuals fro. SVE can
meet LDRs through regeneration, a Treatability variance for'
contaminated carbon residuals would be established.
Alternative S2 SVE / Cap A SVE / cap B
Costs
Capital Costs $ 1. 45 M $ 2.81 M
o , M per year
during SVE $ 49,000 $ 49,000
~---------- --------- ~'---
o , M per year
after SVE $ 5,800 $ 5,800
Net Present Worth $ 1. 61 M $ 2.97 M
Alternative 83 - 80il vapor Bztraction/ Bzcavation
(U.8. BPA an4 XDKR-reco..en4e4 ao~l cleanup alternative utili8in9
80li4 .aate 4iapoaal of PAR contaainate4 aoil)
Volatile contaminants in the source areas at Kirsch and Wade
Electric would be remediated using SVE. SVE uses a partial
vacuum in wells to withdraw volatile contaminants from soil.
Activated carbon would be used to capture the off-gases. The
remaining contaminants in the source areas would be excavated
and either disposed of off-site or incinerated (burned) on-
site. In accordance with RCRA Land Disposal Restrictions,
excavated soils would be treated, as needed, prior to
disposal. Cleanup levels for the excavation would be
determined using Michigan Act 307 Type B cleanup levels. It
is estimated that 10,000 cubic yards of soil with low levels
of contamination would need to be excavated from Kirsch and
890 cubic yards of soil with low levels of contamination would
-------
27
need to be excavated at Wade Electric until the remaining
soils would be at or below cleanup levels. Excavated soil
would either be taken off-site and placed in a solid waste
disposal facility or incinerated (burned) on-s~te.
This alternative would include construction of a fence around
the site to prevent site access during the SVE and excavation
process. No restrictions would be placed on future use of the
site because the remedy will be protective.
The construction "o'f the SVE"'~'SY~"t:em"""l:!f.esti::\ated to"t:ake one'
;::~7istructi'.)n season. SVE remediation at the Site is likely to
-- 3DDroxi~atelv 3 v~ars. After ~~e SVE Drocess is
complete~ excavation of remaining con~aminants. could be
completed in an additional construction season. Costs are
presented below for SVE with excavation to Michigan Act 307
Type B levels. Please note that costs are also dependent on
the method used for disposal of excavated soil.
Although it is expected that carbon residuals from SVE can
meet LDRs through regeneration, a Treatability Variance for
contaminated residuals would be established.
Alternative S3 SVB with BEeavation to Type B
levels
Costs Off-site CD-Site
Disposal of Incineration
Soils of Soils
Capital Costs $ 2.18 M $ 6.33 M
o & M per year !\
during SVE and $46,600 !1 $46,600
Excavation ii
---------- ---- -----
o & M per year
after SVE and $ 0 $ 0
Excavation
Net Present Worth $ 2.31 M $ 6.46 M
8.0
COKPARA'1'ZVE ANALYSIS OP AL'1'ElUIA'1'IVBS: DB BIBB CRI'1'ERIA
The following nine crite~ia, outlined in the NCP at Section
300.430(e)(9)(iii), were used to compare the alternatives and to
determine the most appropriate alternative for remediation of the
soils and groundwater contamination that is protective of human
health and the environment, attains applicable or relevant and
-------
28
appropriate requirements (ARARs), is cost-effective and represents
the best balance among the evaluating criteria. An alternative
providing the "best balance" of tradeoffs with respect to the nine
criteria is determined from this evaluation.
A.
Threshold Criteria
1.
OVerall Protection of Human Health and the Environment
This criterion addresses whether a remedy provides
adequate protection and describes how risks posed through each
pathway are eliminated, reduced or controlled u~rouS~
treatment, engineering controls, or instit~tional controls.
All alternatives except GWl and Sl would minimize the
risks posed by direct contact or ingestion of site-related
contaminants by treatment and/or containment. A1l remedial
technologies except GWl and Sl are considered protective of
human health and the environment.
Comparing groundwater Alternatives GW2 through GW4,
Alternative GW2 is considered least effective in remediating
the contaminated aquifers, because contaminants are not
actively contained or extracted. Alternative GW2 would be
most reliable in protecting the water quality of the existing
municipal system. However, Alternative GW2 would not be
protective of private well users who are or may in the future
be impacted by the groundwater contamination. To increase the
reliability of Alternatives GW3 and GW4 in protecting the
municipal well system, contingency plans can be prepared to
allow for swift and effective well-head remediation if
municipal wells show signs of contamination. Upon ultimate
cleanup of the contaminated groundwater, Alternative GW4 (and
possibly Alternative GW3) would protect the health of private
well users.
Alternative S3 is considered more reliable than
Alternative S2 because, once completed, the site would require
no engineering or institutional controls.
2.
ComDliance with ADDlicable
Reauirements (ARARs)
or
Relevant
and
ADDroDriate
Compliance with ARARs addresses how the proposed
alternative complies with all applicable or relevant and
appropriate requirements of Federal and more stringent state
environmental laws (ARARs), and also considers how the
alternatives comply with advisories, criteria or other
guidance to be considered (TBCs) that do not have the status
of laws, but. that the u.s. EPA and the state have agreed are
"appropriate" for protectiveness or to carry out certain
-------
29
actions or requirements.
A summary of identified ARARs for the groundwater and
soils alternatives is included in section 8.A.2 below. Only
ARARs necessary for on-site remedial activities bave been
identified. In some instances, rules cited contain both
substantive and procedural or administrative requirements.
Only the substantive requirements are ARARs for the purpose of
on-site activities. Examples of administrative or procedural
requirements which are not consid~red ARARs include, but are
not limited to, reporting requirements and permit application
r~~w~remen~z. A ~=re detailgd discussion of: site ARARa is
l.;.-:at~d in AppendL~ A of :.:.:.e :'5. - - - . -
a.
Clo8ur.
1.
Federal
Portions of 40 CPR Part 264, regarding standards for
owners and operators of bazardous waste treatment, storage,
and disposal facilities, are ARARs. 40 CPR 264.94 and 264.100
requirements are relevant and appropriate since they regulate
circumstances sUfficiently similar to those at the site
because source area soils are likely to be cbaracterist-ically
hazardous. 40 CPR 264.301, 264.303-304, 264.310, 264.91-100,
264.111, and 264.116-117 requireaents are applicable since
RCRA bazardous wastes (i.e., cbeaical precipitation 8ludge)
aay be placed in a landfill, after treataent to satisfy LDRs.
i1.
state
Certain Hazardous Waste Manageaent Act 64 requireaents
are ARARs. MAC Rule 299.9612 requirements are not applicable
because source area soils are not contaminated with Hazardous
Waste Management Act (BWMA) waste. However, the requirements
are relevant and appropriate since they regulate circuastances
sufficiently similar to those at the Site. Levels of
contamination in the source area soils indicate that these
soils would be cbaracteristic bazardous wastes. MAC - Rule
299.9602-9604, 299.9611-9613, and 299.9619.9622 are applicable
if HWMA wastes are placed in a landfill.
Micbigan Act 641, P.A. 1978 (Solid Waste Management Act).
This Rule prescribes the requirements for new and existing
sanitary landfill facilities. The rules cover construction
permitting procedures, landfill design, groundwater quality,
permeability testing, licensing procedures and operating
requirements. These requirements are applicable for
construction of a solid waste cap, as described in Alternative
S2 - Cap A.
-------
30
iiL
Discussion
The caps considered in Alternative S2 can comply with state
and Federal ARARs. Residuals from groundwater and soil
treatment alternatives can be handled in such a way that
Federal Land Disposal Restrictions will be met (for a further
discussion, see subsection (e) on page 33 of this ROD).
b.
Groundwater standards
i.
Federal ARARs
Maximu.m Contaminant I.,;,avels (MCLs), and to a certain
extent, Maximum contaminant Level Goals (MCLGs), the Federal
drinking water standards promulgated under the Safe Drinking
Water Act (SDWA), 40 CFR 141, are applicable to municipal
water supplies servicing 25 or more people. MCLs and MCLGs
are relevant and appropriate to circumstances at the Site,
since the aquifers are used as sources of groundwater for
drinking water. MCLGs are relevant and appropriate when the
standard is set at a level greater than zero (for non-
carcinogens); otherwise, MCLs are relevant and appropriate at
superfund sites. The point of compliance for MCLs and MCLGs
is throughout the plume if wastes are removed from the Site.
Federal Water Pollution Control Act (Clean Water Act)
Section 304 provides for the development and publication of
water quality criteria for human health and aquatic life.
Water quality criteria are not legally enforceable standards
and are therefore not applicable. However, a modified water
quality criterion for TCE for drinking water consumption would
be relevant and appropriate for all groundwater alternatives.
iL
State ARARs
The Michigan Department of Natural Resources has
promulgated administrative rules governing "Environmental
Response Acti vi ty" pursuant to the Michigan Environmental
Response Act, 1982 PA 307, M.C.L. 299.605 ("Act 307 Rules").
These Act 307 Rules, codified at M.A.C. Rule 299.5101 et sea.,
contain provisions establishing procedures for response
activities (M.A. C. Rules 299.5501-5519), selection of remedial
action (M.A. C. Rules 299.5601-5607), and cleanup criteria
(M.A.C. Rules 299.5701-5727) at sites of environmental
contamination where response activities are taken pursuant to
Act 307 and ARARs for this remedial action. These Rules
contain requirements that are applicable to remedial action at
the sturgis site since: (1) contaminant levels currently
found in the Site groundwater exceed levels set by application
of the cleanup criteria; and (2) contaminants currently found
in the site I s soils are a potential continuing source of
contamination to the groundwater.
-------
31
The selected remedy will attain requirements contained in
the Michigan Act 307 Rules. Part 7 of the Rules provides, in
general, for three different types of cleanup criteria,
designated as Types A, Band C. Type A criteria are based on
reduction of hazardous substances concentration to background
or detectable levels. Type B criteria are based on reduction
of hazardous substapce concentrations to an acceptable risk
level (e.g., 1 x 10-0[0r carcinogens), using standard exposure
assumptions. Type C criteria involve a site-specific
assessment of risk. """
""
SDWA - Act 399, P.A. 1975'(Michigan Safe Drinking Water
Act) provides" regulations"establisbinq" MCLs-:;for certain:
contaminants in addition to Federal MCX. and would be relevant
and appropriate for all groundwater alternatives. MAC Rule
325.10601-10607 requirements are applicable since the aquifer
underlying the Site is used to supply a co_unity water
system.
MWRCA Act 245, P.A. 1929 (Michigan Water Resources
Commission Act) contains State water quality standards,
treatment plant operator requirements, and wastewater
reporting requirements. The rules also impl_ent a waste
effluent discharge system compatible with NPDES requirements
and provide for the non-degradation of groundwater. Because
NPDES requirements regulate discharge, these water quality
standards are applicable to Alternatives GW3and GW4 for
discharge to surface water. The state has identified Act 245
Part 22 Rules as an ARAR for the sturgis Municipal Well Field
site. The united states disagrees that Act 245 Part 22 Rules,
as interpreted and applied by the State, is an ARAR. This
issue is the subject of litigation in u.S. v. Akzo coatinas of
America, in appellate case numbers 89-2902 and 89-2137.
iii. Discussion
The No Action alternative does not comply with the
requirements of the Act 307 Rules, Michigan Water Resources
Commission Act, or the Federal and state Safe Drinking Water
Acts as they pertain to qroundwater. Adoption of this"
alternative would not prevent further migration of
contaminated groundwater.
Both the Federal and State Safe Drinking Water Acts are
relevant and appropriate (the aquifer under the site is used
for a community water supply, but applicability is determined
at the tap) to sturgis groundwater considerations since they
regulate Maximum contaminant Levels in drinking water for
protection of human health. Alternative GWl would not meet
ARARs. Alternatives GW2 would meet Federal and state ARARs,
but would only attain a Michigan Act 307 Type C cleanup.
-------
32
Alternative GW3 would meet Federal and state ARARs, but would
not actively remediate to Type B levels. It is, however,
possible that Alternative GW3 would meet Type B levels in the
process of containing the plume and treating extracted
groundwater. Alternative GW4 would meet Federal and state
groundwater ARARs specific to individual component actions
(e. g., air stripping, carbon adsorption, chemical oxidation) .
c.
soil st.andard.
L
state ARARs
HERA Act 307, P.A. 1982 (~ichiqan Environmental
Response Act) provides rules regarding the procedures for
determining cleanup criteria for contaminants in groundwater,
surface waters, soils, and air. Act 307 is applicable for all.
soil alternatives.
i1.
Discussion
Alternatives S2 and S3 would meet Federal and state
ARARs, while the No Action Alternative does not comply with
the identified ARARs for the soils areas. Both the multi-
media and Michigan Act 64 clay caps comply with the
requirements found in RCRA at 40 CPR 264 et sea. However, the
containment proposed under Alternative S2 would not meet
Michigan Act 307 Type B requirements. Please refer to the
Feasibility study for further discussions of the soils
alternatives and ARARs.
d.
Air st.andard.
1.
Federal ARARs
Regarding the Clean Air Act Requirements, 40 CPR 50.1-
50.12 requirements are applicable because emissions from the
groundwater and soil treatment systems would be subject to
primary and s~condary Ambient Air Quality standards.
construction and treatment system activities are potential
sources of fugitive dust, particulates, and VOCs and would be
subject to the TSP standard.
i1.
state ARARs
Certain state Air Pollution Act requirements are ARARs.
Act 348 contains rules regarding emission limitations and
prohibitions for particulate matter, fugitive dust, and VOCs.
MAC Rule 336.1702, 336.1901, and 336.1371-1373 requirements
are applicable since emissions from the treatment system would
be subj ect to state and standards for VOCs. construction
activities are potential sources of fugitive dust.
-------
33
iii. Discussion
It is expected that all action alternatives for soils and
qroundwater could meet state and Federal air standards.
..
acaa Land Di.po.al a..triotion. (LOa.)
1.
Federal ARARs
RCRA LDRs, 40 CFR Part 268, are applicable for all actlon
alternatives considered for the sturqis ~unicipal Well Fiel~
S1 ':.e . Filter media for :,oth groundwater and-soil t::-'2'--: "'-t
alternatives are likely- to become characteris't..- - .:
hazardous, and therefore, LDRs are applicabl-. prior to re-uo8
or disposal. If characteristic spent carbon i. to be land
disposed, it aust be placed into a Subtitle C landfill. When
treatment standards for newly identified Toxicity
Characteristic (TC) wastes are pro.ulqated, the characteristic
carbon will have to be treated to below the treatment
standard, but then may be disposed in a Subtitle D landfill.
i1.
Discussion
Althouqh it is expected that all action alternatives
considered for both soil and qroundwater could meet RCRA LDR
restrictions, a Treatability Variance for contaminated carbon
residuals would be established to address the possibility that
regeneration/theraal destruction cannot reduce TCB and PCB in
the spent carbon to LDR acceptable levels. The treataent
level ranqe established throuqh a Treatability Variance that
regeneration/thermal destruction will attain for each
constituent of concern are:
TCE
PCE
95-99.9' reduction (TWA)
95-99.9' reduction (TWA).
f.
~raD.portation Requlation.
Federal ARARs
i.
certain Department of Transportation requirements are
ARARs at this Site. If contaminated soil, contaminated
qroundwater or treatment byproducts (i. e. , chemical
precipitation, spent filters) are determined to be hazardous
wastes under RCRA and are transported off-site, the Department
of Transportation Rules for transportation of hazardous
materials (40 CFR Part 263) and- RCRA would be applicable to
any off-site transportation or handlinq of the hazardous
wastes. In addition, 49 CFR 107 and 171 requirements are
applicable when hazardous wastes are transported to an off-
site disposal facility.
-------
34
i1.
state ARARs
MAC Rule 299.9301-299.9311 generator requirements are
applicable for all wastes transported off-site.
iii. Discussion
It is expected that all action alternatives could meet
the transportation ARARs identified above.
q.
:o81laneoU8 &DR.
i.
Federal ARARs
Since the levels of TeE and PCB in the groundwater are
high enough to make the groundwater a RCRA hazardous waste by
characteristic, any unit in which the groundwater is treated
is subject to RCRA regulations for tanks, 40 CPR Part 264
Subpart J and/or RCRA regulations for miscellaneous units, 40
CPR Part 264 Subpart X.
If characteristic carbon is to be regenerated, RCRA
regulations for miscellaneous units are applicable, as set
forth in 40 CPR Part 264 Subpart X.
i1.
state ARARs
Excavation and construction activities must comply with
substantive requirements promulgated pursuant to Michigan's
Soil Erosion and Sedimentation control Act (MCL 282.106). MAC
Rule 323.1701-1714 requirements are applicable since
construction would involve earth changes and the potential for
. soil erosion.
certain Mineral Well Act requirements are ARARs. MAC
Rule 299.2211-2229 requirements are applicable if extraction
and monitoring wells are installed on-site.
iii. Discussion
It is expected that all
alternatives could comply with
identified above.
soils and groundwater
the miscellaneous ARARs
B.
PrimarY Balancina Criteria
3.
Lana-Term Effectiveness/Permanence
This criterion delineates the residual risk and evaluates
the ability of an alternative to maintain reliable protection
of human health and the environment over time, once cleanup
-------
35
objectives have been met.
All alternatives other than GW1 and Sl would offer
protection of public health and the environaent over the lonq
term by treating or containing contaainants. However,
treatment-oriented alternatives are .ore effective over the
long-term than containment alternatives. If groundwater
contamination were to be treated only as needed (Alternative
GW2) or contained to prevent further. contamination of the
municipal wa~er system (Alternative GW3), the public water
supply and private water supplies would. require ,permanent
oversight to:issure that. -=on~aminant .l.evels in the water
supply were safe. In addition, it i. not possible to insure
legally enforceable institutional controls on qroundwater
usage fro. private wells. Containaent of contaJlinants at the
source areas (Alternative S2) would require perpetual
maintenance of the cap to assure its effectiveness.
4.
Reduction of Toxicity. Mobilitv or Volume Throuah Treatment
This criterion evaluates the anticipated performance of
the treatment technologies a remedy aay employ.
Alternatives GWl and Sl would not reduce toxicity,
mobility or volume (THY) of contaainants at the Site.
Alternatives GW2 through GW4 would reduce the THY by removing
contaminants fro. the qroundwater. Although Air stripping and
Liquid Phase. carbon Adsorption only transfer contaainant. from
qroundwater to another media (and thus do not reduce THY), the
reqeneration of the filter carbon used in the processes would
reduce the THY through treatment. Chemical oxidation destroys
contaminants and thus reduces THY through treatment.
Alternatives S2 and S3 both reduce THY, through the
reqeneration of the filter media used for SVE.
5.
Short-term Effectiveness
Short-term effectiveness addresses the period of time
needed to achieve protection and any adverse impacts on human
health and the environment that may be posed during the
construction and implementation period.
Alternatives GWl and GW2 would not actively reduce levels
of contamination in the groundwater. Alternatives GW3 and GW4
would actively reduce levels of contamination. Alternative
GW4 would be more efficient than Alternative GW3 in removing
contaminants from the groundwater, and would therefore reach
cleanup levels before Alternative GW3. Until groundwater
cleanup levels are achieved, only institutional restrictions
on the placement of private wells can be used to discourage
the use of private wells. The potential also exists for
-------
36
worker exposure to contaminated material during treatment of
groundwater for Alternatives GW2, GW3 and GW4.
Alternative Sl would not reduce levels of contamination
in the soils at either source area. Alternatives 82 artd 83
can be constructed and reach cleanup levels in approximately
3 years. Until cleanup is realized, fencing can be used to
restrict access to contaminated areas. There is also
potential for worker exposure to contaainated material during
disturbance of soils. The potential for exposure increases as
the amount of water being treated or soil being disturbed
increases. Alternative 53 possibly poses ~e grea~est threat
of exposure to workers because of the larqe quantities of soil
to be excavated for off-site disposal or on-site incineration.
A health and safety program which include. worker protection,
dust suppression, and air monitoring should mini.ize these
risks. \
6.
ImDlementability
This criterion considers the technical and administrative
feasibility of implementing an alternative, including the
availability of material and services needed to implement a
particular option.
All of the treatment options and technologies considered
for groundwater remediation would be relatively siaple to
implement. However, Alternatives GW3 and GW4 would include
construction of new wells for extracting groundwater, which
would have to be carefully planned so as not to spread
existing contamination. Both discharge options for treated
groundwater in Alternatives GW3 and GW4 are implementable,
although re-injection may be administratively difficult to
implement because of legal questions concerning the
applicability of Michigan Water Resources Commission Act as an
ARAR. '
Alternatives Sl, S2 and S3 are all easily implementable
at the site. The use of incineration with Alternative S3
could be difficult to implement because the substantive permit
requirements would need to be met. ,
7.
~
The net present worth (NPW) costs of all alternatives
considered are presented on the following tables. See the'
detailed description of each alternative for a breakdown of
capi tal costs, O&M costs and assumptions used for cost
estimation.
-------
37
8mamay 01' IIP8 COST8 :rea GIIOUlJDDUR U,ftJtPTrflI8
GW1 $ 2,085,000
Air 8trippinCJ Liqui4 Pbaa. Cheaioal
Carbon ozi4ation
A48orption
I GW2 $ 7,450,000 $ 8,940,000 - $ 16,000,000
j
.
Stonl Reinject Stonl ..lnJect $tonl- ..a.inJ8Ct
S8wr tan s-r tan ..... tan
GW3 . 9.56 II "0.90 II "0.50 II "2.00. II 121.50 II 122.90 II
GW4 111.20 II 114.40 II 114.40 II 117.60 It 121.50 It S24.6O It
81JJ11mJtY 01' IIP8 C08'1'8 :rea 8O:IL U,HJtD'frfllS
Sl $ 104,000
8V1I with Cap & 8V8 with Cap B
S2
$ 1.61 M $ 2.97 M
8V8 with 8V8 with Off-8it.
S3 :Inoin.ration of Di.po.al of
...i4ual PH ...i4ual PH
Contaainat.4 80il Contaainat.4 80il
$ 6.50 M $ 2.34 M
c.
Modifvina criteria
8.
state AcceDtance
MDNR concurs with the selected alternative.
9.
communitv AcceDtance
community acceptance is assessed in the attached
Responsiveness Summary. The Responsiveness Summary provides
a thorough review of the public comments received on the
Proposed Plan, and the Agency' s responses to those
comments.
-------
38
9.0
'filii SBLBC'fBD UDDY
Based upon consideration of the requirements of CERCLA, as
amended by SARA, and the NCP, the detailed analysis of alternatives
and public comments, u.S. EPA and MDNR have selected Alternatives
GW4 and S3 for remediation of contamination at the Sturgis
Municipal Well Field Site. Alternative GW4 (employing Air
stripping with treated groundwater to be discharged to a Storm
Sewer) was selected for remediation of groundwater contamination.
The Aaencies have also determined that SVE, with excavation and
Off-site disposal of residual PAR contaminated soils, shall be
implemented under Alternative S3 to address the source area aoi1
c:::1tamination at both the Wade and Kirsch prop.a~ies. Both
groundwater and soil alternatives will use Michigan Ac~ 307 Type B
levels for establishing cleanup levels.
The detailed evaluation of groundwater alternatives found that:
.
Alternative GW1 is not protective of human health and the
environment:
.
Alternative GW2 would not restore the city's aquifers to
beneficial use and would not be protective of private well
users;
.
Alternative GW3 would not actively restore the city's aquifers
and would not achieve ARARs as quickly as Alternative GW4; and
Re-injection of treated groundwater in Alternatives GW3 and
GW4 does not increase effectiveness proportional to its high
cost.
.
The detailed evaluation of soil alternatives found that:
Alternative Sl is not protective of human health and the
environment:
.
.
Alternative 52 "is protective of human health and the
environment but would require long-term maintenance of a
multi-layer cap and permanent restrictions on the use of the
contaminated land; and
.
On-site incineration of residual PAR contaminated soils
removed under Alternative S3 would be significantly more
costly than excavation with disposal.
Based on the information available at this time, the U.S. EPA
and the State of Michigan believe that the selected alternatives
(Alternatives GW4 and S3) would reduce the risks to human health
and the environment by removing and treating solvent contamination
-------
39
in the groundwater and the source area soils. It is iaportant to
note that Alternative GW4 will eventually allow for unrestricted
use of the city's aquifers. Alternative S3, with excavation, will
remove contamination in the soils to acceptable heal th-based
levels, thus restoring the contaJainated source areas to
unrestricted use. Alternatives GW4 and S3 would also be cost-
effective, attain ARARa, and use permanent solutions to the maximWl
extent practicable.
It ~s necessary ":0 clearly?~tate that the goal of Alternative
GW4 is ~o restore groundwater to its beneficial use, which is, at
-:::lis Si-c.e, a drinking water source. 3ased'-cm information obtained
during the RI and on careful analysis of all remedial alternatives,
u.s. EPA believes that the selected remedy will achieve this goal.
It may become apparent, during 1apl...ntation or operation'of the
groundwater extraction and treataent syst- and its modifications,
that contaminant levels have ceased to decline and are r-.ining
constant at levels higher than the cleanup levels over some portion
of the contaminated plume. In such a case, the systea performance
standards and/or the remedy may be reevaluated.
Alternative GW4 will include groundwater extraction for an
estimated period of 20 years, during which the system' s perfo~ance
will be carefully monitored on a regular basis and adjusted as
warranted by the performance data collected during operation.
Modifications aay include any or all of the following:
.
Discontinuing pumping at individual wells where cleanup levels
have been attained;
~ternating pumping at wells to eliminate stagnation points;
.
.
Pulse pumping to allow aquifer equilibrium and to allow
adsorbed contaminants to partition into groundwater; and
Installing additional extraction wells to
accelerate cleanup of the contaminant plume.
It also is necessary to note that the selection of Alternative
GW4 by this ROD does not restrict the desiqn of the final
groundwater treatment system to the design developed in the FS.
The 90al of Alternative GW4 is the restoration of the well field.
The design of such a sy~tem may, at the discretion of u.s. EPA,
include: (1) surface water discharge of treated water (via the
storm sewer) or municipal use of treated water, and/or (2)
centralized treatment or decentralized treatment of groundwater.
.
facilitate or
-------
40
10.0
STA'fU'IORY DB'nJUlIIlATIOBS
The selected remedy must satisfy the requirements of Section
121 (a-e) of CERCLA, as amended by SARA, to:
a.
Protect human health and the environment;
Comply with ARARs (or justify a waiver);
Be cost effective;
Utilize permanent solutions and alternative treatment or
resource recovery technoloqies to the maximum extent
practicable; and,
Satisfy the preference for treatment as a principal
element or provid~ ~~ explanation as to why this
preference is not sQ~isfied.
b.
c.
d.
e.
The implementation of Alternatives GW4 and S3 satisfies the
requirements of CERCLA, as amended by SARA, as detailed below:
a.
Protection of Human Health and the Environment
This selected remedy will provide protection of human
health and the environment through treatment.
Implementation of the selected alternative will-reduce
potential risks to human health and the environment posed by
possible use of contaminated qroundwater and will eliminate
the Kirsch and Wade properties as sources of groundwater
contamination.
No unacceptable short-term risks or cross-media impacts
would be caused by the implementation of the remedy. Durinq
the period required for remediation, institutional controls
would be used to mitigate the interim threats from possible
use of contaminated groundwater and possible exposure to
contaminated soils. The community and Site workers may be
exposed to TCE and PeE in the soils and air, and to dust and
noise nuisances during implementation of the groundwater and
soil remedies. Standard safety proqrams, such as fencing, use
of protective equipment, moni.toring and dust control measures,
should mitigate any short-term risks.
b.
ComDliance with ARARs
The remedy selected will meet or attain the applicable or
relevant and appropriate Federal and State requirements, and
will be implemented in a manner consistent with these laws.
In particular, the remedial action selected for
implementation at the Sturgis Site is consistent with the
National Contingency Plan, the State's Act 307 Rules, and with
MAC Rule 323.1041 - R323.1116.
-------
41
Under Rule 299.5705(4), "[t]he reaedial action type
(i.e., A, B, C or a combination thereof) proposed shall be at
the option of the person proposing the reaedial action.. u. S.
EPA has selected extraction and treataent of the hazardous
substances in the groundwater and soils to Type B levels
determined through Rules 299.5709.
The cleanup llvels for. Act 307 Type 8 carcinogens are
based on a 1 x 10- cancer risk and are consistent with Rules
299.5709 and 299.5723. Where the cleanup level for a
particular chemical is based on analytical detection (MDL) ,
Rule 299.5707 (b) has been applied. Rules 299.5709 (b) and
299.5125 provide for cleanup .levels Dased on Human Lifacycle
Safe Concentrations (HLSC); taste and odor (T'O) thresholds
are consistent with Rule 299.5709(d); and Rule 299.5725(b) (i)
is the citation for Maximua Contaainant Levels (MCLa), where
used.
Al though it is anticipated that spent carbon from
groundwater and soil treatment units can be reqenerated to LDR
acceptable levels, a Treatability Variance is granted to
address the possibility that regeneration/thermal destruction
cannot reduce TCE and PCE in the spent carbon to LDR
acceptable levels.
Chemical Soecific
The treatment systea will co.ply with Michigan's Safe
Drinking Water Act (Act 399) requireaents to provide water of
an acceptable quality. Monitoring, with the provisions of
installing necessary 'treatment, provides water supply
protection. If treatment is needed, the water will be treated
to levels at or below the Act's Maximum Contaminant Level
(MCL) for TCE and the proposed MCL for PCE.
The selected al ternati ve for groundwater, Al ternati ve
GW4a - Aquifer Restoration without Re-injection, can co.ply
with the Act 307 rules for Type B clean-up levels. Type B
cleanup levels for selected chemicals of concern in
groundwater are identified in the following table.
-------
42
Michigan Act 307
Selected VOC Cleanup Levels for Groundwater
(in ug/L)
Trichloroethene (TCE) 3
Tetrachloroethene (PCE) 1
1,2 Dichloroethane 1
--
Bromodichloromethane I 1
1,1,2 Trichloroe~hane 1
Vinyl Chloride 1
Benzene 1
Cyanide 100
The selected alternative for soils, Alternative S3, SVE
of VOCs with excavation and solid waste disposal of residual
PAH contamination, can comply with the Act 307 rules for Type
B clean-up levels. Type B cleanup levels for SVE removal of
VOCs in soils, based on twenty times the qroundwater standard,
are identified in the table below.
VOC Cleanup Levels for Soil
(in ppb)
Trichloroethene (TCE) 60
Tetrachloroethene (PCE) 14
Leach tests may also be performed to demonstrate that
contaminants are not leaching into groundwater at levels above
the groundwater cleanup standards.
After the SVE removal of VOCs is complete, soils
containing residual PAH contamination will be excavated until
the following indicator PAHs in the remaining soils are at or
below their respective Type B cleanup levels, which are equal
to their Method Detection Levels.
-------
43
Excavation Cleanup Levels for PABs
(in ppb)
Benzo(a)pyrene 330
Benzo(b)flouranthene 330
The levels of contamination reaching the plume management
wells could exceed the TCLP limits. The TCLP, replaces thg
Extraction Procedure (EP)ln determinin,g.if CERCLA-wastes ar,g
characteristical:ty.hazardous wastes. .-' When th..-_-TCLP replaces
the EP in Michigan I s Hazardous Waste Manageaent Act (Act 64),
hazardous substances exceeding the TCLP liait. would also be
a hazardous waste under state regulations. To caaply with
these requirements, the water would be 1I8.naged as a hazardous
waste.
Tbe spent carbon from the air or water adsorption systems
could potentially be a characteristically hazardous waste.
The spent carbon would be tested to determine if it would be
a hazardous waste. Following the generator and transportation
requirements of Michigan I s Hazardous Waste Managaaent Act (Act
64) would allow this alternative to comply with the Act.
The soil and groundwater treatment systeas will caaply
with Michigan's Air Pollution Act (Act 348) requiraaents.
Best Available CUrrent Technology (BACT) requires treataent of
the air eaissions from the air stripper systea. Tbese
requiraaents are satisfied through the use of a vapor phase
activated carbon system to remove VOCs from the stripper air.
Location-Soecific
Because the Wade Electric property is adjacent to an old
cemetery, Alternative S3 will co.ply with the following ARARs:
Protection of National Historic Landmarks (36 CPR Part 65),
Protection of Historic and CUltural Properties (36 CPR Part
800), and Protection of Archeological Resources (36 CFR Part
296, 43 CrR Part 7, 32 Part 229, 18 CrR Part 1312).
Action-Soecific
SESCA (Act 347) regulates erosion and sediment control
measures during construction. The selected alternative will
comply with these requirements during construction of the
treatment systems.
MCLA (Act 257) requires a lower maximum axle load on some
highways during March, April, and May to prevent roadway
damage. Depending on the weight of the spent carbon units,
-------
44
special routes or schedul ing around these months may be needed
to allow unit replacement.
c.
Cost Effectiveness
A cost-effective remedy is one for which the cost is
proportional to the remedy's overall effectiveness.' The
detailed costs associated with the implementation of
groundwater and soil alternatives can be found in Section 7,
with costs summarized in Section 8.
Alternative 53 affords a high degree of effectiveness by
treating the principal threat present in source area soils,
with excavation and off-site containment of low-level residual
PAR soil contamination. Alternative GW4 also affords a high
degree of effectiveness by treating the contaminated
groundwater and returning it to its beneficial use.
Alternative GW4 is protective of human health and the
environment and will achieve cleanup levels more rapidly than
other groundwater alternatives. The goal of Alternative GW2
is to protect the municipal water supply. The goal of
Al ternati ve GW3 is not to remediate the aquifer, but to
protect the municipal wells from contaminant migration.
Although Alternatives GW2 and GW3 would remove contamination
from the aquifer, only Alternative GW4 would actively
remediate the resource to beneficial use.
Alternative S3, with SVE of VOCs and off-site disposal of
residual PAR contaminated soils, is protective of human health
and the environment. Clean closure of the source areas to
reduce levels of contamination to acceptable levels
'(Alternative S3) does not cost significantly more than
Alternative S2, where low-level contamination would be
contained on-site. Alternative S3 will allow for unrestricted
land use while alternative S2 would require perpetual
operation and maintenance of the cap and future use
restrictions .'
d.
utilization of Permanent Solutions and Alternative Treatment
Technoloaies or Resource Recoverv Technoloaies to the
Maximum Extent Practicable
u.S. EPA, with the State of Michigan's concurrence, has
determined that the selected remedy meets the statutory
requirement to utilize permanent solutions and alternative
treatment technologies to the maximum extent practicable for the
Sturgis Municipal Well Field Site. Of the alternatives that are
protective of human health and the environment and comply with
ARARs, u.S. EPA has determined that the selected remedies for
-------
45
addressing soil and groundwater contamination provide the best
balance of tradeoffs in terms of long-tera effectiveness and
permanence, reduction of THY through treatment, short-term
effectiveness, implementability, cost and state and co..unity
acceptance.
The selected remedy represents the aaxiaua extent to which
permanent solutions and treatment can be practicably utilized
for this action. The principal threat present in soils will be
treated through SVE of VOCs. The low-level, long-tera threat
posed by PAR contamination in the source are soils is addressed
::r..:::-ugh excavation and ~ff-site solid-waste disposal of
contaminated soils. Grou."1dwater ccn~al:li.na~ion, .which represents
the principal source of site risk, will be actively r&88diated
to restore the resource to its beneticial us.. Institutional
controls will be used during reaediation to mitigate the risks
posed by use of contaminated groundwater and exposure to
contaminated soils.
e.
Preference for Treatment as a Princinal Element
The selected remedy uses treataent as a principal eleaent.
Alternative GW4 will utilize Air Stripping to reaove
contaminants from the groundwater. Although Air Stripping and
Liquid Phase Carbon Adsorption only transfer contaminants trom
groundwater to their filter media (and thus do not reduce TMV),
the reqeneration of the filter carbon used in the processes
would reduce the TMV through treatment.
A1ternative S3 utilizes SVE to remove volatile
contamination from the source area soils. Reqeneration of the
filter media used for the SVE will reduce TMV through treatment.
11.0
DOCUIIIDI'IA'1'IOli 0.. 8ICDII..IC&I1'1' CBaIIG88
u.s. EPA reviewed all written and verbal comments submitted
during the public comment period. Upon review of these comments,
it was determined that no significant changes to the remedy, as it
was originally identified in the Proposed Plan, were necessary.
-------
ItBSPOBSIVBDSS SI1IIInIlY
This Responsiveness Summary has been prepared to meet the
requirements of sections 113(k) (2) (B) (iv) and 117(b) of the
Comprehensive Environmental Response, compensation, and Liability
Act of 1980 (CERCLA), as amended by the Superfund Amendments and
Reauthorization Act of 1986, which requires the United States
Environmental Protection Agency (U.S. EPA) to respond "...to each
of the significant comments, criticisms, and new data submitted in
written or oral presentations" on a proposed plan for remedial
ac-:.ion. The Responsiveness Summary addresses concerns expressed by
t.:-~e public and potentially responsible parties (PRPs) in the
wri~ten and oral comments received by the U.S. ~PA and ~~e State
regarding the proposed remedy for the Sturgis Municipal Well Field
site.
A.
I.
OVBRnBW
BACJ:GROmJD/PROPOSBD PLAB
In 1982, during routine sampling and testing of the water supply
wells in the city of Sturgis, the Michigan Department of Public
Health discovered that the water from two city wells contained low
levels of volatile organic compounds (VOCs): trichloroethene (TCE)
and tetrachloroethene (PCB). Subsequent investigations during the
Remedial Investigation (RI) determined the nature and extent of
contamination and identified two source areas.
The major components of the selected remedy include:
Soil vapor extraction of Volatile Organic Compounds (the
principal threat) in the source area soils to Michigan
Act 307 Type B Cleanup Levels: .
Excavation of soils contaminated with low-level
Polynuclear Aromatic Hydrocarbons (PAHs) to Michigan Act
307 Type B Cleanup Levels for indicator contaminants and
disposal of soils in a Michigan Type 2 Solid Waste
Landfill:
Extraction and treatment of groundwater to Michigan Act
307 Type B Levels using air stripping, with vapor phase
granular activated carbon to be used to treat the off-
gasses:
Discharge of treated water to surface water or to the
municipal system:
1
-------
B.
C.
Thirty year groundwater monitoring program to assure the
effectiveness of remedial action and quality of municipal
water supply.
II.
PUBLIC COvvw.. PBRIOD
A public comment period was held fro. June 10, 1991 to
August 10, 1991 to allow interested parties to co_ent on
the Proposed Plan in accordance with section 117 of
CERCLA. On June 20, 1991, a public .eeting was held in
sturgis, Michigan at the sturgis City Library. u.s. BPA
and Michigan Department of Natural Resources (MDNR)
. .-. J
presented the Proposed Plan, answered quest1ons, and-
accepted comments from the public. During the 60-day
public comment period, u.s. EPA received approximately
five sets of written co..ents and no verbal-co...nts
concerning the proposed plan. one set of co_ents, fro.
Cooper Industries, was especially extensive.
COIlll1JllIn IIIVOLVBJIBJI'1'
Public interest regarding the $ite has been .inimal.
Although the Proposed Plan was distributed to .
approximately 300 residents in the vicinity of the source
areas, only a handful of residents attended the public
meeting. One resident who attended the .eeting expressed
his concern that the u.s. BPA and the MDNR were _king
too much of the well field contaaination. However,
several other residents expressed their concern to MONR
and U.S. BPA representatives about the contaaination
spreading to areas where private wells are in use.
Smnn.RY 01' SICDlII'ICAft mllift.... I'JtOK PUBLIC OI'l'ICY.'.A
Comments 1 and 2
Mr. Paul Wartner, state Senator, and Mr. Glenn Oxender,
state Representative, submitted letters to u.s. BPA
during the public comment period. Both letters requested
that the U.S. BPA respond to cooper's concerns about the
Remedial Investigation/Feasibility study ("RI/FS"). The
letters also requested that the U.S. EPA work with the
Kirsch Division of Cooper Industries to develop and
implement an appropriate remedy.
ResDonse 1 and 2
u.s. EPA has responded to Cooper's comments in Section D
of this Responsiveness Summary. u.s. EPA has selected
the remedy in the Record of Decision and will negotiate
2
-------
D.
B.
with Potentially Responsible Parties for its design and
implementation.
Comment 3
Mr. John Brand, the sturgis City Manager, submitted a
letter during public comment period, stating that he
appreciated the efforts of the u.s. EPA and MDHR with
relation to the sturgis Municipal Well Field site. He
also stated that he had reviewed the water treatment plan
submitted by Cooper and felt that it was consistent with
their objective to develop a cost effective treatment
system.
Response 3
U.s. EPA also finds merit in the proposal made by Cooper
Industries. However, the remedy selected must also be .
protective of those individuals who utilize private wells
and are not on the municipal system. Please see section
D for a further discussion of the Cooper proposed
groundwater alternative.
81JJOm11Y 01' 8IGllII'ICAII'l' co.,........s nIOli '1'BB PUBLIC
Comment 1
Mr. Frank Kalasky, Plant Engineer for the sturgis Foundry
Corporation, offered the sturgis Foundry process well for
use as part of the aquifer restoration and offered spent
foundry sand to replace any contaminated soil removed
during remediation.
Res'Donse 1
The U.S. EPA appreciates Mr. Kalasky's offer.
noted.
Comment
81JJOm11Y 01' SIGllII'ICAII'l' COIlllD1TS .-0. COOPB. IBD08ftIBS
(PUB'" CODDY 01' KIRSCH CODDY)
Oraanization of Comment ReSDonse
u.s. EPA carefully reviewed all material submitted by
Cooper Industries and has summarized and responded to the
significant comments. Comments will be responded to in
the following order:
.
section I - Comment submittal letter dated August 12,
1991 from Mr. Maclay R. Hyde.
3
-------
.
section II - cooper Industries, Inc.'s comments in
regard to u.s. EPA/MDNR identification of PRPs
including a list of 89 firms and their location with
respect to the groundwater contamination and a aap of
locations with respect to groundwater contamination.
section III - Cooper Industri.., Inc.'s co..ents on
the clean-up standards under CBRCLA, the National
contingency Plan ("NCP") and other ARARs under federal
or Michigan Law and their application to the selection
of an aquifer restoration_remedy by the u.s. EPA and
MDNR .
.
.
.- . -
. - --
section IV - Comments from Cooper geologist Chris
Smitll regarding mOJl~tor~n9 wE;;;11::ons-cruction
techniques.
section V - A critique of Warzyn's baseline risk
assessment for the sturgis Well Field Superfund Site,
Gradient COrPOration, August 1991.
.
section ~ - correspondence from Cooper geologist
Chris Smith regarding NPL site pumping, treating and
use of treated water in municipal system.
section VII - Hydrogeological review comments of-the
RI/FS by Hydro-Search, Inc.
.
.
section VIII - correspondence/...orandumfroa Mark E.
Siaonett, Esq. and P.E., literature review of
effectiveness of pump and treat remedy to attain
ARARs .
.
Section IX - Alternate groundwater remediation
concepts, Sturgis Well Field, Sturgis, Michigan,
Hydro-Search, Inc., August 1991.
section X - CERCLA S 104 (e) response by Cooper
Industries, Inc. submitted December 13, 1989.
.
.
section XI - Transcript of public meeting at Sturgis
Public Library, June 20, 1991.
.
section XII - Correspondence from Robert W. Teets,
Director of Risk Management and Environmental Affairs,
to Terese A. Van Donsel, Remedial Project Manager,
U.S. EPA, dated August 12, 1991, and including Freedom
of Information Act requests. .
section XIII - Correspondence dated August 12, 1991,
from Mark E. Mulhollam, Attorney for Cooper
Industries, Inc., to Terese A. Van Donsel, Remedial
.
4
-------
project Manager including a summary'of the TCE and PCE
toxicological literature.
.
section XIV - correspondence from Robert W. Teets,
Director of Risk Management and Environmental Affairs,
to Terese A. Van Donsel, Remedial Project Manager,
U.S. EPA dated september 12, 1991.
U.s. EPA overview
While Cooper Industries' contractors have provided
sUbstantial comments on the RIjFS, tha majority of the
c~mments stem from professional differences related to
the meael used to examine the extent of contamination and
evaluate remedial alternatives. U.s. EPA would like to
clearly state that the purpose of the modeling done
during the RIjFS was D2t to develop a final design for
remediation. The modeling was done purely to better
understand the flow. of contamination and to develop a
backdrop against which the alternatives could be
evaluated. Therefore, many of the comments would be more
appropriately addressed during Remedial Design/Remedial
Action ("RD/RA").
In fact, U.S. EPA and MDNR are currently preparing a
Supplemental Modeling Report which outlines areas where
further refinement of the model will be needed for RD/RA.
Many of the concerns raised by cooper will be more
specifically addressed in this report. The report will
bean attachment to the RD/RA Scope of Work and draft
RD/RA consent Decree and will be provided to potentially
Responsible Parties at the time of special Notice.
It is important to note that Cooper Industries' own
proposed groundwater alternative (discussed below in
section IX) utilizes the same groundwater technolOgy
recommended in the Proposed Plan, air stripping with
carbon adsorption. The primary difference is the
location and number of the wells and the use of the
treated water. These differences can easily be
considered consistent with the groundwater alternative
selected in the ROD. The Feasibility Study ("FS")
clearly states that the system could be centralized or
decentralized, and the Proposed Plan states that
municipal use of the treated water is an option. The
U.S. EPA feels that these are issues that are best
decided in RD, when further refinement of the model can
be done. To commit to any design at this point in time
is imprudent: it would restrict the opportunity to
develop an efficient and cost-effective treatment system
5
-------
for the restoration of the well field.
Cooper's criticisms of the technical feasibility of
aquifer restoration groundwater reaedies are important
considerations in the design of a groundwater treataent
system. However, U.S. EPA has detarained that aquifer
restoration is necessary in Sturgis to return the
resource to its beneficial use. The well field is the
source of drinking water for the City of sturgis
municipal system and the source of water for private well
users not served by the municipal system. While the
~~estions and concerns raised by Cooper are important and
will be critical considerations'in=:the.development of the
groundwater remedial design, these questions are not
appropriate at this tima. Theae ~cstions can and should
be addressed once perforaance data froa an operating
systea can be evaluated. A valuable resource 8bould not
be "written off" merely because a re88diation action is
complex and lengthy.
6
-------
8J1c-.rX08 X
Cooper Industries, Inc.'s counsel, Mr. Maclay R. Hyde,
submitted a comment transmittal letter dated August 12, 1991.
ReSDonse
Legal and technical concerns raised in this transmittal
memo are summaries of the concerns expressed in .ore
detail in the actual comment package. These concerns
will therefore be addressed in the appropriate comment
section.
u.s. EPA, however, would like to emphasize two points.
The commentor stated that the U.S. BPA and MDNR's remedy
is "premised upon a policy disfavoring use of groundwater
which is treated to meet all applicable federal and state
water quality standards in theaunicipa1 .yetea.- This
is false. In the Proposed Plan, the u.s. BPA and the
MOHR clearly stated that municipal use of treated water
is a possibility which should be further explored.
Municipal use of the treated water was not exaained in
the FS because possible municipal use of the water is
dependent on the interest of the City and the acceptance
of its use by the people of sturgis. However, U.S. BPA
believes that groundwater is a very valuable resource and
should be preserved rather than wasted, if there is a
means of using it. Thus, City use of the water would be
acceptable under Alternative GW4 and is allowable under
this ROD.
In addition, the commentor states that the Feasibility
Study and the administrative record do not support
selection of any of the active groundwater alternatives.
The commentor also argues that Alternatives GW3 and GW4
are not consistent with the KCP because neither
alternative can be considered reliable or cost effective.
u.s. BPA disagrees. Alternatives GW3 and GW4 are
reliable and cost-effective. u.s. EPA would like to
stress that there is a difference between an alternative
being costly and an alternative being cost-effective.
Active remediation of groundwater is costly, but it is
cost-effective. Returning the well field to its
beneficial use is the preference of the KCP. Restoration
of the well field would provide a permanent solution to
risks that currently exist at the Site from contaminated
groundwater, thus providing protection to private well
users as well as to the municipal supply. In addition,
u.s. EPA believes that the technologies evaluated for
remediation of groundwater contamination are proven and
reliable. Also, see Section VIII for a discussion of the
ability of "pump and treat" to meet cleanup standards.
7
-------
.BC'.rX08 XX
Cooper Industries, Inc. provided comments concerning u.s.
EPA/MDNR identification of PRPs including a list of 89 firms
and their location with respect to the groundwater
contamination and a map of locations with respect to
groundwater contamination.
ReSDonse
Although u.s. EPA appreciates information submitted
concerning possible additional PRPs, these comments are
nc~ =elevant to the remedy selection precess and the
a~?ropriateness of the pro?osed remedy_and therefore will
not be responded to in ~~isQocum6n~.How~ver, thg ,
inforJIation will be carefully reviewed by both u.s. EPA
technical and l89al ataff. xt i. neceaaary to note that
neither u.s. EPA nor MONR have foraally identified PRPs
at this point in time.
u.s. EPA would like to respond specifically to one point
raised by the co_entor. cooper' s comment that 0 . S. EPA
needs to perforJI further investigations to identify
additional PRPs, and therefore the adainistrative record
does not support remedy selection of an aquifer
restoration reaedy is not on point. A public co..ent
period on the adainistrati ve record ia required by the
National contingency Plan (NCP), aa set forth in 40
C.F.R. 300.430 (f)(3). The purpose of the public co..ent
period is to provide the public with the opportunity to
review and co_ent on the documents that fora the basis
for the selection of a particular response at a site.
The adainistrative record is a record of documents
considered or relied on in selection of a remedy. Since
identification of PRPs does not necessarily bear on what
remedies might be appropriate for the site, the documents
relating to investigation of PRPs are generally not
included in the administrative record supporting reaedy
selection.
8
-------
SBCTXOII XXX
Cooper Industries, Inc. submitted comments on the clean-up
standards under CERCLA, the NCP and other ARARs under federal
or Michigan Law and their application to the selection of an
aquifer restoration remedy by the u.s. EPA and MDNR.
ResDonse
section 121(d) of CERCLA, as amended by SARA, requires
attainment of Federal ARARs and of State ARARs in State
environmental or facility siting laws when State
requirements are promulgated, more stringent than Federal
laws, and identified by the State in a timely manner.
In its Comments on page 5 of the submittal, Cooper
indicates that section 121 (d) (2) (A) requires compliance
with ARARs and states that Maximum contaminant Level
Goals (MCLGs) are ARARs for this site. MCLGs are the
Federal drinking water standards promulgated under the
Safe Drinking Water Act (SDWA). Cooper argues that,
under S 121 (d) (2)(A), MCLGs are the ARAR levels which
must "at least" be attained in the onsite area where such
a goal is "relevant and appropriate." The current MCLGs
for TCE and PCE are both zero parts per billion (ppb),
which Cooper states cannot practically exist.
In fact, MCLGS are relevant and appropriate when the
standard is set at a level greater than zero (for non-
carcinogens); otherwise, MCLs are relevant and
appropriate (see 53 Fed.Rea. 51441 and 40 C.F.R.
300.430(e)(2).) The MCL for both TCE and PCE is 5 ppb.
The state of Michigan has promulgated Act 307 and its
implementing rules, which specify that protection of
human health, safety and welfare, and the environment and
natural resources is achieved by a degree of clean-up
which conforms to one or more of three clean-up types.
Michigan's Act 307 sets more stringent clean-up standards
than MCLs. Therefore, Michigan's Act 307 is an ARAR for
the. sturgis Municipal Well Field site remedial action,
and the clean-up levels to be attained at this site shall
be developed using Act 307.
On page 6 of the submittal, Cooper stated that ARAR
compliance may be waived in circumstances that satisfy
the criteria of paragraph (4) of S 121 (d) of CERCLA, 42
U.S.C. S 9621 (d) (4). section 121 (d) (4) (C) sets forth
the criteria for a technical impracticability waiver, and
Cooper states that because cleanup to the MCLG of zero is
not possible, this requirement should be waived. As
stated above, a clean-up to the levels of MCLGs for TCE
and PCE is not required at this site. In addition, the
9
-------
logical requirement to support an ARAR waiver for
technical impractability would require a demonstration of
engineering impracticability. Cooper baa not provided
u.s. EPA with any information indicating that it i.
technically impracticable to achieve a clean-up at tbia
~, and no such infonaation was docuaented during the
RI at this Site. Cooper's sub.ittala concerning the
practicability of aquifer restoration are significant
considerations, yet questions concerning the ability to
reduce the contamination in the well field to health-
based levels are best answered once performance data from
a treatment system in sturgis can be evaluated.
In its Comments on page 9, Cooper states that the RI does
not provide adequate information to characteri~e the
groundwater flow systea or to evaluate reaedial
alternatives with sufficient accuracy, and therefore aay
be considered incomplete. u.s. BPA disagrees with this
assumption. There has been sufficient inforaation
gathered concerning contaminant characterization and
groundwater flow to allow for selection of a groundwater
reaedy.
In its Comments on page 10, Cooper states that the
alternatives developed during the FS were based on a
subjective interpretation of the broad develop.ent and
screening requirements of the NCP, as set forth in 40
C.F.R. 300.430(e)(1)-(7). The U.S. EPA disagrees with
this assertion. In fact, the al ternati ves developecl
during the FS were based not only on the requir_nts of
the NCP, but also on u.s. EPA'. quidance docuaent,
Guidance for Conductina Remedial Investiaations and
Feasibi1itv Studies (october 1988, OSMER Directive
9355.3-01) which further specifies the approacb to be
used in developing and screening alternatives.
In its Comments on page 12, Cooper states that U.S. EPA
performed a nine criteria evaluation of alternatives, but
then skipped the requirements of 40 C.F.R. 300.430(f) (i).
U.S. EPA disagrees. In its nine criteria evaluation, .
U.S. EPA did evaluate the threshold criteria, the primary
balancing criteria, and the modifying criteria (40 C.F.R.
300.430 (e)(9)(1)(A),(B) and (C». .
In its Comments on page 13, Cooper states, "EPA and DNR
have selected a ground water and soil remedy as
documented in the Proposed Plan," issued in June 1991.
In fact, U.S. EPA and Michigan did not select a remedy in
the Proposed Plan: U.S. EPA and the state of Michigan
issued a Proposed Plan which recommended a remedial
alternative for the soil and groundwater. No remedy is
selected until after all public comments are received and
10
-------
considered.
In its comments on page 17, Cooper states that the FS
report did not consider remedial alternatives for
Michigan's Act 307 Type C compliance. This is incorrect.
Alternative S2 could be considered a Type ccleanup
because contamination would be left contained on-site.
As specified in Act 307, the application of Type A, S, or
C cleanups in the state of Michigan is made on a case-by-
case basis, considering site-specific information. The
u.s. EPA agrees with the state of Michigan that Type C
cleanup objectives are not appropriate for this site
~onsidering the current and future use of groundwater
migrating from the source areas. Type C criteria have
baen applied to sites where, due to the nature and extent
of contamination, the availability of ~echnolog~es for
remediation of the contaminants, and the location of the
site and its surroundings, the use of containment or
institutional controls is the most appropriate way to
eliminate all the expOsure pathways considering the
factors in Act 307 and the HCP.
In its comments on pages 19 and 20, cooper criticizes the
FS report, and states that technical literature supports
cooper's assertion that groundwater remediation through
pumping is an engineering infeasibility. u.s. EPA
disagrees. As stated earlier, claims of technical
infeasibility are best evaluated once performance data
from an operating system in sturgis can be evaluated.
On page 21 of its comments, cooper requests a waiver of
the ARAR requirements pursuant to CERCLA 1 121(d) (4) of
the HCP, 40 C.F.R. I 300.430(g) (2) (v) and (f) (1) (ii) (C)
because the cleanup levels established pursuant to
Michigan's Act 307 for TCE and PCE cannot be reached.
cooper has not provided any support for this argument,
and U.S. EPA disagrees with cooper's determination that
any Michigan Act 307 levels cannot be reached.
On pages 21 and 22, Cooper criticizes the use of
"quickness" of restoration as a factor for evaluating
alternatives. U.s. EPA disagrees with this assertion and
suggests that the comment or reference the HCP (I 300.4~0
(e)(9)(iii)(E)(4». One of the items for consideration
under "Short-term Effectiveness" is "time until
protection is achieved".
On page 23, the commentor states that the reliability of
the pumping schemes has not been established. U.S. EPA
would like to state that the groundwater desiqns used to
represent alternatives in the FS are not final desiqns
for remediation. verification of the pumping schemes is
11
-------
not necessary at this point in tiae.
On page 23, Cooper co..ents that the 80deling work
stretches the available data too far to make it reliable
and that there were too many unknown input paraaeters.
u.s. EPA disagrees; the .adeling conducted in the PS is
sufficient for use for evaluation and .election of
remedial alternatives.
On page 24, Cooper states that U.S. EPA cost esti.ates in
the FS are inaccurate. u.s. EPA disagrees; the cost
estimates are sufficiently accurate to evaluate remedial
alternatives. .
On page 24, Cooper concludes that the FS iilld the' ,;'":
administrative record do not support .election of active
groundwater restoration alternative., and that ARAR8
compliance i. technically infeasible. As stated above,
u.s. EPA disagrees. Cooper finally states that the
selection of active groundwater restoration is not
consistent with the NCP. In fact, for the reasons
described in the previous comments on technical issues,
groundwater restoration is consistent with the NCP. The
NCP states that "EPA expects to return usable ground
waters to their beneficial uses wherever practicable,
within a timefraae that is reasonable given the
particular circuastances of the site" (see I 300.430
(a)(l)(iii)(P». U.S. EPA believe. that restoration is
practicable at sturgis and that the ti.efraae for
reaediation is reasonable.
On page 25, Cooper proposes an alternative of contraction
of the pl\DIe, and argues that this alternative is an ARAR
under Act 307, Rule 705(5). U.S~ EPA considered nuaerous
alternatives, and carried many forward in the PS report.
The alternative of contraction of the plume was not
developed in the PS because active restoration is
possible, and is preferred under the NCP (see I 300.430
(a) (1) (iii) (A» .
On page 25, Cooper references the Gradient critique of
the Baseline Risk Assessment and the HydroSearch, Inc.
critique of the RI/FS. u.s. EPA would like to refer the
reader to the Section V and VII of this Responsiveness
Summary. Cooper also states that the distribution and
volume of soils requiring remediation is not known. u.s.
EPA agrees that it is difficult to estimate the volume of
VOC contaminated soils requiring remediation. However,
the volume of PAR contaminated soil bas been estimated in
the FS. Supplemental sampling to further delineate the
extent of VOC and PAR contamination in soils will be
considered during the pre-design phase of RD/RA. The
12
-------
issue does not affect the selection of SVE as the
technology for VOC treatment or the selection of
excavation and off-site disposal for the remediation of
PAH contamination.
On page 26, Cooper states that more soil investigation
work is necessary to outline the extent of contamination
and associated risks, and therefore, until further soil
investigation work is performed, any remedy other than
institutional controls, cannot be considered cost-
effective, and would be inconsistent with the NCP. u.s.
EPA strongly disagrees. The RI/FS conducted at the
sturgis Municipal Well Field was quite extensive and
significant soil sampling at the source areas was
conducted, showing tha~ remediation is necessary and
allowing a valid comparison of alternatives. If, during
RD/RA, it is determined that additional sampling is
necessary to develop an effective design for source area
remediation, additional sampling will be considered. It
is critical to note that all questions need not be
answered in order to make an intelligent decision
concerning appropriate remediation for the site.
13
-------
.BC'f:I08 :IV
Cooper geologist Chris Smith provide co..ents regarding
monitoring well construction techniques. These comaents are
responded to in the order presented in Mr. Saith's letter to
Mr. Maclay Hyde, dated August 5, 1991.
Comment 1
"The water table wells were not a consistent twelve foot
length (ten foot screened interval plus two feet of sand
and gravel). Warzyn's ~ell compl~tion logs indicate that
the sandpacked o~en inter/al varied from 11.1 - 17.9 feet
in length (adaitional improper well construction
techniques can compound to result in a aaximum open
interval of 11.7 to 20.0 feet)."
ReeDonse 1
This comment ignores the position of the water table.
The saturated lengths of the water table are generally
less than 10 ft.
Comment 2
"The deeper confined well open intervals were not a
consistent seven foot length (five foot screened interval
plus two feet of sand indicate that the sandpacked open
interval varied from 5.4 - 20.4 feet in length
(additional improper well construction techniques can
compound to result in a maxiaua open interval of 6.1 to
120.3 feet)."
Response 2
In a high hydraulic conductivity aquifer, the interval
which produces water to the well during a low pumping
rate sampling event is adjacent to the well screen zone.
The primary purpOse of the seal is to prevent connection
between aquifers. This suggestion that the "actual open
interval" is up to 120.3 ft is inaccurate. The well that
the comment implies has a 120.3 ft open interval, W39D,
has a bentonite pellet seal mixed with sand through a
silty clay and a fine sand. Since the well is screened
in a high permeability sand, the lower seal is effective.
14
-------
Comment 3
"An upper bentonite seal above the sandpack was
nonexistent in four wells and too thin in one well.
In water table wells, this bentonite plug prevents the
downward percolation of rainwater or perched water
through the wellbore. In deeper wells, the plug defines
the top of the investigated open interval."
ReSDonse 3
This comment suggests that ~~e bentonite pellet seal
above the sand pack is the only seal to prevent downward
m~gration of rain water or p~rched groundwater. Rather,
a seal with a lower permeability than the soil or rock is
required from the ground surface to the top of the sand
pack. At sturgis, we used bentonite slurry or
bentonite/sand slurry at most locations, so the seal was
assured to have a lower permeability than the sand and
gravel. .
Comment 4
"Collapsed formation was utilized as the sandpack in
three wells.
A collapsed formation sandpack (versus a clean silica
sand sandpack) surrounding the screened interval may
compromise the efficient hydraulic operation of the
screen slots, and result in a less accurate measurement
of well hydraulic characteristics."
ReSDonse 4
The wells where collapsed formation was used for sand
pack were installed primarily in SP, GP or SP-SM type
soils (W26I, W27I, W35I). The hydraulic conductivity
tests at three o! the wells r~!ulted in relatively high
values (8.4 x 10 3to 1.6.x 10 em/s) indicating the
collapsed formation probably formed an appropriate sand
pack for hydraulic conductivity testing.
Comment 5
"In some instances, collapsed formation was utilized as
the backfill for the upper portion of the borehole above
the screened interval, instead of a bentonite slurry
grout.
15
-------
It is unclear whether collapsed foraation will create an
adequate seal to prevent vertical aiqration of
contaminants wi thin the borehole. As best, the use of
this material results in uncertainty as to the extent of
the true seal or whether vertical co..unication exists."
ResDonse 5
The seal at the two wells W36S and W15S that do not have
bentonite added to the backfill 8Uat rely on the surface
and lower seal to prevent migration through the
unsaturated zone. However, in these two cases the
construction ~~pears to be adequate because of the lack
0: contaminan~s detected. -
Comment 6
"In some instances, collapsed formation was utilized as
the backfill for the lower portion of the borehole below
the screened interval in deep wells instead of a
bentonite slurry grout.
It is unclear whether collapsed formation will create an
adequate seal to prevent vertical migration of
contaminants wi thin the wellbore. At best, the use of
this material results in uncertainty as to the extent of
the true seal or whether vertical communication exists."
ReSDonse 6
See the response to Co_ent 2 above. In addition, a
bentonite slurry does not provide the stability upon
which to install well and sand pack. The collapsed
formation typically does not extend more than a few feet
below the well screen.
Comment 7
"In some instances where a deep boring was modified to
accommodate a shallower well, no lower bentonite plug was
placed in the bottom of the screened interval.
since collapsed formation was utilized as the backfill in
many instances, a bentonite plug becomes an essential
component to alleviate uncertainty as to the possibility
of vertical communication within the wellbore."
ResDonse 7
See the response to comment 2 above.
16
-------
Comment 8
"Well cuttings were utilized as the backfill material for
well W-34SR.
It is unclear whether the cuttings were contamination
free or of a proper composition to create a sufficient
seal to prevent vertical migration of contaminants within
the wellbore."
ResDonse 8
The backfill at W34SR used a granular bentonite cuttings
mix, so the seal should be effective.
Comment 9
"Since these well construction deficiencies tend to
result in an ambiguous open interval with no way to
verify the true open interval, technical decisions based
upon information gathered from these wells may be
unreliable.
In an extreme example, an improperly constructed well
which pierces a confining layer can act as a conduit for
vertical migration of contaminants within the wellbore."
ResDonse 9
As stated in the responses to comments 1 through 8 above,
the wells were constructed to provide a good connection
between the well screen and 'the adj acent permeable sand
and gravel aquifer. The seal above the filter pack
provides an effective seal for the well screen within a
small zone within the aquifer being monitored and to
prevent migration between aquifers.
Comment 10
"The descriptive boring logs for the individual wells
should contain symbols which are representative of the
type of lithologies present within a particular
subsurface interval in the wellbore. Several Warzyn logs
utilize different lithologic symbols for descriptions
containing identical verbiage. Additionally, there
exists a poorly defined overlap of acceptable verbiage
between several lithologic units."
17
-------
ResDonse 10
The example of the inconsistent use of litholO9ic symbols
and descriptions occurs at only one location. The
description and symbol at WSD is an obvious typographical
error. The soil descriptions and symbols received
adequate scrutiny for boring 1098 on the cross sections
to be a reliable indication of the soil type present.
Comment 11
"':'herefore, i.n light of the above-mentioned,
l.:!conSl.st.. -c::s, it is unclear whether Warzyn's
~~terp~et~c~on of a "five layer" system of continuous
till and aquifer units across the study area is an
accurate interpretation of the local stratigraphy.-
ResDonse 11
u. S. EPA disagrees. The Agency can se. no
inconsistencies in well construction that would call into
question the conclusions of the RI.
18
-------
SBC'f:IOB V
Gradient corporation provided a critique of the Baseline Risk
Assessment for the sturgis Well Field Superfund site. In
qeneral, Gradient corporation comments on the Baseline Risk
Assessment relate to the conservative nature of the '
assessment. Gradient corporation asserts that the assess.ent
is based on "worst case" assumptions which result in an
"unreasonable" versus "reasonable" risk.
ResDonse overview
Although Gradient has raised comments abOut the
assumptions made in the Risk Assessment, it is necessary
to note the very clear risk at ~is Site. Groundwater
has been found to be contaminated with TCE and PCE at
levels many orders of 8I1gnitude above MCLa. Tbia level
of contaaination will surely be unacceptable and require
action on any scale of risk, and debates over specific
assumptions are academic. In fact, the exceedances of
MCLs in qroundwater are, by theaselves, sufficient to
require action at the site. Once action is justified by
unacceptable risk and/or MCL exceedance(s), all ARAR.s .
apply.
ReSDonse
Gradient's review expressed concern at the conservative
nature of the assumptions used in the Baseline Risk
Assess.ent. Assumptions used in the Baseline Risk
Assess.ent were develoPed from the draft Michiqan Risk
Assessment Guidelines prepared by the Michiqan Council on
Environmental Quality and from the u.s. EPA Risk
Assessment Guidance for superfund (RAGS). U.S. EPA
supplemental guidance was not available when the risk
assessment was prepared. The new suppleaental guidance
may have chanqed some of the assumptions used to assess
exposure as indicated by Gradient corporation (e.q., 30
year exposure duration versus a 70-year lifetlae
exposure). However, it is necessary to note that RAGS
states (emphasis added):
"National statistics are available on the
upper-bound (90th percentile) and averaqe
(50th percentile) number of years spent
by individuals at one residence (EPA
1989d). Because of the data on which
they are based, these values may
underestimate the actual time that
someone miqht live in ,one residence.
Nevertheless, the upper-bound value of 30
years can be used for exposure duration
19
-------
when calculating reasonable aaxiaum
residential exposures. In.088 O88e.,
however, a lifetiae ezpo.ure (70 year. by
convention) -y be a aore appropriate
a..uaption. Con.ult with the ...
reqardin9 the appropriate ezpoaure
duration for re.idential ezpoaure..-
Gradient also challenged the use of maxi.um detections to
represent contaminant concentrations. U.5. EPA guidance
recommends that the concentration of a contaminant in a
g~ven ~edium (groundwater. soil, etc.) used to represent
the exposure-point concentration be derived by
c31cula~ir.q the 95% ~ppec ~onfidence limit on t~e mean of
sample concentrations (9~~ UCLM). If this val~ ~xceeds
the maximum value identified, the aaxi8ua value ..3 to be
used instead. In the 5turgisMUnicipal .ell Fi~~i
assessment, 95' UCIJ( value. were. calculated for ~ie .ost
frequently identified contaainants in groundwate; TCE
and PeE. These 95' UCIJ( values were greater than the
aaximum concentrations identified for these c08PQunds,
which is due to the large degree of variability within
the contaminant concentration data.- Therefore, aaxi.ua
contaminant concentrations were used to represent
exposure-point concentrations for these data, as well as
for all contaminant concentrations in other media. The
use of maximum contaminant concentrations for other
contaainants was based on professional judgement after
reviewing the distribution of saaple results for other
contaainants of concern. Assuaing a lognorJlal
population, the use of aaxi.ua contaainant concentrations
is reasonable when the variance of sa8Ple results is
high. In addition, U.5. EPA Region V support. the use of
maxi.um contaainant levels in groundwater exposure
scenarios. The philosophy is that if an individual
constructs a private well he may very well pull
groundwater from the most contaminated part of the plume.
Many of the proposed methodoloqies and approaches put
,forth by Gradient in their document would not be
acceptable to U.5. EPA. For example U.5. EPA does not
consider it valid in a risk assessment that a fence
(institutional control) will always exist at the site and
completely eliminate childhood exposure to contaminated
media. This is particularly important in light of the
fact that access has not always been restricted.
Assumptions regarding limited access, such as a fence,
are generally considered by U.5. EPA as an unacceptable
approach to determine baseline risks.
The U.5. EPA Environmental criteria and Assessment Office
(ECAO) is currently working on a draft approach for risk
20
-------
assessment for PARs at superfund sites (i.e., relative
potencies). Presently, there is no u.s. EPA position on
this issue despite the claim made by Gradient corporation
that the toxicity equivalence factor concept is
acceptable to the Agency. A conversation with ECAO
(september 4, 1991) indicates that the Agency is
requiring that slope factors for benzo(a)pyrene be
adopted for PARs with B2 carcinogen classifications.
Gradient corporation has used relative potencies for
carcinogenic PARs obtained from scientific literature.
The Gradient approach would therefore not be acceptable
to the U.s. EPA.
Comments made by Gradient concerning other possible
sources of conta~ina~ion at the Kirscn property are
no~ed.
21
-------
.BC'f'%08 n
Cooper geologist Chris Smith provided comments regarding NPL
site pumping, treating and use of treated water in municipal
system.
RESPONSE
Mr. Chris Smith provided a summary of sites where
groundwater is currently being treated for public
consumption. u.S. EPA has no underlying objection to the
use of treated water, as demonstrated by the statement in
the u.s. EPA and MDNR Proposed Plan that the municipal
use of treated water should be investigated as an option
in RD/RA. Preli~~~ary discussions between the MDNR and
the City of Sturgis confirmed that the City was indeed
interested in this option.
The comparison of the Sturgis Municipal Well Field Site
to other sites where treated water is being consuaed is
not appropriate. The decision concerning whether or not
to discharge treated water to the City's water supply is
not one which can be made unilaterally by the u.S. EPA,
MOHR or Potentially Responsible Parties. This is a .
design issue which will require the approval and
involvement of the City of Sturgis.
We encourage Potentially Responsible Parties to
investigate this issue further. It is necessary to note
that, in keeping with this ROD, the final design 8ust
restore the aquifer. The health and welfare of private
well users must be held in as high a regard as that of
municipal users.
22
-------
,BC'rIO. VII
HydroSearch, Inc. (HSI) provided extensive review co_ents of
the RI/FS for Cooper Industries. Comments are presented and
responded to in the sequence followed in the HSI co_ent
document. "Text" is the quoted portion of the RI or FS
document on which Cooper provided comment.
RI Executive SummarY Comment 1 CRI.ES.l):
Text:
"By May 1983, Gove Associates had installed and sampled
four wells, but were unable to locate the VOC plume or
source."
Comment:
"The actual finding of the Gove report was that no
VOCs were detected in the four ~onitor wells or
PWl/PW2 . "
Response:
Comment noted.
Comment RI.ES.2:
Text:
"The types and concentrations of VOCs present in the Ross
Laboratory groundwater samples were not made available
during the course of the RI."
Response:
"If the data were not available, how is the conclusion
drawn that the wells are impacted? Where is the
documentation to support this?"
The data relating to Ross LabOratories was provided to
MDPH by Ross Laboratories and is contained in Michigan
Department of Public Health files, but is considered
confidential.
Comment:
Comment RI.ES.3:
Text:
"The groundwater flow model developed for the area
confirmed observations made regarding groundwater flow
direction (Drawings 12868-23 and 12686-24)."
Comment:
"We disagree that the ground-water flow model for the
area confirmed observations on Section 9.0." .
Response:
See responses to comments made in section 9.0.
23
-------
Text:
Comment RI.ES.4:
Comment:
Response:
"The advective particle tracking model coupled to the
flow model, showed potential travel path directions for
non-dispersive, non-reactive solutes entering the flow
system at discrete locations and tiae intervals (Drawing
12868-25), and helped to reinforce ob..rvations regarding
areal and vertical extent of contaaination."
"We disagree that the particle tracking .odel
reinforces observations regarding the areal and
vertical extent of contamination. See comments in
Section 9.0"
See responses to comments made on Section 9.0.
Text:
RI Section 2 Comment 1 (Comment RI.2.1):
Comment:
Response:
Comment
Text:
Comment:
Response:
"The City currently plans to impl88ent treatment for VOCs
at well PW3."
"An air stripper pilot test was perforaed at
However, the City does not plan to impleaent
for VOCs at well PW3, but rather installed a
at Thurston Woods (PW6)."
Comment noted.
PW-3.
treatment
new well
RI.2.2:
"Although previous sampling showed that well PW4 was not
contaainated, the fourth groundwater saapling round
(August 1989) performed during the RI detected low level
contamination (1 ug/L of TCE)."
"Subsequent sampling perforaed by the DepartJlent of
Health has demonstrated that the fourth round
detection was a false positive."
CODent noted.
Text:
Comment RI.2.3:
Comment:
"Although reportedly installed in close proximity to a
former landfill site, well PW5 has remained free of VOC
detections, and currently serves the City as the primary
well in the system."
"According to the City Manager, Mr. John Brand, the
24
-------
Response:
well is installed through the landfill."
Comment noted.
Text:
Comment RI.2.4:
Comment:
Response:
"Although Kirsch Co. Plant No.1 ba8 two wells and Kirsch
Co. Plant No.2 has one well, none of these wells
currently function as water supply wells."
"Kirsch Plant No.1 has three wells: two are equipped
with pumps but not currently used, and one has been
capped with a welded plate."
Comment noted.
RI Section 3 Comment 1 (Comment RI.3.1\:
Text:
Comment:
Response:
Comment
Text:
Comment:
Response:
"In addition to groundwater sample collection at each
monitoring well, depth-to-water and total well depths
were measured at monitoring wells and compared to as-
built well depths to establish well identity."
"What were the conclusions of this well survey? Were
any wells found unsuitable for sampling and redrilled?
Were any wells constructed using PVC glue?"
Conclusions of the comparison were not formalized in
writing. No wells were found to be unsuitable for
. sampling and redrilled. No wells were constructed
using PVC glue.
RI.3.2:
"Appendix B contains details of the water level and
hydraulic conductivity testing proqraas, Table 5-2 shows
groundwater elevations of monitoring wells."
"Hydraulic conductivity test programs are not
described in Appendix B and the raw data were not
included in any appendix."
Comment noted. Cooper representatives obtained the
hydraulic conductivity data from Warzyn's office on
August 5, 1991.
25
-------
RI section 4 Comment 1 (Comment RI.4.11:
Text:
Comment:
Response:
"The upper bedrock unit in the area i8 a8sumed to be
Coldwater Shale of the Kinderhook Series, Mississippian
in age."
"What is the reference from which the upper bedrock
unit is Coldwater Shale? How thick is the shale?
What units underlie the shale? Bow effective of an
aquitard is the shale?"
The reference is the Western Michigan University 1981
Hydrogeologic Atlas of Michigan. Plate 6 of the
reference indicates that the Coldwater Shale is
bedrock geology wi~h a thicrJ1ess of bet~een 180 and
340 feet. Beneath the Coldwater Shale is Sunbury .
Shale, Ellsworth Shale and Atria Shale, and following
that, Traverse Formation Carbonate Sequence. Plate 7
from this document indicates that Coldwater Shale is
an aquiclude, although it is also called a aarqinal
aquifer.
Text:
Comment RI.4.2:
Comment:
Response:
Comment
Text:
Comment:
"In the absence of pumping, groundwater flow may be
radially outward from the City. This probable pre-
pumping conditions was recreated in the groundwater
aodel (see Section 9)." .
flow
"How does this assumption of radial flow apply to both
the shallow and deep aquifers? How was the pre-
puaping condition recreated in the flow .odel? What
head levels were used?"
Radial flow is assumed to apply regionally to the
shallow aquifer only. The model was never calibrated
to pre-pumping conditions. Because of the lack of
pre-pumping statics, head levels fro. pumping
conditions were used.
RI.4.3:
"Therefore, there does not appear to be a strong
influence of a regional gradient on the groundwater flow
system."
"The regional flow direction is generally to the west
towards the st. Joseph River and ultimately Lake
Michigan. The local flow in the deep aquifer is
consistent with this as shown in Drawings 12868-11,
26
-------
Response:
12686-12, and 12686-13. Therefore, the regional
gradient does have a strong influence on the local
qroundwater flow system."
Comment noted. Both regional flow and groundwater
pumping affect local flow.
Text:
Comment RI.4.4:
Comment:
Response:
"As described above, the multi-layer aquifer system
underlying the site consists of thick sand and gravel
sequences separated by low permeability till and/or
lacustrine units."
"The permeability of the till is assumed to be low
although there is little or no bard evidence to
support this assumption. out of 38 slug tests
performed at site monitor wells, only two were
performed in till units. One test in the lower till
(beneath the lower aquifei) indicated a hydraulic
conductivity of 8.4 x 10- em/see (W4DD). The second
test was done in the upper till and indicated a
hydraulic conductivity of 1.3 x 10-4cm/sec (W6S). No
lab permeability tests were conducted to determine the
vertical hydraulic conductivity."
Co_ent noted. The need for confiraation of hydraulic
gradients can be further discussed during RD/RA.
Text:
Comment RI.4.5:
Comment:
Response:
"Total monthly precipitation in the vicinity of the site
ranges from 1.6 to 4.0 in., with total precipitation
approximately 34 in. per year."
"What is the reference for these data? Based on
climatic maps from the Climatic Atlas of the united
states,' u.s. Department of Commerce, National Climatic
Center (1979), the normal annual precipitation in the
sturgis area is approximately 36 inches."
NOAH Climatic Summary
RI section 5 Comment 1 (Comment RI.5.1\:
Text:
"These boring logs were used to construct six geologic
cross-sections through the study area (Drawing 12686-15
through 19)." .
27
-------
Comment:
Response:
"Two of these cross-sections (BB" and FF") were
independently reinterpreted using the data available
in the RI documents. This data included borehole
logs, grain size analyses, hydraulic conductivity test
resul ts, and gamma ray logs. Tbe reinterpreted cross
sections are presented in the HSI Plates I and II.
In general, the reinterpreted cross sections show auch
less continuity in the till units c08pared to the
cross sections given in the RI. We do not believe the
data justify some of the interpretations presented in
the RI and, as a result, the correlation of geologic
units between boreholes has been oversimplified and
overinterpr~~ed. The simplification of the site's
hydrogeology into five layers aay facilita~e the
modeling effort at the site, but i. not coapletely
justifiable."
Comments ~d submittals are noted. Co..entors raise
important issues which should be discussed further
during RD/RA.
Text:
Comment RI.5.2:
Comment:
Response:
". distal outwash (silt, clay and silty clay deposited in
a low energy environment typically at larger distances
from the glacier)."
"In the context of the geologic cross sections and the
subsequent ground-water model, were areas of distal
outwash treated as high para.ability outwash or low
paraeability till?"
There was a typographical error on Cross section CCI.
Lower till was mistakenly labeled as "lower outwash".
In general, this geology would be low permeability.
Comment RI.5:3:
Text:
Comment:
"The predominant facies in this sequence consist of
proximal, coarse grained outwash, with distal fine
grained silt and clay outwash deposits varying from 0 ft
thick at well KW1C to 33 ft thick at well W5DD (Table 5-
1) .
"Thirty-three feet of dis~al fine grained silt and
clay outwash does not appear in the geologic log for
well W5DD. In addition, the reference to Table 5-1
appears to be incorrect. This facies model was not
used in the ground-water flow model, which used
28
-------
Response:
homogeneous units. Why wasn't the heterogeneity
described here used in the model?"
See response to comment RI.5.1 above.
Text:
comment RI.5.4:
Comment:
Response:
The intermediate depth till unit was encountered between
approximate elevations 750 ft and 850 ft. msl at most
wells drilled to this depth."
"Till unit thicknesses were tabulated in HSI Table 1
and isopach maps of ~he "upper" and "intermediate"
da~th till units were prepared and are presen~ed in
HSI Plates III and IV. HSI Plate V combines the data
from the two and indicates where no confining till
units are present in the aquifer. HSI Plate V
indicates that the till "window" extends fro. the
central part of the city to the southwest towards well
PW-4. Do the till unit geemetries presented here
match what was used in the model? See co.-ents on
Figure C.4 in FS Volume 2 of 2."
See response to comment RI.5.1 above.
Text:
comment RI.5.5:
comment:
Response:
"The aajority of this till consists of clay silt, with
sandier till encountered at wells W7D, W8D, and City test
well TW84A."
"Because of the sandier till, were these areas
simulated as "windows" in the till in the ground-water
model?"
No.
Comment RI.5.6:
Text:
Comment:
Response:
"A surficial or upper till unit was encountered above the
previous outwash sequence except at well locations R-2,
WilD, W30D, and W32D."
"See Comment 4."
(Comment RI.5.4 above)
See response to comment"RI.5.1 above.
29
-------
Text:
Comment RI.S.7:
Comment:
Response:
co_ent
Text:
Comment:
"The magnitude of vertical gradients aay also be
indicative of the proximity of a well nest to an active
pumping well or point of ground-water discharge."
"Vertical hydraulic gradient data were calculated for
selected water level rounds and are su.aarized in RSI
Table 2. HSI Plate V graphically presents vertical
gradient data for well nests around the site. The
highest gradients (>.05) are associated with the thick
~ill units along the southwest side of the site. The
lowest qradients «.01) are associated with the till
wi~~ow ~~2~ extends from the center of the site
towards P*-4 to the sou~~east. This data indicates
that the vertical gradients are primarily controlled
by the vertical hydraulic resistance created by the
till. The lack of puapingwell influence observed in
the deep aquifer suggests that the water table surface
configuration and vertical gradient distribution are
not controlled by deep aquifer pumping."
Pumping influence was observed in .the shallow and deep
aquifers. Pumping influence in the shallow aquifer
depends on whether the confining layer was sealed
during pump construction (i.e., Ross Wells).
RI.S.8:
"Drawings 12686-7 through 12686-10 are water table aaps
which display the relationship of the flow eyet.. to low
permeability till units and pumping cent~rB."
"Shallow water level data fro. April, 1989 (Drawing
12686-9) were reinterpreted and are presented in HSI
Plate VI. The major differences between the two
include the following: Data fro. W-9S was considered
anomalous compared to previous and subsequent events
and was adjusted according to changes observed in
adjacent wells between January 31, 1989 and April 26,
1989 measurement events. This results in a lesB
pronounced ground-water mound extending from the
sturgis Foundry. The contours along the southeast
side of the site were opened to correlate with the
till window that extends towards PW-4 (see well 109 in
HSI Appendix B). We disagree that the water table
depressions are caused by the pumping in the lower
aquifer. We believe these depressions would exist
without aquifer pumping in response to vertical
hydraulic resistance."
30
-------
Response:
Depressions ~ exist without aquifer pumping in
response to vertical hydraulic resistance. See
response to comment RI.5.1.
Text:
Comment RI.5.9:
Comment:
Response:
"The water table maps show water table conditions which
are relatively consistent during this investigation."
"The maximum and minimum water levels recorded between
November, 1987 and November, 1989 were tabulated and
are included in HSI Table 3. This table shows the
lowest water levels occurred in late 1988 and are
probably associated with the drought. In general, the
maximum change in water level over this period was
about 2 feet. The table also identifies wells with
anomalous water level changes including W98, W26D,
MWlD, and TW84A. The range observed at TW84A is
probably the result of its proximity to City well PW-
5."
Comment noted. U. S. . EPA maintains the water table
conditions were relatively consistent.
Comment RI.S.lO:
Text:
Comment:
Response:
"Drawings 12686-11 through 13 are potentiometric surface
1I&pS which show the head in the. lower aquifer. Pullpage
fro. municipal and industrial wells appear to have
reduced the head in the lower aquifer throughout the City
and provides the mechanism to produce the downward
vertical gradients discussed above."
"Deep water level data from April, 1989 (Drawing
12686-13) were reinterpreted and presented in HSI
Plate VII. The reinterpretation is siailar to that
provided in the RI report and indicates that the flow
is east to west with little observed influence of the
pumping wells. Because of the lack of response in the
lower aquifer due to well pumping, we disagree that
pUJIp&ge in the lower aquifer is the mechanism that.
produces the water table depression (See Comment 7)."
U.S. EPA disagrees. There is a response in the lower
aquifer due to well pumping and well pumping
contributes to the downward vertical gradient.
31
-------
Comment R1.5.11:
"Test methods are contained in Appendix B."
Text:
Comment:
"Test methods are not included in Appendix B nor are
raw data.
The hydraulic conductivity data were obtained from
Warzyn's office on August 5, 1991. Based on a review
of this data we have the following co..ents:
. A least squares analysis is appropriate in most
cases, .bu ~:: when there is extreme scatter :..n data
pc~nts, J~~gemen~ ~h~uld be used ra~~er t~~n
statistical method. Wells that show sca~ter snould
be retested if they are not representative of the
morpholoqical unit the vells vere coapl.teeS in.
W40D
W39D
W36S
W32D
W30D
W28D
W20S
W15S
W6D
Retest
Retest
Recalc/Retest
Recalc/Retest
Recalc/Retest
Recalc/Retest
Recalc
Recalc
Recalc/Retest
. The tabulated data does not show how the translation
was made from water level to drawdovn.
Of the hydraulic conductivity tests perforaed during
the R1, only one was perforaed on a till unit. This
was at W5DD which is screened in the lower till and
showed a hydraulic conductivity of 8.4 x 10- em/sec.
Due to the lack of tests within the till units, an
approximation of hydraulic conductivity was made using
the grain size distribution data included in Appendix
C-5 of the RI. The grain size analyses were evaluated
using the Hazen method (HS1 Appendix C) to estimate
the hydraulic conductivity (BSI Table 4). Although
this is most applicable for sand size materials, it is
the best equation available and its limitations for
use on clay are recognized. To establish the validity
of the results, the Hazen method results were compared
to the slug test results if the grain size analysis
was performed on a sample from the screened interval
of the well. Unfortunately, only one measurement
meets this criteria which was W35S (BS1 Table 4). The
slug test result shows a hydraulic conductivity of
32
-------
Response:
9.68 x 10-3which indicates about one order of
magnitude difference between the two. Therefore, when
reviewing the Hazen results, the calculated hydraulic
conductivity is approximately one order of magnitude
lower than comparable slug test results.
HSI Table 4 indicates that 11 grain size analyses were
performed on till units. Tbe Hazen resufts for these
units range from 2 21 x 10-~0 2.5 x 10- em/see (2.21
x 19-4to 2.5 x 10-6em/!ec adjusted) and average 8.16 x
10- em/see (8.19 x 10- em/see adjusted). These
calculations indicate that the hydraulic conductivity
of the till is approximately two to three orders of
magnitude lower than the outwash."
Comment noted. U.S. EPA maintains that the method
used in the RI was sound. Best 'professional judgement
was used.
RI section 6 Comment 1 (Comment RI.6.1\:
Text:
Comment:
Response:
"Duplicate analyses of samples at well W42S during Round
4 showed very poor agreement for TCE (6,500 ug/L, W42S-04
and 3.0 ug/L, W42S-94). These data were qualified as
unusable ("R")."
"If the data are qualified as unusable, why were they
included on the isoconcentration aap in drawing
700840-B26."
Co_ent noted. The 1,000 ppb contour should have been
dashed to show it was estbaated.
Text:
Comment RI.6.2:
Comment:
Response:
"VOC data ,fro. the analysis of six .-.ples collected
during Ro~d 2 were determined to be unusable as a result
of poor laboratory performance (failure to meet internal
standard or surrogate recovery criteria). These samples
include W4S-02, W6S-02, W8S-02, MW2C-02, FB01-02 and
FB03-02."
"TWo out of five field blanks failed to meet internal
recovery standard or recovery data and a third field
blank contained 2 ppb PCE. This leaves only two field
blanks for this round of data. How can the RI report
state that the loss of these samples is not critical."
The loss of the five samples was not critical to the
33
-------
RI because results from other sampling rounds were
available.
Text:
Comment RI.6.3:
Comment:
Response:
"Data from these analyses are presented in Tables 6-4, 6-
5 and 3-3."
"Table 3-3 does not present data.
reference should be to Table 6-6."
We believe the
Comment r.c":ed.
Text:
Comment RI.6.4:
Comments:
Response:
Comment
Text:
Comment:
"Results of field GC analysis of soil headspace (Table 6-
5) were, qualitatively, in general agreement with results
of the CLP analyses of soil borings ( i . e., cOllpOunds
detected at concentrations greater than approximately 30
ug/kg were generally identified by both methods."
"As shown on HSI Figures 1 and 2, there is a very poor
correlation between the total ethene values derived
from field GC analysis and CLP data for samples -from
the same sampling location (boring and depth). If any
correlation can be made between the field GC and CLP
data, the trend would indicate that the field GC
indicates total ethene concentrations an order of
magnitude greater than CLP data of the same soil
sample. The use of the field GC is not valid and
cannot be justified or defended with properly
validated CLP data."
The text states that GC and CLP were aualitativelv,
not quantitatively, in agreement. The field GC was an
important tool for identification of possible source
areas by detecting the presence of VOCs.
RI.6.5:
"Results of VOC analyses measured by field GC were in .
general agreement with corresponding CLP analyses. These
results were considered acceptable as a semiquantitative
screening tool in the RI."
"See Comment 4. These results were not used just for
screening: they were also used as the basis for
conceptual design of remedial alternatives. Because
the field GC data greatly exaggerate the magnitude and
34
-------
Response:
extent of impacts, the resultant FS alternatives S2
and S3 were overdesigned."
GC was used for determination of the presence, or lack
thereof, of VOC contamination. GC is a auch .ore
efficient and cost effective tool than extenaive CLP
analyses. u.s. EPA will consider conducting
addi tional CLP soil sampling as part of the pre-design
phase of RD/RA.
RI section 7 Commentl (Comment RI.7.l):
Text:
Comment:
Response:
"VOCs were detected in 21 of the 35 samples."
"The 21 samples with detected VOCs included 4
duplicate analyses. The actual nuaber of detects was
17 out of 35 samples."
Comment noted.
Text:
Comment RI.7.2:
Comment:
Response:
"Each of the 47 soil samples collected for headspace
analyses from these borings contained quantifiable
concentrations of VOCs, except two samples (SB-01R at 2.5
ft, and SB-05 at 5 ft) which had results at below aethod
detection limit (BMDL)."
"There is virtually no correlation between soil gas
. and CLP analyses (See BSI Figures 1 and 2). Soil gas
results are roughly one order of magnitude higher than
CLP lab results for soil saaples collected from the
same intervals. Approximately 15' of soil gas
detections corresponded to non-detectable CLP results.
Therefore; the soil gas data greatly exaggerates the
magnitude and extent of iapacted soil and should not
be used as the basis for determining areas needing
remediation."
See response to comment RI.6.5 above.
Text:
Comment RI.7.3:
"VOC concentration at wells W23S and W34S decreased from
over 5,000 ug/L during the Round.3 groundwater sampling
to no detection (ND) during Round 4 sampling. This
decrease may be due to reconstruction of wells W23S and
W34S between sampling Rounds 3 and 4."
35
-------
Comment:
Response:
"What was the reason for the well reconstruction?
Table 5-2 does not indicate W-23S was replaced, How
was is reconstructed? Why would well reconstruction
affect results in such a dramatic way? without
confirmation, the results from these wells should be
considered suspect and not used. The isoconcentration
map in Drawing 70084-B26 used the data from the old
wells. Why?"
Well W345, which was damaged during Phase II hydraulic
conductivity testing, was replaced with W345R. Well
construction details are contained in Appendix Band
summarized in Table 4 of the Phase I18 Technic~l HeMO
available in t~e ~dministra~ive Racord.
During sampling, the sub88rsible puap became lodged in
well W235. It was necessary to r-ove the well,
section by section, to retrieve the pu.p. The well
was then replaced.
U.5. EPA cannot speculate on why groundwater r.sul ts
changed so dramatically. Data fro. the "old" wells
was used because at the time of sampling there was no
problem with the wells.
Comment RI.7.4:
Text:
Comment:
Response:
"The soil and groundwater analytical results suggest a
major source of VOC contamination in the vicinity of well
nest W34 based on:
. high concentrations of VOCs in shallow soils;
. the decrease in voc concentrations in groundwater with
depth in the upper aquifer; and
. the low voc concentrations in the deeper aquifer (3
ug/L at W26D)."
"Why do soil samples at 5B-22 and 5B-23, adjacent to
W-34, contain no to low level VOC's (ND to 2 ppb,
respectively)? How does the decrease in VOCs in
groundwater with depth in the upper aquifer and low
VOC concentrations in the deeper aquifer indicate a
major source of VOCs at well nest W34? Considering
the results at W345, these conclusions are highly
speculative (5ee COJDJDent 3)."
The difference between W34 and 5B-22/5B-23 results
could indicate an area where VOC contamination is not
consistently distributed due to the pattern of
release(s) and/or the input of "clean" fill.
36
-------
Text:
Comment RI.7.5:
Comment:
Response:
"polyaromatic hydrocarbons (PARs, refer
determine specific chemicals classified
identified in soils from several boring
generally at and near the surface."
to Table 7-1 to
as PARs) were
locations
"Buildings that once occupied the open lot at the
Kirsch Plant '1 were razed in 1973 on the eastern
half, and 1982 on the western half of the lot. When
the western half was razed in 1982, clean fill was
brought in to bring the site to grade. If the PARs
f~und on this lot were a result of Kirsch activities,
why are PAR concentrations highest near the surface
(more subject to biOdegradation and photolysis) and
decrease with depth in the deaolition fill?
The highest values of PARs were found in the northwest
corner of the site and decrease to the south and east
(See Figures 3, 4 and 5). Given the PAR distributj.on
vertically and horizontally, the possible sources of
PARs are the bulk oil terminal (Dury Oil Co.) located
adjacent to the northwest corner of the Kirsch lot and
the buried railroad spur along the north end of the
property. During a site visit on July 9, 1991, a
heavy odor of diesel fuel was noted at the Kirsch
property coming from Dury Oil. It is highly likely
that aerial deposition of PARs occu~s on the Kirsch
property as a result of fugitive eaissions fro. the
bulk oil terminal. pieces ot coal and slag tro. the
buried railroad spur were also noted and could
contribute to the PAR concentrations. oil from
locomotives couid be another source.
Why weren't PAR samples collected at the bulk oil
terminal, other properties surrounding the terminal,
and other potential source areas, to evaluate the real
source distribution of PARs? Why weren't PARs sampled
along other railroad tracks? Why were PARs analyzed
for in the first place when the impacts to the water
supply are VOCs." .
Information concerning other possible sources of PARs
is noted. Because of the lengthy manufacturing
history of the Kirsch property, it would not be
surprising to find PARs and other common industrial
contaminants in the soil. Therefore, the sampling and
analyses was not restricted to VOCs. u.s. EPA will
consider performing additional PAR sampling during the
predesiqn phase of the RD/RA to shed further light on
the extent and source of PAR contamination.
37
-------
Comment
Text:
Comment:
Response:
RI.7.6:
"One sample collected from borinq locations at a depth of
2.5 ft contained low concentrations (6 uq/kq or less) of
benzene, ethylbenzene, and xylenes."
"Which borinq?"
SB-02 sampled in Auqust of 1988 contained 4 uq/kq of
benzene, 3 uq/kq of ethylbenzene and 5 uq/kq of total
xylenes. -
Text:
Comment RI.7.7:
Comment:
Response:
Comment
Text:
Comment:
Response:
Comment
Text:
"Samples from wells W11S and PW6 were considered
representative of backqround conditions tor inorganic
qroundwater quality. Sa8ples fro. locations SB15, SB16,
SB17, SB29 and SB31 were used to characterize backqround
levels of inorqanic constituents in soil."
"The number and location of backqround soil and
qroundwater samples is inadequate to describe the
potential variability of naturally occurrinq eleaents
in the soil and qround water."
U.S. EPA disaqrees and feels that sufficient
backqround Samples have been taken. AI thouqh cleanup
standards are to Micbiqan Act 307 Type B levels, U.S.
EPA will consider additional inv.stiqation of local
backqround levels durinq the predesiqn phase of the
RD/RA.
RI.7.8:
"Groundwater collected from well W23S contained
approximately 18 uq/L of Chromium."
"Two samples were collected from this well with
results of 6 and 18 uq/L/ The text should reflect the
ranqe observed or state that 18 uq/L was the maximum
detected." .
Comment noted.
RI.7.9:
"The chromium concentration in soil borinq SB02 at the
2.5-ft depth was 62.8 and is a180 qreater than Site
backqround concentrations which ranqed up to 20 uq/kq."
38
-------
Comment:
Response:
"See Comment 7. The range and average concentration
of elements in natural soils across the U.S. i.
presented in USEPA Office of Solid Waste and Eaergency
Response, Hazardous Waste Land Treataent, SW-874
(April, 1983) Page 273, Table 6-46 (BSI Table 5).
This table indicates the range of naturally occurring
chromium is 1 to 1000 ppa with an average ot 100 pp..
The chromium concentrations at S802 of 62.8 pp. is
below the average of 100 ppm."
The aver~~e of 100 ppm is not specific to the locale.
Samples -aken as background references provide a much
be~~er comparison than does a national average.
Text:
Comment RI.7.10:
Comment:
Response:
Comment
Text:
Comment:
"Groundwater samples from well W23S and well W34S
contained cyanide at concentrations of approximately 30
ug/L and 284 ug/L, respectively."
"The results from W23S were 23 ug/L and 34 ug/L. The
text should reflect this range. The results for W34S
were 174 ug/L and 284 ug/L (duplicate analysis 247
ug/L). Stating the concentration in W34S is 284 ug/L
is misleading considering the duplicate analysis and
range. A round of samples collected on Septeaber 9,
1990 was split with Kirsch. The Kirsch sample results
indicate a concentration of 190 ug/L which i. below
the MCL of 200 ug/L.
Comment noted. Please note that the Act 307 Type 8
level for cyanide is 100.
RI.7.11:
"Thirty-nine of the 49 soil samples collected at Wade
Electric (Table 6-6) contained detectable VOCs fro. field
GC methods.... Of the 38 CLP samples collected from
borings S809, S810, S811, S812, and well W33 (Appendices
F-3 and F-6), 13 samples had quantifiable levels of TCE."
"As previously stated in Comment 2 (Section 7.0) and
Comment 4 (Section 6.0), soil gas results greatly
exaggerate the magnitude of contamination. At the
Wade site, 39 of 49 samples (80') had detectable VOCs
with the in-field GC. This compares to 13 of 38
samples (34%) with detectable VOCs determined from VOC
samples. Why was there such a significant difference
in results?"
39
-------
Response:
See response to co_ent RI.6.5. As to why there is a
significant difference between GC and CLP results,
that is an academic question not necessitating a
response in this document.
Text:
Comment RI.7.12:
Comment:
Results:
"Twenty-four groundwater samples were collected and
analyzed using field GC methods during the installation
of :~l7S, W18I, W19S! and W33S (Drawing 12686-5).
Nineteen of the samples had quantifiable VOC levels~.
"Given the poor correlation between field GC and CLP
results, of what value are these data? See Co..ents 2
and 11." .
See responses to comaent RI.7.2 and RI.7.11.
Text:
Comment RI.7.13:
Comment:
Response:
Comment
Text:
Comments:
Response:
"A soil sample from W33S at 1.5 ft in depth contained
lead at a concentration (167 aq/kg) which exceeds the
background level (approxi..tely 35 IIC)/kg). This
concentration, however, is within a typical range for a
suburban area."
"This taplies that
not representative
See Comments 7 and
above)
the background saaple results are
of actual conditions in sturgis.
9." (Co_ents RI.7.7 and RI.7.9
u.S. EPA disaqrees with the
samples are representative.
determined from the mean of
would be a range that would
variability. See responses
co_entor. Backqround
"Backqround" would not be
background samples, but
account for natural .
to RI.7.7 and RI.7.9.
RI.7.14:
"However, this well was constructed with galvanized steel
casing which would explain the zinc concentrations, and
possibly the other metals."
"What other wells are constructed of galvanized steel
casing? Why were metals sampled from a well with this
construction?"
See the well logs in the Administrative Record for
identification of wells constructed of galvanized
40
-------
steel casing. The fact the so.. wells had galvanized
steel casings would not necessarily interfere with all
metal results, especially cyanide, which is a
contaminant of concern.
Text:
Comment RI.7.15:
Comment:
Response:
"Telemark Business Forms has been at its current location
since January 1980."
"The land use at the location prior to 1980 was by
Marvel Manufacturing which used solvents containing
TCE."
Comment noted.
Text:
Comment RI.7.16:
Comment:
Response:
"The low "levels of TCE in the groundwater do not appear
to be related to the facility."
"The TCE may not be related"to its current operations,
but there is every reason to suspect the previous land
use could have caused the observed impacts (See
Comment 15). This past source of TCE could have
directly impacted the contamination at the Ross wells
(R1-R4). An apparent window in the till unit near the
Ross wells (See water table maps) could allow
contaminants in the shallow aquifer to migrate
directly to the deep Ross industrial wells."
The Telemark facility does not appear to be a
significant source of contamination to the City's
aquifer. Soil contamination near the facility is
limited to PCE contamination. The extent of PCE
contamination in the aquifer i. limited to the
immediate vicinity of the facility and does not appear
to be related to the contamination affecting the
Site's water supply wells. At this time, there is no
factual information to support speculation that
contamination from the Telemark property has .
contributed to contamination at Ross Laboratories.
Therefore, Telemark is considered to be a separate
site. Telemark is eligible for ranking to determine
its priority for remediation under Michigan Act 307
remediation program.
41
-------
Text:
. Co~ent RI.7.17:
Comment:
Response:
"Drawing 12686-6 shows that the strong influence of
groundwater withdrawal pulls the contaaination to the
northwest and possibly to the southwest (although
previously uncontaminated in the last round of saapling
detected 1 ppb TCE at municipal well PW4.)"
"The pull of contaminants to the northwest (presumably
by the Ross wells) assumes the contaminants in the
Ross wells are from the Kirsch or Wade Electric
facilities. As stated in Comment-16, an equally
feasible explanation ~~at was never considered in the
RI report is that contaaination in the Ross wells
resulted from a shallow source in that area. Drawings
12686-26 shows two possible source areas, one near
Teleaark (210S) and the other near the airport (W14S).
There could also be other sources of shallow ground-
water contamination that have not been identified.
As far as PW4 is concerned, it is located to the
southeast and not to the southwest so the RI report
statement is incorrect. Secondly, TCE was detected in
one out of four rounds at 1 ppb and is hardly a basis
to conclude contaminants have migrated to that well
from the main plume. Pinally, saapling perfOraed by
the Department of Health has not found PW4 to be
btpacted."
Shallow water contaaination appears to be at low
levels in the areas cited by the co_entor; whereas
contamination at the Ross Laboratories is quite high.
Therefore, it is much more likely that the
contaaination is being drawn from the lower aquifer.
Comments concerning PW4 are noted.
Text:
Comment RI.7.18:
Comment:
"This drawing is a cross section which shows a
contaminant plume beginning near the Kirsch Company Plant
NO.1. The plume proceeds along a curved route to well
R4.... The high levels of contamination remain in the
lower outwash unit until intercepted by well R4."
"See Comments 16 and 17. The interpretation in
Drawing 12686-20 implies the contaminants in the Ross
wells potentially came from the Kirsch facility and
ignore the potential for a shallow source close to the
Ross wells to cause the observed impacts. Drawings
12686-20 shows potentiometric lines and
isoconcentration contours of contaminants. The
42
-------
Response:
potentiometric data indicate a downward flow direction
in the vicinity of the Ross well, yet the
isoconcentration contour suggests the flow is upward.
The potentiometric data support the concept of a
shallow source near the Ross wells and not migration
from the Kirsch facility.
Please note that potentiometric lines do not take into
account the influence of the Ross wells on the
hydrogeologic regime.
Text:
C::>lre!lent RI.7.19:
CODlD1ent:
Response:
"Chloroform was identified in sampling Rounds 1, 2, and 3
at well MWlA at concentrations ranging from 2 to 16
ug/L. "
"What activity occurs in the vicinity of this well?
The field observations (Table 7-9, 7-10, 7-11)
indicate water from this well is black with a septic
odor. This could be another shallow source area near
the Ross wells."
See response to comment RI.7.17.
Text:
CODlD1ent RI.7.20:
CODlD1ent:
Response:
"To determine if specific metals in site groundwater were
present in excess of expected natural concentrations, the
sample results were compared to results considered to
represent background conditions (wells W11S and PW6)."
"One shallow and one deep well are not considered
adequate to characterize background conditions across
the City of sturgis."
See response to cODlD1ent RI.7.7.
Text:
CODlD1ent RI.7.21:
CODlD1ent:
"Although well below the EPA Primary Drinking Water
Standard of 1,000 ug/L, these values may be elevated in
comparison to naturally occurring values.:
"See CODlD1ent 20. The RI report also appears to
question the representativeness of the background data
by stating 'these values ~ be elevated in comparison
to naturally occurring values'."
43
-------
Response:
Comment
Text:
Comment:
Response:
See response to comment RI.7.7.
RI.7.22:
'''Althouqh below the EPA Primary Drinkinq Water Standard
of 50 uq/L, these values may be above the backqround
chromium levels for the Site, which are less than
approximately 8 uq/L."
"See Comment 21."
(Comment RI.7.21 above)
See response to c=mment RI.7.7.
Text:
Comment RI.7.23:
Comment:
Response:
Comment
Text:
Comment:
Response:
"Variations in iron and aanqanese are expected in sand
and qravel outwash due to the presence of s8all areas
with hiqh concentrations of these aetals in the sand."
"This statement supports Coament 20."
U.S. EPA disaqrees.
See comment RI.7.13.
RI.7.24:
"...qalvanized steel well. Hiqh zinc concentrations are
co-on in wells of this type ot con8truction and is due
to the qalvanized coatinq applied to the steel."
"See Co_ent 14."
(Co..ent RI.7.14 above)
See response to comment RI.7.14.
Text:
Comment RI.7.25:
Comment:
Response:
"This precipitation may cause operation and maintenance
problems if the precipitation interferes with the
treatment process (e.q., pluqqinq of an air stripper
tower, precipitation of minerals on well screen openinqs,
etc.)"
"If this is the case, why weren't pretreatment options
for water hardness considered in the FS."
If treatability studies to be performed in the
predesiqn phase of RD/RA indicate that pretreatment is
needed, it will be incorporated into the desiqn.
44
-------
Text:
RI Section 8 Comment 1 (Comment RI.8.1}:
Comment:
Response:
"Based on the calculated Kc1 values and aSSWled aquifer
effective porosity (n) and bulk densi~y (Pb), re~arda~ion
factors for TCE(Appendix G) were de~ermined ~o be
approximately 1.1 to 1.8."
"Why weren't these retardation factors used in the
particle tracking model? In PATH3D, input veloci~ies
can be easily adjusted for retardation."
The particle trackinq model was only used as a
tracking tool and not as a tr~'sport model.
Text:
Comment RI.8.2:
Comment:
Response:
"Volatilization may be a significant process resulting in
the loss of some contaminants from soils."
"We agree with this statement and, as a resul~, wonder
why such high values of VOCs were de~ected in surface
soils at the Kirsch property when no activities have
occurred in these areas for approximately 10 years."
U.S. EPA shares the commentor's wonder. The"levels of
contamina~ion mus~ have his~orically been even higher
since significant leaching ~o groundwater and
volatilization to air mus~ have occurred. Any
information the commentor has concerning the 8pecific
releases at the property would be apprecia~ed.
Text:
Commen~ RI.8.3:
Comment:
Response:
"As shown in Section 7, ~he produc~ion wells are capable
of drawing con~aminants from as far as a mile."
"See Comments 16, 17 and 18 in Sec~ion 7.0."
(Commen~s RI.7.16, RI.7.17, and RI.7.18 above.)
See responses to comments RI.7.l6, RI.7.l7 and
RI.7.l8.
Text:
COJlUQent RI.8.4:
"As stated in Section 7.0, four major chlorinated
contamination source areas were identified during
These source areas were identified during the RI.
source areas are as follows:
ethene
the RI.
These
45
-------
Comment:
Response:
.
Kirsch Co. Plant No.1
Wade Electric
Teleaark Business Forms
sturgis Newport Business
Forms."
.
.
.
"section 7.0 discusses the investigative results at
these four facilities, but it does not state that
these four areas are aajor sources of cont&aination.
Only the Kirsch and Wade facilities are identified as
such, and the others are dismissed as possible source
areas. The FS repo~ states only two sources: Kirsch
and Wade."
.. .
The Kirsch and Wade properties vere the only tvo
facilities where contaaination vas shown to have
impacted the vell field.
Text:
Comment RI.8.5:
Co_ent:
Response:
"Based on the Kd dependence on organic carbon shown in
Appendix G, it is apparent that soils vith high organic
carbon content, such as surface soils, are much .ore
effective in sorbing chlorinated ethene compounds than
the underlying deposits. This relationship may explain
the elevated contamination levels observed at the Kirsch
Co. Plant No.1 and foraer Wade Electric facilities (the
age of the spill would also affect the vertical
distribution)."
"The surface aaterial at the Kirsch facility i. a
combination of deaolition backfill and i.ported .ine
spoils from a gravel pit that were placed to grade the
area after building deaolition in 1982. These
materials would not contain a high organic carbon
content compared to naturally developed topsoil. In
addition, see Comment 2 related to volatilization
processes in shallow soils."
using the logic of the commentor, if surface organic
carbon levels at the Kirsch property are not as high
(compared to naturally develoPed topsoil) and less
sorption is taking place than proposed by the quoted
text, the original release(s) at the facility must
have been higher than previously imagined to account
for the current distribution of VOCs. See response to
comment RI.8.2.
46
-------
Text:
Comment RI.8.6:
Comment:
Response:
"Volatilization has likely been an important process in
reducing the TCE and PCE levels in soil. Since some of
the source areas impacts appear to be relatively shallow
and site soils are generally coars. grained, gaseous.
diffusion and release to the atmosphere should readily
occur."
"See C~mments 2 and 5. In addition, at well W1IS,
duplicate analyses of soil samples showed very
different analytical results which causes the high VOC
data to be suspect. At WllS, a sample collected on
October 6, 1987 fro. 1-foot depth bad 260,000 ppb PCE
and 8,200 ppb TCE. On October 19, 1987 (two weeks
later), a sample at W11S froa a depth of 1 foot had
950 PCB and 21 ppb TeE. Is this a r.sult of
volatilization or an indication of the reliability of
data reproducibility."
u. S. EPA will not speculate on the cause of the
different sample results.
Text:
Comment RI.8.7:
Comment:
Response:
"One major qroundwater contaminant plume consistinq of
chlorinated ethene compounds has been detected within the
aquifer. TWo small plumes were identified below the Wade
Electric and Telemark properties in section 7.0."
"A plume was also presented in the vicinity of the
airPOrt (W14S),and another source .ay be present near
the MW1 well nest based on repeated cbloroethane
detections."
No TCE or PCE was detected at MW1. The plume atW14S
consisted of 15 ppb PCE and .6 ppb TCE. The
contamination at W14S appears to be .inimal, primarily
PCE, and does not appear to be a significant source of
contamination to the City's aquifer.
Text:
Comment RI.8.8:
"The plume may be further subdivided into northwestern,
western, southeastern, and southern legs according to the
production wells affected (R1 and R4, northwestern; F1
and PW3, western; PW4, southeastern; and PW5, southern).
An additional leg appears southwest of the plume due to
18 ug/L detected in W39D. Since this southwestern leg is
not caused by a production well, the contaminant
47
-------
Comment:
Response:
migration is apparently due to regional groundwater
flow."
"The northwestern leg to R1 and R4 is one
interpretation and ignores the potential for a nearby
shallow source of contamination (See Co..ents 16, 17
and 18 in section 7). The .oat recent data do not
support the presence of contaminants in PW3:
therefore, the western leg is speculative. The 1 ppb
TCE in PW4 has never been confirmed based on previous
or subsequent sampling and is therefore highly
specula~ive. No ~ontaminants have ever been detocted
in PWS, so the southern leg is illaginative. The data
at W39D were qualified because TCE was also found in
the blank: therefore, the detection has never been
confirmed and the southwest leg of the plume is highly
speculative. In addition, neither the ahallow or deep
flow aapa presented in the RI report support the
concept that contaminants from the main plume can flow
in the direction of well W39. Finally, if the VOC
detection data at W39D were to be confirmed, the
concept of another potential source was never
discussed."
See responses to RI.7.16, RI.7.17 and RI.7.18 for
comments concerning the northern leg of conta8ination.
PW3 is used on a standby basis. Thus, the presence or
absence of contaminants at the well may be linked
directly to periods of puaping. PW3 has shown
contamination.
u. S. EPA maintains that, based on the model in the FS,
low level contamination in the direction of PWS is
likely when PWS is pumping at relatively high rates
and Ross Laboratories is pumping at relatively low
rates. In fact, preli.inary data from the .ost recent
sampling event (August, 1991) haa ahown low level
contamination at TW-84A. Once validation of this data
is complet&, copies will be made available to Cooper
Industries.
Once validated, the data from the August 1991 sampling
event should also serve to clarify the existence of
possible contamination at W39D and PW3 and
contamination in the direction of PWS.
Although the commentor has raised some important
questions concerning the extent to which contamination
has spread from the source areas, these aspects of the
model can be further examined with additional sampling
and refinements of the model in RD/RA. The comments
48
-------
question the boundary of the contaaination and do not
question the use of "PUJIP and treat" as the
appropriate means to remediate groundwater
contamination. If the commentor is correct in his/her
assertion that the "legs" of contamination are not as
extensive as those modeled in the RI, the restoration
of the aquifer will be facilitated.
The comment concerning contamination at PW4 is noted.
Comment RI.8.9:
Text:
"The highest overall 1,1,1 TCA concentration was observed
at well W34S near Kirsch Co. Plant No.1 (9 ug/L)."
Comment:
"In subsequent saapling, this cOlipound was not
detected."
Response:
Comment noted.
Comment RI.8.10:
Text:
"The compound appears relatively consistently at wells
MWlA and MW1B. The potential source of this compound is
unknown."
Comment:
"See Comment 19 in section 7.0." (Comment RI.7.19)
Response:
See response to comment RI.7.17.
Comment RI.8.11:
Text:
"Dibromochloromethane was detected in only two samples
(P1-Round 1 and PW-Round 3) at low concentrations (1
ug/L) . "
Comment:
"Which PW well?"
Response:
Dibromochloromethane was found in PW-3 during Round 4
of sampling.
Comment RI.8.12:
Text:
"The total PAR concentration for this sample was 2,500
ug/kg. No the soil samples collected at Kirsch Co. Plant
No.1 contained detectable levels of PARs."
Comment:
"This statement is not true.
According to the lab
49
-------
Response:
reports, other soil samples at the Kirsch property
contained detectable levels of PABa."
Comment noted.
RI Section 9 Comment 1 (Comment
RI.9.l):
'!'ext:
Comment:
Response:
Comment
Text:
Comment:
"Preliminary stages of the model were also used in
p=ssen~a~~ons to City officials in an effort to describe
the cur~ant problem and the potential effects of planned
expansion cf the ~uni=ipal well system. Simulations were
conducted to evaluate several alternate well locations.
The city ultimately selected the Tburaton Woods Park as
the location of a new municipal wells."
"In the PS, modeling work indicate. that PW6 vill be
impacted under the no action alternative. Of the
model indicates that, why vas the well placed where it
was."
According to the preliminary model, all locations
within the City showed the potential for
contamination. The selection of the Thurston Woods
Park site vas made by the City of sturgis.
RI.9.2:
"The geology and groundvater flow systea within the City
is described in detail in section 5 of the RI report.
This summary provides a general conceptual model used as
a basis for construction of the groundvater flow model
simulating the Site conditions."
"The conceptual model of the flow system prior to
pWlping appears to be that the modeled area is a
ground-vater recharge area with radial discharge to
surrounding surface water bodies, but very little
information is presented regarding this critical
point. The transient model used a simulated
potentiometric surface based on this non-pumping (pre-
1950s) conceptual model as initial condition. But
there is no figure showing the simulated
potentiometric surface for this condition, nor are the
surface water bodies along the boundaries shown. Even
the report seems to have some doubt about the flow
system prior to pWlping: "In the absence of pumping,
groundwater flow mAl be radially outward from the
City" (p. 4-3, underline added). Yet, on page 9-2,
"regional flow in this lower aquifer is from the
northwest toward the southwest." This statement
50
-------
Response:
implies that flow is D2t radially outward in the lower
aquifer.
Although not stated in the report, it seems that the
steady-state model was not calibrated owing to the
absence of field data prior to 1950.-
The radial flow in the shallow aquifer is regional
(Prairie River to the north, Fawn River to the south
and east, and Minnewaukan Lake to the west). Flow at
the City of Sturgis is D2t radial. If, in the second
paragraph, the commentor is referring to pre-pumping
conditions, the model was not calibrated to pre-
~ping statics. See the response to comment RI.4.2.
Text:
Comment RI.9.3:
Comment:
Response:
"Horizontal groundwater flow occurs principally in the
three outwash deposits due to their relatively high
hydraulic conductivity."
"A hydraulic conductivity of around 40 ft/day is not
especially high for outwash deposits. This value
would typically be representative of clean to silty
sand, not sand and gravel. The field data presented
in Table 5-5 indicate that the field-measured values
are higher than those used in the model. How were the
field data used to deteraine plausible ranges of
hydraulic conductivity values for model calibration."
See response to comment RI.S.1.
Text:
Comment RI.9.4:
Comment:
Response:
"It appears that regional flow in this low.r aquifer is
from the northeast toward the southwest."
"Discussions in Section 4.0 indicated that the pre-
pumping flow directions were radial from the City. We
agree that regional flow is northeast to southwest (or
generally east to west) which is in conflict with the
. previously described radial flow concept."
u.S. EPA agrees that regional flow in the deep aquifer
is generally east to west. See response to comment
RI.9.2 for clarification of the radial flow concept in
the shallow aquifer.
51
-------
Comment
Text:
Comment:
Response:
RI.9.5:
"recharge is estimated to occur between 8 and 12 in/yr."
"The recharge rate used in the .odel was 10.5 in/yr
which is a rate of about 30' of precipitation. This
seems high considering the aaount of paved and covered
surfaces. Why wasn't a range of recharge rates (8 to
12 in/yr) used in the model as part of a sensitivity
analysis? When there are no fie1d-aeasured flux
estimated for use as calibration targets, calibration
can be achl8ved by adjusting recharge and/or hydraulic
ccnduc~ivi~y. It is 1ike~y th~t o~~er ~ombinations or
recharge and hydraulic conductivity could have
provided an acceptable calibration. Using other
hydraulic conductivity values would change the
velocity distribution which would alter the
simulations used in the feasibility study."
u.s. EPA would like to note that the modeling
conducted as part of the RI/FS is for purpose. of the
evaluation of the extent and distribution of
contamination and the development and evaluation of
reaedia1 alternatives. Therefore, it is not
necessarily for the model used to identify and
simulate the universe of possible inputs or seek out
every possible design that could be used to represent
a remedial alternative. Further refinement of the
model in RD/RA may indicate that a different recharge
rate would be appropriate.
Comment RI~9.6:
Text:
Comment:
Response:
"Therefore, the recharge rate from this absorption
ponding is approximately equal to the pumping rate from
the well."
"The pumping rate from F-1 is approximately 225 gpm,
which was also the recharge rate used for the
absorption pond in the ground-water model. The.Kirsch
Plant '2 discharges 65,000 gPd (45 gpm) plus storm
water runoff into the pond. This recharge was not
accounted for in the ground-water model. In addition,
no recharge was input to the model for the Ross pond."
Comment noted. Further refinements of the model may
be made during RD/RA.
52
-------
Text:
Comment RI.9.7:
Comment:
Response:
Comment
Text:
Comment:
Response:
"PATH3D, a three-dimensional advectiv8 particle trackinq
model, was used to help interpret results of the
qroundwater flow model. The proqraa, developed at the
Department of Geoloqy and Geophysics at the University of
Wisconsin (Zhenq, 1989)...." .
"This is not the latest version beinq distributed by
Papadopulos and Associates, Inc. Early versions had
various bugs. The latest version of PATH3D has many
improvements and all known errors have been
corrected."
Comment noted. The latest version of the proqram will
be used for the development of rea8dial desiqn.
R:I.9.8:
Trajectories which are computed by PATH3D show how
particles move under advective processes in relation to
the heads that are computed by the flow model." .
"The advective process does not account for dispersion
or retardation that will prolonq the cleanup of an
aquifer. Recent studies (Kenoyer et al, 1990, HSI
Appendix D) indicate that up to 22 pore volwaes may be
needed to flush contaminants fro. an aquifer.-
Comment noted.
Text:
Comment RI.9.9:
Comment:
Response:
"The model consists of five layers, which simulate the
three principal sand and qravel aquifers (model layers 1,
3, and S) and two confininq units (model layers 2 and
4)."
"No documentation was
or thickness was used
qeometries were input
in RI Section s.o and
Volume 2 of 2.
qiven to indicate what elevation
for these layers. What layer
into the model? See Comment 4
Comments on Fiqure C.4 in FS
Comment noted. Further refinements of the desiqn may
be made durinq RD/RA.
S3
-------
Text:
Comment RI.9.10:
Comment:
Response:
"Where the confining units are shown to pinch out,....
the hydraulic parameters of model layers 2 and 4 are
changed to simulate portions of the sand and gravel
aquifers."
"Where were the windows in the till simulated? No
documentation of this was provided. See Comment 4 in
RI Section 5.0 and Comments on Figure C.4 in FS Volume
2 of 2." .
Maps of window areas can be provided in the final
design.
Text:
Comments RI.9.11:
Comment:
Response:
"No flow boundaries surround the model on all sides.
Surface water features near the model boundaries
(simulated as river nodes) serve aa functional hydraulic
boundaries (see Drawing 12686-22) either receiving water
from the aquifer or discharging water to the aquifer,
depending upon the stage in the river nodes and the head
in the aquifer."
"Assuming no flow boundaries surround the model on all
sides in invalid. This implies that surface water
bodies create no flow boundary conditions for not only
the upper aquifer, but the intermediate and deep as
well. It cannot be envisioned under any scenario how
a small lake, pond, or river will create a no flow
condition on the deep aquifer at depths of 100 to 200
feet."
See response to comment RI.5.1.
Text:
Comment RI.9.12:
Comment:
'"This no-flow boundary appears reasonable in this area,
because it is parallel to what is expected to be a flow
line toward the Fawn River from the topographic high to
the north, and it is a large distance away from stresses
applied to the lower aquifer by municipal and industrial
well pumpage."
"Anticipation of flow toward the Fawn River is
inconsistent with two previous interpretations
provided in the RI report: 1) that flow is radially
toward the City, and 2) flow is from northeast to
southwest. The second interpretation (flow from
54
-------
Response:
northeast to southwest) is correct which is not
parallel to the no flow boundary."
See response to comment RI.S.1. We asaume that the
commentor's item (1) refers to stateaents in the RI
that flow is radially outward fro. the city.
Text:
Comment RI.9.l3:
Comment:
Response:
"The northern half of the eastern and western model
boundaries, and the entire northern model boundary are
separated from pumping in the City of Sturgis by two
chains of lakes, described above. Therefore, the no-flow
boundaries on the northern half of the .odel do not
appear to bave an effect on pumping within the City of
sturgis."
"The chain of lakes are discontinuous; therefore, a no
flow boundary in the shallow aquifer, much less the
deep is invalid (See comment 11). There could very
well be underflow beneath the chain of lakes. Did
heads along the northern boundary change in response
to pumping?"
See response to comment RI.S.l.
Text:
Comment RI.9.l4:
Comment:
Response:
"Aquifer parameters were developed by mapping the
physical limits and thicknesses of each unit and
assigning hydraulic properties to each layer to compute
model input parameters."
"Where is the documentation for this?"
Documentation can be provided as part of the final
design.'
Text:
Comment RI.9.1S:
Comment:
"Vertical resistance between model layers were computed
using the weighted average of the vertical hydraulic
conductivity from one layer to another for each cell of
the model."
"No vertical of horizontal conductivity data were
collected, so how were values established? The
horizontal and vertical hydraulic conductivity of all
modeled units are represented as the same. What is
55
-------
Response:
the basis for this? Because of the depositional
nature of glacial units (especially outwash), the
vertical hydraulic conductivity should be less than
the horizontal."
u.s. EPA agrees that vertical hydraulic gradients
should be less than horizontal hydraulic gradients.
See response to comment RI.5.1.
Text:
Comment RI.9.16:
Comment:
Response:
"For example, in areas with significant head differences
between the shallow and deep aquifers, such as near well
nests W-6 and w-a, the resistance between layers was
raised, if necessary, to aatch ob.erved conditions."
"The difference in head at well W-6 is implied to be a
result of layer resistance, yet this well occurs in
the area .of the foundry pond recharge which could also
create the significant head difference. It 8aY be
that the amounts and timing of recharge fro. the pond
were not simulated correctly."
Well logs indicate a lower permeability material.
Text:
Comment RI.9.17:
Comment:
Response:
"The hydraulic conductivity of layers 1 and 5 were
adjusted during the calibration phase, within the tested
limits, to obtain a match with observed head."
"How were the plausible ranges of hydraulic
conductivity (the "tested limits") established from
the field data?"
See Table 5-5 in the RI. In addition, test results
for specific areas of each aquifer were used where
-needed to achieve model calibration.
Comment RI.9.18:
Text:
Comment:
"The vertical hydraulic conductivity and bed thickness
used (see Table 9-1) is based on experience in similar
environments for obtaining a reasonable head loss across
a l-ft thick bed."
"What was the basis for input data used for the
vertical hydraulic conductivity of the lake/river bed?
How were these parameters determined in the
56
-------
Response:
"experience in similar environaents"? The value used
for river beds is of the saae order ot llagnitude as
sand and gravel. Are these river bed sediments truly
composed of sand and gravel? What is a 'reasonable'
head loss?"
See response to comaent RI.S.1.
Text:
Comment RI.9.19:
Comment:
Response:
"Recharge was modified in the model calibration phase to
arrive at a best fit to observed conditions."
"See Comment 5."
See response to comment RI.9.5.
Text:
Comment RI.9.20:
Comment:
Response:
Comment
Text:
Comment:
"Pumping rates input to the model were developed from
pumping records supplied by the City, Ross Laboratories,
and Sturgis Foundry Corporation for the period January
1977 through December 1988."
"Commonly, both municipal and industrial water use
tends to be non-linear with qreatest use durinq
specific periods of the day (See HSI. Piqure 6). This
was not taken into account in the model. In addition,
Kirsch Plant '1 operated two pwaping wells on their
property (HSI Appendix A). Why weren't these pumping
rates accounted for in the model? What about puaping
from the numerous agricultural irrigation systems in
the area?"
See response to comment RI.S.1. U.S. EPA and MDNR
would appreciate any additional pumping information
Kirsch or Cooper could provide for these wells (in
addition to what was provided as part of the comment
package).
RI.9.21:
"The transient groundwater flow model was simulated
between January 1950 and December 1988 in increments,
referred to as stress periods."
"The transient model used the same boundary conditions
as .the steady state model. The effect of pumping is
to draw water from the boundaries toward the center of
57
-------
the modeled area. It is impossible to assess the
changes in the flow field caused by the pUlipiDCJ since
a contour aap of the pre-pUliping head distribution is
not provided. The boundaries are siaulated as no flow
(impermeable boundaries) with surface water at the
top. This is a peculiar way to siaulate the
boundaries as it requires water to flow down
vertically from the surface watarat the edges of the
model and then flow inward. It is doubtful that this
is what happens in the field. Are the surface water
bodies really acting as fully penetrating no flow
boundaries? Is there any field evidence to suggest
this ~assive diversion ~: surfaC3 watar to the wells?
The water budget and the water budget error for the
transient siaulation should be presented."
Response:
See response to co..ent RI.5.1.
Comment RI.9.22:
Text:
"The first stress period in the model (1950 to 1958) used
the average pumping rates for the City production wells
existing at the time (PW1, PW2 and PW4), and the average
1977 pumping rates for Ross Laboratory wells R1 and R2."
Comment:
"How many time steps were used in each stress period?
Normally, there should be at least five time steps per
stress period."
Response:
This information can be found in the MODFLOW input
files provided to Cooper representatives on August 5,
1991.
Comment RI.9.23:
Text:
"Groundwater levels measured at monitoring wells between
December 1987 and November 1989 were used to test the
calibration of the flow model. Simulated heads were.
compared with observed heads at well nests W1, W2, W6, W8
and W26 for this time period... To further test the
calibration of the flow model, simulated water table and
potentiometric surfaces of December 1988 were compared to
the surfaces developed from heads measured in the field
in January 1989."
Comment:
"The calibration is evaluated by comparison of the
simulated and measured contoured heads in the shallow
and deep aquifers. The comparison appears to be only
fair. Such comparison is, of course, subjective. For
this reason, it should never be used as the sole
58
-------
measure of calibration. Another measure of
calibration offered is to show .easured and si.ulated
hydroqraphs of five monitor wells. However, these
hydroqraphs are flat and tell us little about the
transient nature of the system. Why not show the
simulated hydroqraphs for the aunicipal and industrial
wells shown in Figure 11 For e~~le, a coaparison of
simulated versus observed head levels at TW-84A shows
a difference of 11 feet, indicating very poor
calibration.
In any case, five monitor wells, two of which show
poor comparison with the model, are simply not enough
calibration targets for this model. The heads from
Table 5-2 and gradients from Table 5-3 should be
compared (tabulated) with staulated values. Then the
root .ean squared error of the residuals should be
calculated. A scatterplot of .easured and si.ulated
heads is also a useful way to illustrate the
calibration. Location of the calibration targets with
respect to the nodes in the model should be shown on a
figure.
Response:
See response to comment RI.S.1.
Comment RI.9.24:
Text:
"Although these maps are not identical, the principal
groundwater flow directions and the heads in critical
areas are very similar."
Comment:
"There are significant differences between the
observed (January, 1989) and simulated (December,
1988) water table and potentiometric surfaces.
Water Table
1) The observed water table map indicates a narrow
elongated cone of depression (866 ft. contour) west of
.the Ross wells. The simulated shows a much broader
depression centered over the Ross pumping wells.
2) It appears that the simulated recharge at the Sturgis
Foundry is not located over the absorption pond, but
rather to the north across the railroad tracks. In
addition, the influence of the recharge on local flow
conditions is much greater than in the simulated.
3) What is the source of recharge simulated near well W8
(872 ft. contour). No discussion was given in the
text. No recharge is observed on the water table map.
59
-------
4) Comparison at WJ9 indicates a difference of 7 feet
(867 simulated vs 874 observed).
5) The si.ulated map shows a cone ot depression around
PW4; the observed map shows flow away from PW4.
Potentiometric Surface
1) The major difference is that the siaulated map shows
flow from the northeast and flow from the southwest
convergent on the City's pumping cer:- '~r. By contrast,
the observed map shows flow from no~~east to
south~est. This is alse the conclusion of the ~r.
Th~a is a major and significant difference
demonstrating poor calibration.
2) The siaulated shows distinct con.s of depres8ion
associated with wells PW-4, PW-5, Rl-R4, and RS. The
observed data do not show any depression associated
with PW-5, the slightest hint of one at PW-4, and an
offset depression associated with R5. The bottoa line
is that the pumping wells have little influence on the
horizontal direction of groundwater flow and are,
therefore, less capable of redirecting a plume than is
suggested by the RI report and the ground-water flow
model.-
Response:
See response to comment RI.5.1.
Comment
RI.9.25:
Text:
-The groundwater contaminant plume, as interpreted fro.
analysis of Round 4 groundwater samples collected in
August 1989 (Drawing 12686-6), was used to test the
calibration of the particle tracking model.-
Comment:
-How was this done? What values of effective porosity
were used for each hydrogeologic unit in the particle
tracking simulation? Was calibration achieved by
adjusting effective porosities?-
Because of the errors found in the version of PATH3D
used, this should be addressed during RD/RA using the
latest version.
Response:
Comment
RI.9.26:
Text:
-Once hydraulic properties of the aquifer were
established, the model was started using heads estimated
in a steady state run with the 1950 pumping rates.-
60
-------
Comment:
Response:
How were these
"What starting heads were used?
developed?"
The starting heads for the transient run were the
resulting heads from the steady state run.
Comment RI.9.27:
Text:
Comment:
Response:
"The particles were inserted at the water table at each
location."
"How many particles were used?"
Nineteen (19) particles per five year increment per
source area.
Text:
Comment RI.9.28:
"The flow system is dominated by the well field."
Comment:
Response:
"While this is true based on the simulated heads, this
is not true when compared to actual conditions. See
Comment 24." (Comment 9-24).
See response to comment RI.5.5. No pre-pumping
conditions are available. However, the flow field is
at least influenced by the well field.
Text:
Comment RI.9.29:
Comment:
Response:
"However, recreating the head at well W6S is not critical
to accurately simulating the groundwater flow in the
principal water supply aquifer under the City."
"The recharge at the sturgis Foundry has a tremendous
influence on the shallow flow system as can be
.observed on the water table maps. This influence can
direct contaminants towards windows in the till and,
therefore, ultimately into the deep aquifer."
u.s. EPA agrees, however, if the sturgis Foundry
recharge were not occurring, the direction of the flow
in the shallow aquifer may be the same.
Comment RI.9.30:
Text:
"Contaminant migration simulated by inserting particles
at Kirsch Co. Plant No.1 were shown to migrate toward
61
-------
Comment:
R'::!spor.se:
the center of the City."
"These simulations did not account for the pu.ping
that occurred at the Kirsch property prior to 1981.
During the pre-1981 era, these wells were used and
capable of pumping at rates of 100 to 600 qpa (See HSI
Appendix A). The effect of this pu.ping would retard
any movement toward the center of the city. Note the
locations of the Kirsch wells on cross sections shown
on HSI Plates I and II."
See response to comment RI.5.5.
Text:
Comment RI.9.31:
Comment:
Response:
"The principal migration route was toward City well PW2
during its operation. When City well PW2 was turned off,
the migration route shifted toward the Ross Laboratories
well field." .
"This does not match the observed flow directions in
the deep aquifer. Based on these maps, contaminants'
near PW2 would continue to miqrate west to southwest
along the regional gradient and not towards the Ross
wells." .
See response to comment RI.S.1.
Text:
Comment RI.9.32:
Comment:
Response:
"During periods when city well PW2 was off or pumping at
very low rates (see Fiqure 1 for period 1984 through
1989), the particles originating at Kirsch Co. Plant No.
1 migrated primarily toward Ross Laboratories wells R1
and R2."
"The analytical data show the highest concentrations
occur at Ross well R4, not R1 or R2."
Comment noted.
Text:
Comment RI.9.33:
"The leg of the migration route from Kirsch Co. Plant No.
1 extending toward well PW5 consisted of only a very few
number of particle tracks, indicating that at times when
PW5 was pumping at relatively high rates and the Ross
Laboratories well field was pumping at relatively low
rates, a small amount of contamination may have been
62
-------
induced to move toward City well PWS."
Comment:
"Given the poor calibration of the flow model and the
absence of detectable VOCs in PW-S and W-29D, this
potential migration route is extreaely speculative and
unsubstantiated."
Response:
See response to comment RI.S.1.
Comment RI.9.34:
Text:
"The particles initiated at the Wade Electric facility
mia=a~ed to the nor~h L~rough the Upper aquifer and then
vertically downward in the vicinity of well W32D where no
confining layer was observed."
Comment:
"This does not match observed flow directions which,
in the shallow aquifer, are to the east from the Wade
facility."
Response:
See response to comment RI.S.1.
Comment RI.9.35:
Text:
"After discontinuing use of City well PW2, particles from
the Wade Electric facility migrated toward the Ross
Laboratories well field."
Comment:
"See Comment 31." (Comment RI.9.31 above.)
See response to comment RI.S.1.
Response:
Comment on RI Table S-2
"Water level data for W9S on April 26, 1989 is anomalous compared
to other measurement events. This measurement was used in the
development of the water table map in Drawing 12686-9. The data
for well MW1B on August 14, 1989 is anomalous compared to other
measurement events. .
Where are the water level data for the September, 1987 Round 1 .
sampling event? What does NA stand for in this table?"
ReSDonse:
The comment concerning water level data is noted.
September, 1987 Round 1 water level data is
available in Technical Memo '1 which has been in the
repository at the Sturgis Library since early 1988.
"NAn stands for "not available".
63
-------
Comment on RI Table 5-4
"Which values represent upper, middle, and lower aquifer as
presented in Table 5-5? Only one of these values is froa a
unit (W5DD). Why were no vertical hydraulic conductivities
determined using lab permeability testing?-
till
ResDonse:
The values are shown on the cross sections. Lab
permeability tests were not used because they
typically underestimate bulk permeability of clay.
layers. Values were based on soil type and
calibration of the heads in the upper and lower
aquifer.
Comment on RI Table 9-1
"storage coefficients are not discussed in the text. Table 9-1
indicates that a storage coefficient of 0.3 was used for the
upper aquifer. This is high for a sand and gravel aquifer. An
average S for coarse gravel is .23 and an average S for aediu.
sand is .28. No justification is given for using storage
coefficient equal to 0.005 for the other four layers, including
both aquifers and confining units. The aquifers ought to have
different storage values than the confining units."
ResDonse:
Remedial Design will present an opportunity for
refineaent of the made1. The estimate of .30 is
acceptable considering the fact that well field
storage makes a minimal contribution to the aquifer
because long-term .ade1ing was at a quasi steady-
state.
Comment on RI Drawina 12686-6
"This drawing is highly speculative with respect to the "legs" of
contamination shown. The southeast leg towards PW-4 is based on
one detection of 1 ppb TCE out of four round. of data.
Subsequent sampling by the City indicates that this detection was
a false posi ti ve. The southern leg towards PW-5 is pure
speculation as no VOCs have ever been detected at PW-5. The
southwest leg towards W-39 is not appropriate for two reasons.
First, the data from this well were qualified because TCE was
also found in the associated blank. Second, based on regional
flows presented in Drawings 12686-11, -12, and -13, ground water
impacts at W-39 (if confirmed) are from another unrelated source.
The western leg towards PW-3 is speculative because recent data
indicates that no VOCs are present at this location. The
northern leg towards the Ross well is also in error. First, if
all the contaminants are related and are migrating north due to
Ross pumping, how does the plume migrate past the pumping wells
64
-------
as shown in this drawing? Second, the interpretation totally
ignores the potential for an unrelated source to impact the Ross
wells such as a nearby shallow plume that .oves directly vertical
into the Ross wells."
ResDonse: See response to comment RI.8.8. As for comment
concerning northern leg, dashed lines are a "best
estimate" of the extent of contamination.
Comment on RI Drawina 12686-7
"Sote 2 says water levels were collected on April 26, 1989, and
No~e 3 says .ater levels were collected on November 9 - 14, 1988
and October 20, 1988. It appears that Note 2 should be deleted.
Why were water levels at five wells collected in November, and
the others in october? Given the transient conditions at the
site due to pumping, the data should be collected from the same
time period.
This data suggests a "window" in the till in the vicinity of the
Ross wells (R1-R4) which means a shallow source in this area
could directly impact the Ross wells."
ReSDonse:
The commentor is correct1 note 2 should be deleted.
Water levels for all wells were not sampled at the
same time due to access complications. However, in
a comparison of wells that were collected on both
dates, little difference was seen.
The comment concerning a window in the vicinity of
the Ross wells is noted.
Comment on RI Drawina 12686-11
"Same comments as for Drawing 12686-7. Thirty-three percent of
the data was collected one month earlier than the rest. During
that period, wells PW-4 and PW-5 were pumping 12' more than in
November. The Ross wells could also have been pumping at
different rated. The effect of these differences on the flow
maps could be significant. Why wasn't all the data collected at
the same period."
Response: See above to the response to the comment on Drawing
12686-7.
Comment on RI Drawina 12686-12
"Why was the drawdown around R-5 off center?"
65
-------
ResDonse: The drawdown could be centered on R5, but there was
no data to document such an interpretation.. It is
for this reason that the contours were dashed.
Comments on RI Drawinas 12686-15.
-16. -17. -18. and -19
"Several inconsistencies were noted by comparing the cross-
sections to each other. These inconsistencies are associated
with the piezometric data which are presented differently for the
same well:
I~ Sect~~n AA' at MW-3, the 866 f~. centour is within the
intermediate till, and in section EE' it i8 shown in the upper
till.
.
In section AA' at MW-3, the 864 ft. contour i8 within the
lower part of the intermediate till, and in Section EE', it is
shown in the middle of the intermediate till.
.
In section AA' at W41D, the 870 ft. contour i8 not shown, and
in section FF', it occurs in the upper till.
In Section AA' at W41D, the 868 ft. contour is within the
intermediate till, and in section FF', it is shown in the
upper till.
.
.
In section AA' at W41D, the 864 ft. contour i8 vertical
through the lower outwash, and in section FP', it is oblique
through the lower outwash.
In Section AA' at W32D, the 866 ft. contour i8 approximately
20 feet higher than it is shown in Section PP'.
.
.
In Section BB' at W50D, the 864 ft. contour is within the
middle outwash, and in Section DD', it is shown in the upper
till.
.
In section BB', the 866 and 868 ft. contours were omitted east
of W2D when compared to Section FP'.
In Section DD' at MW1C, the 866 ft. contour is within the
intermediate till, and in Sections EE' and FF', it is shown in
the upper till.
.
.
In section DD' at MW1C, the 864 ft. contour is within the
lower part of the intermediate till, and in Sections EE' and
FF', it is shown in the upper part of the intermediate till.
In sections DD' at MW1C, the 862 ft. contour is at the base of
the lower outwash and shows vertical flow: in Section EE', it
is in the intermediate till, and shows horizontal flow: and in
.
66
-------
Section FF', it is in the middle of the lower outwash and
shows oblique flow.
Several inconsistencies were also noted between the cross-
sections and the corresponding water table map:
.
In Section AA', the 868 ft. contour between MW2C and MW3C is
inconsistent with the water table map.
.
In Section AA', the 872, 870 and 868 ft. contours between W4lD
and W32D are inconsistent with the water table map.
In Section AA', the 874, 872, and 870 ft. contours near TWS4A
should extend up to the water table surface.
.
In Section BB', the 866 and 864 ft. contours are oaitted to
the east of W26D.
.
In Section CC', the 866 ft. contour between W30D and W40D is
inconsistent with the water table map.
In Section DD', the 874 ft. contour at W8D is inconsistent
with the water table map.
.
Except at locations along the south and west portions of the site
(MW-6, W6D, W5D, W39D, TW84A, and W29D), the piezometric data do
not reflect the changes in lithology presented in the cr08S-
sections. For example, the southern part of Section DD' sbows a
steep qradient (stacking of piezometric contours) within the till
as a result of the vertical resistance the till unit creates. In
no other cross-sections do the till units deaonstrate this same
vertical resistance. This data indicates that most of the till
units shown" have little resistance to qround-water flow, and,
therefore, are not significantly different from the outwash with
respect to hydraulic conductivity.
In general, it is believed that the cross-sections and, hence,
subsequent ground-water modeling efforts, over-interpret the
continuity of the till units. As presented in RSI Plates I and
II, the till units are considered to be much more discontinuous
in lateral extent." "
ResDonse: Bu11eted comments are noted. However, commentors
concerns do not affect U.S. EPA's interpretation of
the flow system.
The commentor states that other cross-sections do
not show the same vertical resistance as seen in
Section DD'. However, other locations in the till
demonstrate similar vertical resistance (although
not to the same extent). See Section AA' at MW2C,
W4lD, and T484A. See also Section BB' at W5DD and
67
-------
Section CC' at W29D.
U.S. EPA disagrees with the c0888ntor's stateaent
that the data indicates that .ost of the till units
have little resistance to qround-water flow, and
therefore, are not significantly different than the
outwash with respect to hydraulic conductivity. The
little head difference between the upper and low
aquifer, due to the high peraeability of the outwash
and the close proximity of production wells to the
window, can account for the observed data.
Concerning the HSI submittal, see the response to
comment RI.5.1
Comments on RI Drawina 12686-20
"The contaminant plume is depicted as .oving upward towards well
R-4 which is inconsistent with the piezometric data which
indicates flow is vertically downward at R-4. An alternative
interpretation which was not presented and which is consistent.
with the hydraulics, would be a source near W14S to move directly
into R-4.
In addition, the combination of field GC and CLP data is
inappropriate considering the poor correlation between the two.
Furthermore, using the highest value detected over a two-year
period misrepresents the contaminant plume considering qround-
water velocities on the order of 200 ft/yr..
ResDOnse:
Please see response to
concerning piezometric
and RX.6.5 for comaent
Comment concerning use
co..ent RX.7.18 for co..ent
data. See response RX.6.4
concerning CLP and GC data.
of highest value is noted.
Comments on RI Drawina 12686-21
"The utility of this drawing is uncertain given the poor .
correlation between field GC and CLP data. In addition, the data
presented are from different depths and were collected at
different times during the RIproqram. Nothing is consistent in
this data presentation." .
ResDonse:
Please see responses to comments RI.6.4 and RI.6.5.
Comment concerning utility of drawing is noted.
Comments on RI Drawina 12686-23:
"There are several concerns with this drawing related to the
68
-------
(1)
comparison of the observed water table (Drawing 12686-8).
(2)
(3)
(4)
(5)
The simulated recharge at the sturgis Foundry is
mislocated and is shown north of the railroad tracks.
The absorption pond is actually about 300 feet south of
where it was simulated.
What is the reason for the ground-water mound near W-8?
This does not appear on the observed map, nor is there
any explanation given for it.
The influence of the recharge at the Sturgis Foundry on
the shallow flow system is much greater in ~he observed
than in tne simulated.
The simulated map shows flow towarda PW-., suggesting a
window in the till. The observed show. flow away fro.
PW-4.
There is a 7 ft. difference at W39S, and a 4 ft.
difference at W37."
ReSDOnse: Comments are responded to in the order presented
above.
(4)
(5)
(1)
(2)
Comment noted.
The shallow aquifer is relatively thin and
based on the 109 at W-8, it has a relatively
low permeability. .
Because 01' the shallow flow systeDi through the
upper till unit, the head is very sensitive at
that point. Because of the local nature of the
phenomenon and the high sensitivity of the
head, the use of the head in this area for
calibration purposes should be discounted.
There was little data available in the vicinity
of PW4 to draw the observed water table
surface, so contours are dashed in that area.
No head measurements for W39S on January 1989
were available, so water levels in that area
are not based on observation. However, very
little head difference between simulated and
the observed from October was seen (12686-10).
W37I is an intermediate depth well. The log .
shows a low permeability zone in the water
table. The model represents the head above the
low permeability zone. W37I represents the
head below the low permeability zone.
(3)
69
-------
Comments on RI Drawinq 12686-24
"There are several concerns with this drawing related to a
comparison with the observed piezometric map shown in Drawing
12686-12:
(1)
(2)
(3)
(4)
These two maps do not look alike at all.
The simulated map shows distinct cones of depression
associated ~ith the pumping wells. The observed map
shows very little deflection in the regional flow as a
result of pumping.
The simulated map shows flow convergent toward the center
of the City (flow from northeast and.fro. southwest).
The observed aap shows a regional flow from northeast to
southwest. .
There is a difference of 11 feet at TW84A."
ResDonse: Comments are responded to in the order presented
above:
(1 and 2)
The simulated map is for continuous
pumping at lower rates than the actual
pumping rate at any 1 point in time, In
addi tion, the contour surface fr08 the
model provides more information about the
contours than could be obtained from the
available actual observations.
Co_ent noted. This is an area where
further refineaent of the model is needed
durinq RD.
TW84A is very close to PW5. The head .at
TW84A is very sensitive to the actual
pumpinq rate at PW5 at the instant of
measurement.
(3)
(4)
Comments on RI Drawina 12686-25
"Concerns with this drawing include:
(1)
(2)
The miqration 1eq to the south towards pw-s is
inconsistent with the observed flow direction in Drawing
12686-12 which again demonstrates the poor correlation
between observed and simulated conditions.
The particle tracking shows the axis of the plume moving
towards R-2, while the observed is shown moving towards
R-4 in Drawing 12686-27.
70
-------
(3)
(4)
(5)
The particle tracking from Wade Electric moveS,northeast
in the shallow ground water, and then northwest in the
deep ground water. By contrast, the observed ground-
water flow from Wade is east to southeast in the shallow
aquifer, and then west in the deep aquifer.
The particle tracking does not show leqs towards PW-4,
W39D, or PW-3 as indicated in Drawing 12686-27.
The simulated migration route is not shown for the
Telemark site, and none was performed for the source at
W14S."
Fespon~~ Co~ents are responded to in the order presented
above:
(4)
(5)
(1)
The flow model siaulated variations in time,
whereas the observed is an interpretation of
measurements at a point in time.
Small scale variations in permeability (that
were not simulated) may result in small
variations in flow adjacent to production
wells.
Refer to the water levels at W27S and W32S.
They suggest that there may be flow to the
north and east. Particles are under the
influence of the shallow aquifer until they
pass through the window into the deep aquifer,
where they subsequently are influenced by the
deep aquifer flow. The observed head in the
shallow system suggests the potential for flow
to the north and east at certain times.
The frequency and nU8ber of particles released
in the model are insufficient to have the
resolution to model the low concentrations
observed at PW3 (approx 5 ppb) versus the high
levels observed in other locations throughout
the plume. The observed leq at PW4 may not be
real. Additional monitoring will be necessary
to verify the presence, or lack of,
contamination.
The simulated migration route from Telemark is
shown on the drawing. No source has been
identified at W14S.
(2)
(3)
"( 1)
Comments on RI Drawina 12686-26
Using the maximum detected value in four rounds of
samples collected over a two~year time span misrepresents
the plume and does not account for migration that occurs.
71
-------
.(2)
(3)
(4)
(5)
(6)
A shallow source is depicted at Telemark and at the
Airport (W14S), yet these sources are dismissed or not
even discussed in the report.
The 10 ppb contour around Wade Electric suggests
migration is to the west, yet the shallow ground-water
flow is east to southeast. How do you explain this
discrepancy?
In addition, why does the 20 ppb detection at A-1 fall
be~ween the 1 and 10 ppb ~ontour? ~his data suggests
another source considering flow direction.
The contours around the sturgis Foundry pond should be
closed and separate from the Wade Electric pluae.
The map shows 7.7 ~g/l at W-37I, whereas the tabulated
data in Appendix E show no detectable compounds with a
detection limit of 77.0 ~g/l for this event. Regardless,
the maximua detected value during the four sampling
rounds was 16.0 ~g/l.
An alternative isoconcentration map reflecting these
comments is presented on HSI Plate VIII. To make the
comparison equivalent, the maximum value recorded during
the four sampling rounds was used, although we di~agree
with this procedure."
ReSDonse: Co_ents are responded to in the order presented
above.
(7)
(1)
(2)
(3)
(4)
(5)
(6)
(7)
Co_ent noted. The method used provides a two-
year SUDDDary of where the plume has been and
IlaY not represent anyone point in time.
Contamination from Telemark is discussed in the
RI. No source has been identified at W14S.
The contours were developed from an
interpolation of contamination at both Wade and
the sturgis Foundry.
Co_ent noted.
This well may be a multi-aquifer well and was
therefore given little weight in the
development of contours.
Comment noted.
See response to comment RI.S.1.
Comments on RI Drawing 12686-27
" (1)
The concept of presenting the maximum detected value over
a two-year period misrepresents the actual ground-water
quality and does not account for plume migration over the
72
-------
(2)
sampling periods.
In addition, the comment on Drawing 12686-6 related to
the "legs" of contaminant plumes also apply to this
drawing.
(3)
An alternative isoconcentration map reflecting these
comments and those on Drawing 12686-6 is present on RSI
Plate IX. To make the comparison equivalent, the maximum
value recorded during the four sampling rounds was used,
al though we disagree with this.. procedure. "
Res?on~ Comments are responded to in the order presented
above.
(1)
(2)
(3)
Comment noted.
See response to comment. on Drawing 12868-6.
Submittal of HSI Plate IX i. noted.
Comment on RI Drawina 70084-88
"The soil gas data do not correlate with the CLP soil data and
greatly exaggerate the extent and magnitude of impacts. An
extreme example of non-correlation is the consistent soil gas
detections with depth at S8-02 compared to consistent CLP no
detections with depth at the same location (compare to Drawing
70084-811 and 70084-812). See the CLP/Soil Gas Comparison in
attached Fiqures 1 and 2."
Re~Donse: See responses to comments RI.6.4 and RI.7.11.
Comment on RI Drawina 70084-89
"Same comments as for Drawing 70084-88. Note the consistent soil
gas detections with depth at W17S, W19S, S809. The.CLP data at
these locations (Drawings 70084-815 and 70084-816) show no
detections at any depth."
ResDonse: See responses to comments RI.6.4 and RI.7.11.
Comment on RI Drawina 70084-810
"Same comments as for Drawings 70084-88 and 70084-89. Note
consistent soi1 gas detections with depth at S814. The CLP data
at this location (Drawing 70084-818) shows no detection at any
depth."
ResDonse: See responses to comments RI.6.4 and RI.7.11.
73
-------
Comments on RI Drawina 70084-B11
"The CLP data indicate 770 ppb of PCB at a depth of 2.5 feet at
SB06, yet soil sample adjacent to this location (8B23 and 8B24)
show no detectable PCB at a depth of .5 feet. Either the data
from SB06 is representative of an extremely saa11 area, or ther.
is poor reproducibility of results.
Similarly, the CLP data indicate 260,000 ppb of PCB at a depth of
1 foct at borehole W11S. Thirteen days later, a surface sample
collc:cted at Wl:S at the same depth showed 950 ppb of PCB. This
data indicates ?oor rs?roducibility of soil analytical results o~
.~ ~remendous amount of vola~ilization occu-~ed during the 13 days
between the two sampling events."
ReSDonse: See response to co..ent RI.8.'.
Comment on RI Drawina 70084-B12
"The same comments on Drawing 70084-B11 apply to this drawing.
At SB06, the CLP data indicate 27,000 TeE at a depth of 2.5 feet
while adjacent surface soil samples at SB23 and SB22 at a depth
of .5 feet show no detectable TCE. In addition, the CLP data
indicated 8,200 ppb of TCE at a depth of 1 feet at borehole W118
while the su~face sample from a depth of 1 feet shows 21 ppb of
TCE. What is the reason for these significant differences in the
CLP data?"
Response: See response to comment RI.8.6.
Comments on RI Drawina 70084-B14
"Compared to the tabulated data in Appendix E, a couple of errors
were noted in the data presented on this drawing. The drawing
shows 144 ~g/kg of PAR at a depth of 2 feet at SB19; the Appendix
indicates no detectable PAR. The drawing shows no detectable PAR
at a depth of 4 feet at SB20, and the Appendix indicates 220
~g/kg of PAR. The total PAR at W42S is 2,446 ~/kg, not 2,250
~g/kg as noted in the drawing.
In addition, see Comment 5 in Section 7.0 concerning possible
sources of the PAR found on the Kirsch property."
Response: Comment noted.
RI.7.5.
See also the response to comment
Comments on RI Drawina 70084-B26
"Using the maximum value detected over a two-year period
74
-------
misrepresents the magnitude and extent of impacted ground water.
This is especially true for wells W34S and W23S where subsequent
sampling at the wells indicated no detectable VOCs. Why is this
such a difference between sampling events. The report states it
is because the wells were replaced. How would this affect the
results? Are other wells. suspect also then because of well
installation procedures?
At well W42S, a value of 6.500 ~g/l is shown and used to develop
the 1.000 ~g/l contour. The data validation portion of the
report qualified this data as unusable because a duplicate sample
showed only 3 ~g/l. Why was this data used when it was qualified
as unusable?"
Res90nse:
As discussed in an earlier comment, the maximum
value detected was used to provide a two-year
summary of where the pluae has been and it aay not
represent anyone point in tiae. u.s. EPA will not
speculate on the variation in the sample results.
The data from W24S was used to develop the contour
because it was. consistent with the field analysis of
soils taken from that location. .
Comment on RI Drawina 70084-B27
"This contour map is inconsistent with the
12686-26. The later drawing shows the 100
extending far to the west. Why isn't this
presented?"
one shown in Drawing
and 10 ppb contours
data consistently
Re.oonse:
Drawing 70084-B27 was developed in an attempt to
show Wade contamination independent of the influence
of the sturgis Foundry. .
Comment on RI Drawina 70084-B28
"Why isn't the data contoured such as shown in Drawing 12686-26?"
Resoonse: Comment noted.
Comments on RI Volume 5 of 5. Aooendix H - Groundwater Flow Model
~
(1)
"This appendix includes "Input Data Modules" and output
for the "Steady-State Simulation of Calibrated Transient
Groundwater Flow Model". Is this the pre-pumping steady-
state simulation used for the initial conditions of the
transient model, or the steady-state simulation with
75
-------
(2)
pumping used with the particle tracking siaulationa in
the FS? .
The input data modules include pWlping wells, as does the
listing in the output record; yet, the water balance does
not show discharge fro. pumping wells. There is a .
significant (19') error in the water budget for this
simulation. MODFLOW runs usually have water balance
errors of less than 1', suggesting that something is
wrong with the way the simulation was executed."
ResDonse: Comments are responded to in the order
presented above.
(1)
In Appendix Hit states that it is a
steady-state ai8ulation using period 50
pWlping rat.s (ending Deceaber 1988).
(2)
The apparent error occurred because the
production wells were not considered in
the water budget. When the water budget
is corrected for the production wells, the
error falls to approximately 1'.
Text:
FS Section 1. Comment 1 lFS.1.1)
Comment:
Response:
"The RI identified two sources areas responsible for the
aquifer conta8ination: the Kirsch Co. Plant No.1
property (Kirsch) and the fOrJIer Wade Electric property
(Wade).
"Additional source areas were identified but dis.issed
in the RI report such as the Telemark and airport
properties. "
Additional source areas investigated during the RI
were not found to be impacting the well field.
However, the U.S. EPA would be interested in reviewing
any information the commentor may have concerning
other possible sources.
Text:
Comment FS.1.2
"The primary source of this contamination is from the
Kirsch property. Drawings 12686-26 and 12686-27 show the
total ethene concentrations in the shallow and deep
aquifers, respectively. These drawings show that
contamination from Kirsch enters the shallow aquifer,
follows the groundwater flow to the southwest, and flows
through a window in the till units to the deep aquifer.
76
-------
Comment:
Response:
Contamination in the deep aquifer is pulled towards
active pumping wells. Contamination bas been detected in
the public water supply wells PW1, PW2, and PW4, the
sturgis Foundry well (F-1), and Ross wells R1, R2, and
R4."
"The observed hydraulic data presented in the RI
report do not support the concept of flow fro. Kirsch
to the Ross wells. The RI report also did not
consider an alternative path from a shallow nearby
source directly impacting the Ross wells. At least
two other potential sources fit this scenario
(Telemarx and the airpo~).
The RI data do not support the concept that Kirsch is
the primary source of impact at PWl, PW2, and FI. The
Wade Electric site could be the priaary potential
source for those impacts. Furthermore, the
contamination of 1 ppb of TCE observed in PW-4 has
never been duplicated by subsequent Department of
Health sampling."
u.s. EPA disagrees; the data clearly supports movement
in the deep aquifer to the Ross wells. See responses
to comments RI.7.16 and RI.7.17.
The comment concerning which area is primary source
fOf PWl, PW2 and F1 is noted. Comment concerning PW4
is noted.
Text:
FS Section 2. Comment 1 CFS.2.1\
Comment:
"Considering the overall long-term goals of protecting
public health and the environment, and the. primary Site-
specific goals of remediating the water supply aquifer
and contaminated soils, a number of remedial action
objectives were developed ....-
"We agree with the goals of "remediating the water
supply aquifer", but are very concerned with the lack
of specific inclusion of primary site specific goals
to assure that the City of Sturgis can deliver a safe,
reliable water supply to the citizens of the City of
Sturgis. We believe that the overriding goal and
remedial action objective (RAO) of any action taken to
mitigate the environmental impacts associated with the
"site" must be the delivery of a safe, reliable water
supply for the City, not just to remediate the
aquifer.
. RAO Numbers 1,2, and 6 are not Remedial Action
77
-------
Response:
Objectives, but rather, requir888nts of any reaedy
selected. The NCP requires that a reaedy ..et
ARARs (and meeting MCLs for drinking water is an
ARAR) and that ~e e~tablis~gd risk standards for
EPA acceptance 1S 10 4to 10 .
. The term "exposure" in RAO NUllbers 4 and 5 is too
vague. Is this dermal exposure, inhalation
exposure, or infiltration exposure? The objective
for reduction of exposure should be well defined and
not of a general nature.
. Again, there is no ~ention of ~aintaininq a safe,
reliable drinking water supply for the City of
Sturgis as a RAO."
The City of Sturgis currently has a safe water supply.
The goal of providing a safe water supply is contained
in goal '5, which is in fact more conservative than
that proposed by the co_entor. U. S. EPA is as
concerned with the health of private well users as we
are of municipal supply users. Other comments are
noted.
Text:
Comment FS.2.2
Comment:
Response:
"The effectiveness of biological treataent of TCE/PCB
contaminated soils has not been sufficiently deaonstrated
to deteraine whether it would be effective."
"Why was no consideration given to biodegradation of
PARs instead of excavation and disposal? See
discussion of biodegradation of PARs in the Risk
Assessment comments attached separately and prepared
by Gradient Corporation."
Biological treatment of PARs was an option discussed
internally by agency personnel. The technology was
not carried into the FS for several reasons. The PAR
contamination is unevenly distributed across the Site
and is primarily in the shallow soils. Treatment
would likely require construction of a bioreactor.
Biological treatment of PARs is still in the
developmental stage and would require significant
time, effort and expense to demonstrate a workable
technology before full-scale remediation could begin.
78
-------
FS Section 3. Comment 1 (PS.3.1}
Text:
Comment:
Response:
"The discharge options for treated groundwater considered
in this alternative are limited to City storm sewers
discharging to the Nye Drain, and to groundwater via
injection wells."
"Based on discussions with the sturgis city engineers,
the existing storm sewers can not handle the proposed
hydraulic load on a permanent basis. New sewer
installation or upgrade of the existing system would
be required to provide for increased capacity. Why
wasn1t discharge to an absorption pond considered?"
The use of an absorption pond was discussed internally
by agency personnel, but was not carried forward into
the PS for aany reasons, including: the large volWle
of water requiring treatment and land acquisition
concerns. In addition, warzyn's review of the city
storm sewer determined that a regional upgrade would
not be needed; however, it is possible that portions
of the system would require improvements.
Comment FS.3.2
Text:
Comment:
Response:
"G4A Aquifer restoration by extraction wells located in
the contaminated zone, with disposal of treated extracted
water via storm sewer to the Nye Drain."
"From a resource conservation standpoint, why wasn't
use of water for potable drinking water considered as
opposed to discharging treated water to waste?"
The FS did not consider use of treated water.
However, the Proposed Plan and this ROD clearly state
that the municipal use of treated water should be
investigated further.
Comment FS.3.3
Text:
Comment:
"Non-volatile contaminants would be managed by
containment that would reduce the risk due to direct
contact with the soils and reduce the infiltration
through the soil that may carry contaminants to the
groundwater."
"The non-volatile PABs are relatively immobile at low
concentration levels, have only been found in soils
less than 6 feet deep, and have not been detected in
ground water; therefore, containment to reduce
79
-------
Response:
Comment
Text:
Comment:
Response:
infiltration is not appropriate..
Although PARs are relatively imaobile, containaent is
necessary to reduce the direct contact threat.
FS.3.4
"Wells installed in the unsaturated zone would be
connected to a vacuum header system..
"Why were horizontal well~ not considered given the
large area being suggested for remediation? OVer this
size of an area, horizonta~ wells re more cost-
effective than the vertical wells..
See response to comment PS.2.2. Tbe use of horizontal
wells instead of vertical wells. is a design issue that
can be addressed in RD/RA.
Text:
FS Section 4. Comment 1 (FS.4.1)
Comment:
Response:
Comment
Text:
Comment:
Response:
"Type C criteria cannot be determined at this tiae
because it would require a risk assessment following the
requirements established in the Act 307 rules..
.Why wasn't the risk assessment performed following
the requirements established in the Act 307 rules?
In the Proposed Plan, the U. S. EPA and MOHR proposed a
Type B cleanup, DQt a Type C cleanup, and thus, are
not required to conduct a risk assess.ent to justify a
Type C cleanup.
FS.4.2
"This time estimate is based on the advective velocity of
the groundwater and does not consider the effects of
dispersion/aquifer heterogeneities and aquifer adsorption
(retardation) of the contaminants."
"Because of these factors, the time estimates qiven in
this PS serve no useful function and grossly
underestimate the time required to achieve aquifer
clean-up. Why weren't these factors considered?
The model used in the RI/FS was developed to assist in
understandinq and evaluating groundwater flow and to
assist in the development of remedial alternatives.
Dispersion and retardation are certainly important
80
-------
factors to consider during design. However.. U.S. EPA
and MDNR generally do not estimate their effects
because the tools to do so are still in the
developmental stage.
Text:
Comment FS.4.3
~omment:
Response:
Comment
Text:
Comment:
Response:
"It was assumed that future conditions would
slightly (20'> increased withdrawal rate tor
supply, which would be drawn from wells PW4,
illustrated on drawing 70084-834."
include a
the City
5, and 6, as
"What is the basis for ~~is assunption?
agree with this increase?"
Does the city
The increased future withdrawal rate is an estimate to
account for future growth. This can be further
refined during Remedial Design.
FS.4.4
"As evaluated and described in Appendix C, this system
consists of 7 extraction wells, shown on drawing 70084-
837."
"Why weren't fewer wells pumpinq at hiqher capacity
used to provide a more cost effective alternative?"
See response to F8.2.2. This is a design
consideration best addressed in RD/RA.
Text:
Comment FS.4.5
Comment:
Response:
"Well PW6 is adequately protected from lower aquifer
contamination by Alternative G3a, but could potentially
be subject to shallow aquifer contamination from the
Kirsch Co. source area."
"The RI report stated that the ground-water model was
used to evaluate an appropriate location for PW~6
which was installed in 1989. Why was this location
selected if the modeling indicated the potential for
impact? Furthermore, what is the potential for PW-6
to become impacted from the Deer Run Landfill located
to the east of the well. Impacts have been documented
at this source."
See response to comment RI.9.1.
81
-------
Text:
Comment FS.4.6
Comment:
Response:
"Alternative G3b - Plume Management with Re-injection, by
using groundwater reinjection, provide. aore substantial
hydraulic isolation of PW6 from the Kirsch source."
"Based on the modeling results presented in appendix
C, we disagree that re-injection provides more
substantial hydraulic isolation of PW6. The only.
thing that reinjection adds is cost and the potential
to spread contaminants away from ~e capture zones of
the remedial extraction wells." .
Comment noted.
Text:
Comment FS.4.7
Comment:
Response:
"The advective flushing ti.e for the existing groundwater
contamination is between 5 and 15 years."
"Why is the advective flushing time for plume
management (G3) shorter than aquifer restoration
(G4)?"
The advective flushing time does not include. the
effects of dispersion and retardation. Flow paths for
GW4 are more direct than those of GW3, so it is
anticipated that the ti.e to reach ARARSwould, in
fact, be shorter with GW4 rather than GW3.
Text:
Comment FS.4.8
Comment:
Response:
"The supply well at Kirsch Co. Plant No.1 has not been
operated recently. Repair of this well may be needed to
make it operational."
"What is the Kirsch well going to be used for? This
was not discussed under the alternative description."
There is no planned use for the Kirsch well.
Text:
Comment FS.4.9
"4.2.3.8 Cost"
Comment:
"Several discrepancies exist between the costs
presented here and those presented in the tables in
Appendix D. Capital costs for Alternatives G3b with
air stripping should be $6,060,000, not $5,850,000 and
82
-------
Response:
$12,100,000. O&M costs should be $503,000 and not
$470,000 for the air stripper option. NPW costs for
air stripping should be $9,900,000 and $11,400,000 for
Alternatives G3a and G3b, respectively, and not
$9,560,000 and $10,900,000."
Comment noted. Cost details in Appendix D were
revised without updating text.
Tex~:
Comment FS.4.10
Comment:
Response:
"Alternative G4a - Aquifer Restoration Without Re-
i~jecticn consists of a total of S6ven extrac~icn wells,
ea=h provided with a wellhead trea~ent system (Drawinq
70084-B39). Disposal of treated water is to the City
storm sewer system, which discharges to the Nye Drain
(Drawinq 70084-B34)."
"Why not use fewer wells pumping at higher rated to
accomplish the same objective (e.g. two at 800 qpm
each vs seven at 225 qpm each)? Why not centralize
the treatment to reduce costs instead of using well
head treatment? Why not use the treated water for
potable water supply, thus minimizing the pumpage of
and competition with PW-4, PW-5, and PW-6?"
The FS clearly states that possible centralization of
the groundwater treatment system should be a design
consideration (p. 4-20/GW3, p. 4-28/GW4).
Text:
Comment FS.4.11
Comment:
Response:
"Alternative G4a - Aquifer Restoration Without Re-
injection and G4b - Aquifer Restoration With Re-injection
can comply with the Act 307 rules for Type A or B clean-
up levels." "
"Any of "the alternatives can comply with the clean-up
levels given enouqh time. The advective flushing
times used in the FS are unrealistic and misleading in
terms of meeting these objectives." .
It is doubtful that Alternative GWl could meet ARARs
other than throuqh lenqthy natural attenuation. GW2
could potentially reach ARARs for part of the plume,
but not the entire aquifer.
83
-------
Text:
Comment FS.4.12
Comment:
Response:
"The levels of contamination reachinq the plume
manaqement wells could exceed the TCLP limits."
"Should be "aquifer restoration wells" and not "plume
manaqeaent wells".
Comment noted.
Text:
Commant FS.~
Comment:
Response:
"Alternative G4b - Aquifer Restoration With
by usinq qroundwater re-injection, provides
substantial hydraulic isolation of PW6 fro.
source."
Re-injection,
aore
the Kirsch
"More substantial
demonstrated with
adds costs to the
benefit and could
hydraulic isolation is not
re-injection wells. Reinjection
alternative without recoqnizinq any
spread contaminants."
Comment noted. Reinjection was not recommended by
u.S. EPA and MDNR in the Proposed Plan.
Text:
Comment FS.4.14
Comment:
Response:
"Should Ross wells Rl to R4 shut down (or if discharqe is
siqnificantly reduced), the extraction wells would
provide inco.plete protection to wells PW4 and PWs.
These reliability issues would be addressed in final
desiqn."
"This reliability issue would not be a concern if the
extraction wells were used for providinq treated
potable water. The pumpinq on PW-4 and pw-s would be
minimized or eliminated, thus enhancinq the recovery
of contaminants by the extraction wells. PW-4, PW-S,
and PW-6 could be kept in standby and used for backup
(redundancy) in the event of problems with the
extraction system."
u.s. EPA aqrees (and has stated in the Proposed Plan)
that municipal use of treated water is an option and
should be considered. This is a desiqn issue better
addressed in RD/RA.
84
-------
Text:
Comment FS.4.15
Comment:
Response:
The advective flushing time for the existing
contamination in the aquifer for Alternative G4 - Aquifer
Restoration is estimated to be between 10 and 20 years.
Why is the clean-up time for aquifer restoration
longer than plume management?"
See response to comment FS.4.7.
Text:
Comment FS.4.16
Comment:
Response:
"The estimate of advective flushing tiae is probably an
underestimate of actual aquifer clean-up time."
"The word "probably" should be removed. Actual clean-
up times will be longer than those predicted by
advective flushing (See HSI Appendix D)."
Comment noted.
Text:
Comment FS.4.17
-4.2.4.8
Cost"
Comment:
Response:
"Chemical oxidation for Alternative G4b should be
$13,760,000 and not $13,800.00 (See Table 09). O'M
costs should be $598,000 and not $565,000 for the air
stripper (See Table D7). NPW costs for Alternative
G4a with air stripping should be $11,500,000 and not
$11,200,000. NPW costs for G3a with air stripping
should be $14,700,000 and not $14,400,000 (See Table
D7)."
Comment noted. Cost details in Appendix D were
revised without updating text.
Text:
Comment FS.4.18
Comment:
Response:
"To avoid short circuiting of soil gas near the ground
surface and ineffective contaminant removal in the deep
unsaturated zone, the wells should be screened in the
shallow soils."
"The reported highest VOC impacts were in the shallow
soil zone excluded from possible treatment."
This is a design issue.
The SVE system can be
85
-------
designed so as to optimize VOC removal.
Text:
Comment FS.4.1~
Comment:
"Total ethenes in soil, as determined by field Ge, is
shown in Drawing 70084-88. The TCE and PeE soil
concentrations, by CLP analysis, are shown in Drawings
70084-812 and 70084-811, respectively. Soil
contamination has resulted in groundwater contamination,
as shown in Drawing 70084-B2S. BaSed on the field GC
cc~=entrations and analysis ~h~wn in Appendix G,
approximately 220,000 cy of soil is contaminated with
about 600 pounds of TCE at the Kirsch property. About 50
percent of this mass is from the isolated heavily
contaminated soils."
"We take exception and disagree with the analysis
provided in Appendix G. The calculation of
contaminant concentrations was based solely on the
field GC analysis and not on validated CLP data. As
shown on HSI Figures 1 and 2, there is very poor
correlation between the total ethene values derived
from field GC analysis and CLP data for samples from
the same sampling location (boring and depth). If any
correlation can be made between the field Ge and CLP
data, the trend would indicate that the field Ge
indicates total ethene concentrations an order of
magnitude greater than CLP data of the same soil
sample. The use of the field Ge data is not valid and
cannot be justified or defended with properly
validated CLP data. If the 8aa. analysis perforaed in
Appendix G were conducted on acceptable, validated CLP
data, the calculated concentration of TCE in the soils
at the Kirsch site would be on the order of 80 POunds.
Because of the lack of correlation between the field
GC data and validated CLP data for the same soils
samples, we do not believe that the use of the field
GC data is appropriate for establishment of cleanup
volumes, contaminant extent, or other critical
analysis.
Further analysis of the CLP data indicate a number of
discrepancies. In particular, the high TCE/PCE
concentrations found in the CLP data for 58-6 are
significantly higher than concentrations found in 58-6
are significantly higher than concentrations found in
S8-22 and S8-23, both of which were located within 10
feet of the S8-6 location. The report CLP
concentrations are:
86
-------
Borina TCE (DDb1 PCE (DDb1
88-6 27,000 @ 2.5' 770 @ 2.5'
88-22 ND @ 0.5' ND @ 0.5'
88-23 ND @ 0.5' ND @ 0.5'
"The lack of any detection 'in two .amples taken within
10 feet of the reported soil concentration as high as
27,000 ppb TCE raises question of the validity of the
high reported concentrations at 88-6"
Of the 61 samples taken at the Kirsch site and tested
by the CLP, only 23 samples had TCE detections in
excess of the 60 ppb concentration level re~~ired by
MDNR Act 307 Type B ,cleanup standards. Of these 23
samples, only 10 had concentrations which were .ore
than double the 60 ppb standards. The saaple. with
more than 120 ppb were distributed as follows:
The 2.5' sample at W-428
The 2.5' and 5' samples at 88-6
The l' sample at W-11D
The 2' sample at 88-07
The 45' sample at @-428
The 30, 40' and 50' samples at 88-6
The 30' sample at 88-18
The 30' sample at 88-20
As shown above, half (5) of the high concentrations of
TCE were found at the surface of four locations, all
on the western half of the Kirsch property, and half
,(5) of the high concentrations were found at or just
above the water table at four locations. This
indicates that the soil contamination problem is: 1)
isolated at shallow depth in a few locations, and 2)
there is a deep soil contamination problem that
appears to be the result of organic vapors emanating
from the shallow ground water. The shallow ground
water has organic concentrations beneath the site
which generally range between 100 to 10,000 ppb and,
thus, can easily serve as the primary source for'
organic vapor contamination of the soil.
Because of the significant questions regarding the
validity of the data presented and the interpretation
of that data with respect to soil contamination at the
Kirsch site, it is our position that the data
developed to date provides considerable evidence that
most of the soil contamination at the Kirsch site is
at or below concentrations requiring remediation;
therefore, the extensive soil vapor extraction system
proposed in the FS are not technically justified or
87
-------
warranted. In addition, where high concentrations of
contaainants were identified, they were found in very
isolated areas (less than 20 feet in diaaeter) and the
validity of the high concentrations cannot be
confiraed from adjacent or duplicate saaplinq.
Therefore, it is strongly reco...nded that the
decision to be made at this tiae with respect to the
Kirsch site soils would be to:
. Provide for continued institutional controls such as
the fencing and site access control already in place
at the site, and deed restrictions,
. Conduct addition investigation to verify the nature
and extent of the "hot spots" of VOC contaaination
of the surficial soils,
. If "hot spots" are verified at the site, take
appropriate contingent action, such as excavation
and treatment or in-situ treatment of the "hot
spots" to an acceptable cleanup level,
. Implement a monitoring program to assure that
ambient air quality is not impaired by the transfer
of organic vapor from the ground water into the soil
and subsequently into the atmosphere.
Cooper Industries would support this approach and
would vOluntarily conduct the necessary investigations
required to confira or eliainate the question raised
with respect the soils at the Kirsch site."
Response:
u.s. EPA disagrees. We feel that there is sufficient
information to justify active remediation of the
contamination in the soils at the identified source
areas. If deemed necessary by the party conducting
Remedial Design, additional soil sampling may be
considered as part of the predesign phase of the RD/RA
to better estimate the performance requirements of the
SVE system needed.
Comment F8.4.20
Text:
"Ethene concentrations at 5B06, 5B07, W115, and W425 are
much higher in the surficial soils compared to the
underlying soils."
Comment:
"5ee Comment 18. If high VOC concentrations are
confirmed in the shallow soil, they should be
remediated. The proposed VES system would not
accomplish this."
88
-------
Response:
Comment noted. A1thouqh remediation of shallow soil
contamination will not as easily be obtained as deeper
contamination, it can still be accoap1ished. One
option to consider for Remedial Design is the use of a
temporary synthetic cap to provide a better vacuum.
Text:
FS Section 5. Comment 1 (FS.5.1)
Comment:
Response:
"The G4 - Aquifer Restoration alternative is particularly
effective when combined with injection of the treated
water."
"The model does not show much difference between the
option ~ith injec~ion versus the option without
injection. We do not believe injection adds anythinq
but cost to the proqram."
Comment noted. Reinjection was not recommended in the .
Proposed Plan and has not been selected in this ROD.
Text:
Comment FS.5.2
Comment:
Response:
"Alternative G3 and G4 can meet the identified ARARs
described in Table A3 of Appendix A."
"See comment 11 in Section 4.0."
See response to comment FS.4.11.
Text:
Comment FS.5.3
Comment:
Response:
"The G4 - Aquifer Restoration Alternative would provide
the quickest reduction in residual risk."
"The advective f1ushinq results indicate the plume
manaqement options would achieve clean up faster than
aquifer.restoration by five to ten years. Explain
this contradiction." .
The preliminary design concepts presented in the FS do
not represent in any way a final remedial design. By
adjustinq placement of extraction wells, alternative
GW4 will provide for the most effective and most
efficient remediation possible. A1thouqh it seems
that Alternative GW3 will remediate the aquifer faster
than Alternative GW4, it should be known that
Alternative GW4 will provide for a more direct flow
path and thus will experience less retardation than
Alternative GW3.
89
-------
Text:
Comment FS.5.4
Comment:
Response:
"The Plume Management alternative is not as effective
because more of the aquifer would become contaminated
before contaminants were removed."
"See Comment 3." (Comment FS.5.3).
See response to comment FS.5.3.
Text:
C~~€nt FS.5.5
Comment:
Response:
"The G4 - Aquifer Restoration alternative provides the
most direct and quickest reduction by hydraulically
containing the plume and treating the .ost contaainated
groundwater." ,
"See Comment 3." (Comment FS.5.3).
See response to comment FS.5.3.
Text:
Comment FS.5.6
Comment:
Response:
"The G4 - Aquifer Restoration alternative provides the
most aggressive action and would .ake the quickest
progress towards the criteria. The less aggressive
actions of the G3 - Plume Management alternative would
result in slower progress towards .eeting the criteria."
"See Comment 3."
See response to comment FS.5.3.
Text:
Comment FS.5.7
Comment:
Response:
"Alternative G2 - Public Water Supply Treatment may not
be feasible under Act 307."
"How does Act 307 preclude the treatment of a public
water supply? It is already being performed in
Kalamazoo and Clare, Michigan (See KSI Appendix F)."
The two examples cited by the commentor are sites
where the current treatment systems are either
operable units or interim responses until more
extensive remediation systems can be implemented.
Treatment of the water supply, in lieu of active
remediation of groundwater contamination, is clearly
not preferred when an aquifer is potentially useful
90
-------
for municipal or private well use.
Comment:
FS Table 4-2 - Cost
Response:
"As noted in Comments 9 and 17, section 4.0 of the FS,
there are discrepancies between this summary table and
the detailed cost estimates prepared in Appendix D."
Comment noted.
FS.4.17.
See responses to comments FS.4.9 and
Comment:
FS DraW1na 70084-B7
Response:
"See Comment 19 in section 4.0 of the FS concerning
the validity of using the ~ie1d GC data."
See response to comment FS.4.19.
Comment:
FS Drawina 70084-B23
Also, see
Response:
"See Comment 19 in section 4.0 of the FS.
Comment 4 in Section 6.0 of the RI."
See responses to comments FS.4.19 and RI.6.4.
FS Drawina 70084-B32
Comment:
Response:
Also, see
"See Comment 19 in Section 4.0 of the FS.
Comment 4 in section 6.0 of the RI."
See responses to comments FS.4.19 and RI.6.4.
FC ADDendix C.
Comment 1 lFS.C.1}
Text:
Comment:
"Alternatives were developed iteratively, using the
performance criteria listed above, to provide the desired
performance at both a minimum pumping rate, and with a
minimum number of wells." .
"Minimizing the pumping rate and minimizing the number
of wells are mutually exclusive objectives. You can
either minimize the pumping rate per well which
results in a larger number of wells or you can
maximize the pumping rate. per well which results in a
fewer number of wells. The alternative that maximizes
the pumping rate from the fewest wells would result in
the most cost-effective alternative, yet was not used
91
-------
Response:
in these analyses.
Why not?"
The commentor misunderstood the passage. The document
meant to explain that performance of the systea was
evaluated using many wells at a low pumping rate and
also fewer wells at a higher rate.
':'ext:
Comment FS.C.2
Comment:
Response:
"A set of six extrac~ion wells was located within the
plume to remove ~ontamination, and foundry well Fl was
react~vated as an extrac~ion well."
"Seven extraction wells plus the foundry well Fl were
simulated, not six extraction well.."
Comment noted.
Text:
Comment FS.C.3
Comment:
Response:
Comment
Text:
Comment"
Response:
"Individual extraction well discharge ranges from 0.2 cfs
to 0.5 cfs."
"Compared to alternative G3, these pumping rates are
lower. Why?"
The extraction wells are further from the city
production wells and can achieve the extraction
objectives with lower rates.
F8.C.4
"If Ross wells Rl through R4 were shut down, Figure C23
indicates that the extraction wells, pumped at the
originally specified rate, would provide incomplete
protection to public wells PW4 and PW5."
"If extraction'wells were used for providing treated
potable water, the pumping on PW-4 and pw-s could be
reduced or eliminated, thereby removing this concern.
U.8. EPA agrees (and has stated in the Proposed Plan)
that municipal use of treated water should be
considered during Remedial Design. The comment would
therefore be more appropriately addressed during
RD/RA.
92
-------
FS Table C1
Comment:
Response:
"Why weren't extraction rates equivalent to PW-4, PW-
5, and PW-6 (1.1 to 1.5 cfs) used for the EW wells (.2
to .5 cfs)?"
This is a design issue more appropriately addressed
during RD/RA.
FS Ficrure C.4
Comment:
"This map of Layer 1 heads suggests that a very large
till window occurs in the vicinity of PW-6 D4sed on
the large cone of depression depicted. The log for
PW-6 shows two till units exist at PW-6 above the
screen interval. One is 16 feet thick (79 to 95 feet
deep) and the other is 19 feet thick (128 to 147 feet
deep) (See HSI Table 1). PW-6 is screened fro. 150 to
175 feet deep. Why was a till window placed in the
vicinity of PW-6? If a till window was not placed
there, why does a cone of depression appear at the
water table at that location? .
This map also suggests a till unit is present at PW-4
and PW-5 due to the lack of simulated drawdown. Based
on the well logs, a till unit is present at PW-5~
however, no till units above the screened interval of
the well are present at PW-4.
The apparent till unit geometries used in these
simulations do not match the observed occurrence of
till. This places the validity of these modeling
efforts in question, especially considering the poor
calibration of the base model with the observed heads
as presented in the RI report.
What is the simulated water table mound located at
coordinates 15450E and l8550N? It appears as
significant as the sturgis Foundry recharge, yet was
not discussed.
Drawdown is simulated at coordinates 19450E and 19550N
suggesting a window in the till. No borings are
located in this area. What data were used to support
this?
Numerous nodes in Layer 1 go dry during the
simulation, but there is not discussion on this
There is a significant (19%) error in the water
for this simulation. MODFLOW runs usually have
balance errors of less than 1%, suggesting that
point.
budget
water
93
-------
Response:
something is wrong vith the vay the siaulation vas
executed."
The dravdovn noted by the co_entor at PW6 is induced
dravdovn from vi thdraval from the lover aquifer. The
magni tude of the dravdovn is a conaequence. of the
overall construction of the model and the geologic
interpretations from place to place vi thin the City of
sturgis.
The commentor questions the presence of a till unit at
PW4. According to the measured values, a localized
till uni~ .as identified at the water table near PW4.
See Cross Section EE' in the RI.
u.s. EPA disagrees vith the c0888ntors assertion that
the till unit geoaetries used in these siaulations do
not match the observed occurrence of till.
The aound at 15450E and 18850N is due to relatively
lover peraeability materials observed in the vicinity
W-8S (i.e., the boring log at W8 shoved silty fine to
coarse sand from a depth of 39' to the top of the
upper till unit (63.5') compared to fine to coarse
sand vith only a trace of silt and clay in other
areas. )
The apparent dravdovn at coordinates 19450E and 19550N
is a s8&ll overshoot problea in the interPOlation
routine. This issue can be further discussed during
Reaedial Design.
The co..entors concern about dry nodes is valid and
should be discussed further in Reaedial Design.
The commentor also expressed concern about the error
in the vater budget for the simulation. The apparent
error occurred because the production vells vere not
considered in the vater budget. When the vater budget
is corrected for the production vells, the error falls
to approximately 1'.
Comment:
FS Fiaure C.7
"Assuming the capture zone predicted for PW-6 is
reasonable, it vould appear that tvo to three vells
constructed similar to PW-6 could capture the entire
area of impacts. Why were extraction wells simulated
vith such lov pumping rates?"
94
-------
Response:
This is a design question which should be addressed in
RD/RA.
FS AQDendix D. Comment 1 CFS.D.1)
Comment:
"This is a general comment with r..Pect to the cost
estimating approach taken:
All cost estimates included the following assumed cost
factors:
. Contractor OVerhead of 20%
. Engineering of 20%
. Administra~ion of 5%
. Construction Documentation
of 10'
These factors
standards.
are high
when' compared to
industry.
The uni t rates provided in the estimate appear to
include the contractor's overhead and profit (as
generally included in the unit costs derived from the
"Means Manual"). Therefore, the 20% Contractor's
overhead does not appear to be warranted, nor
reasonable.
Engineering Costs typically range between 10' and 15'
for design purposes.
Construction Documentation Costs typically range between
5' and 7' unless "Construction Documentation" includes
construction management.
Response:
The Net Present Worth (NPW) factor of 10' is
unrealistic. The NPW factor, as defined in the FS, is
the difference between the inflation and interest rates.
At this time, inflation is running at 4.5' average, and
the interest rate is typically 9.5'. This would
indicate that a more appropriate NPW factor would be 5'.
Best professional judgement was used for development of
cost estimates. These costs were applied consistently
across all action alternatives, resulting in a sound
basis for cost comparisons. In addition, a 10% NPW
factor is required by the National Contingency Plan.
FS Table D1
Comment:
"No costs were included for pretreatment of water
hardness to prevent scaling in the air stripper. No
95
-------
Response:
FS Table D3
Comment:
Response:
costs were provided for a water reservoir to aeet
chlorination holding tiaes after air stripping." .
If treatability studies show pretreataent for inorganics
is necessary, it will be included in the design. If
municipal use of the treated water is eventually.
determined to be the best option, discussions would be'
necessary wi th the ci ty of sturgia to detenaine how
expenses would be split between Potentially Responsible
Parties and the city. For example, the cost of the
water reser~oir for chlorination may be the
responsibility of the city.
"No costs were included for pretreatJaent of water
hardness to prevent scaling."
See response to FS Table D1 above.
Comment:
FS Tables D4. DS. and D6
(1)
"Two injection wells were simulated (INJ1A and
INJ1B); yet costs are only shown for one
inj ection well. Why are seven booster pumps
required for the two injection wells?
The contingency for these alternatives was
mobile well head treataent units that could be
placed on PW-4, PW-5, and PW-6. Why weren't
these costs included in the analyses?
No costs were included for stOr8 sewer
installation/upgrade (existing sewers can't
handle hydraulic load on a permanent basis)."
INJ1A and INJ1B are the same well; one injects
into the upper aquifer and the other injects
into the lower aquifer. Booster puaps were
assumed to be necessary for each of the
extraction wells for delivery of water to
treatment or injection systems.
A contingency plan will be developed as part of
Remedial Design. Inclusion of the cost for the
actual contingency treatment unit is
discretional, as it mayor may not be needed.
A regional upgrade would not be needed; however,
it is possible that portions of the system would
require improvements.
Response:
(2)
(3)
(1)
(2)
(3)
96
-------
FS Table 07
Comment:
Response:
"Al ternati ve G4 pumps 3 . 5 cfs versus 5.9 cfs for-
Alternative G3. If G4 pumps and treats less water, why
are costs higher than G3? Why is the air stripper cost
for G4 at $55,000 and for G3 (Table D-4) at $40,000?
Why is the vapor phase GAC unit costs for G4 at $75,000
and G3 (Table 0-4) at $60,0001 Why are costs for liquid
phase GAC polishing unit included for G4 ($50,000) and
not for G3 (Table 0-4)?"
The comment concerning the air stripper costs is noted.
Vapor phase GAC unit costs for GW4 are greater than
those for GW3, because the ant::..cipated con~aminant 10CiG.",
expected at wells in Alternative GW4 are greater than
those in GW3. Liquid phase GAC polishing is included
for GW4, and not GW3, because of the higher levels of
contamination that GW4 wells would be expected to
encounter.
Comment:
FS Tables 07. 08. and 09
Response:
"Costs for using sturgis Foundry wells (F-l) were not
included. Contingency costs for mobile well head
treatment units were also not included. No costs were
included for storm sewer installation/upgrade."
C'omment concerning the sturgis Foundry well is noted.
See the response to comments on FS Tables 04, D5 and
06 for responses to additional concerns.
97
-------
SBC!'IO. VIII
Mark E. Simonett, Esq. and P.E., provided comments and a
literature review of the effectiveness of pump and treat
remedies in attainment of ARARs.
RESPONSE
U.S. EPA acknowledges that research into groundwater puap
and treat remedies have suggested that, in certain
instances, low level cleanup levels cannot be achieved.
Uncertainty in meeting cleanup objectives is an
in:"...:fficient reason for not initiating groundwater
ex-craction. At a minimum, impleaen'Cing groundwater
extraction would remove .ost of the dissolved contaainant
mass in the most contaminated portion of the aquifer in a
relatively short period of ti.., aa haa been the case at
many other sites documented by u.s. BPA. Proqress towards
cleanup will be evaluated after the groundwater extraction
system is implemented and, if it is determined that cleanup
objectives cannot be achieved, consideration will be given
to enhancing the groundwater extraction and treataent
system and/or establishing alternative concentration limits
(ACLs).
Mark E. Simonett's August 7, 1991 letter report appears to
be a thorough analysis of the current state-of-the-art for
DNAPL's contamination cleanup. In recoqni tion of the
difficulties of such cleanups, .uch research is underway
both nationally and internationally. This research holds
auch promise to achieve cost-effective in-si tu cleanup.
For example, Cooper Industries itself is undertaking field
research by the recent installation of a vadose zone water
flushing system at the former McGraw-Edison Plant site in
Albion, Michigan. This project will provide the ability to
enhance the cleanup rate of the existing pump and treat
system. Future modifications may include the addition of
surfactants or nutrients to enhance flushing and/or
bioremediation.
Another promising ,adaptation to pump and treat is the use
of in-situ methanotropic and/or aerobic bioremediation.
Under this technoloqy, DNAPL' s would be effectively treated
in the dissol ved and absorbed phases. The liquid DNAPL
phase may be more resistant, but here again sparqinq
technology is emerging. Sparqing by air, heat, nitrogen,
etc., is potentially effective to reach all of the DNAPL
areas.
Until it is shown that the selected technologies cannot
achieve cleanup levels for this Site, the remedy must be
designed to be protective of human health and the
98
-------
o
environment and comply with ARARB. since the stated
cleanup levels are based on protection and ARARs, U~S. EPA
believes that the selection of Alternative GW4 is
consistent with Section 300.430(e) of the NCP.
99
-------
.BC'1'1:08 1:Z
Hydro-Search, Inc.
remediation concepts.
(HSI)
provided
alternate
qroundwater
RESPONSE
HSI's alternative groundwater remediation concept proposes
to conduct aquifer restoration via extraction wells, with
treat!t'ent and use of the extracted water in the City's
potable water supply sys't.em. HSI proposes that PW6 be
utilized to augment the potable supply, and that PW4 and
P"..i5 be used for back-up purposes. Treatment of t.'1e
extracted water could be a aultistage process, with iron
and manganese removal, air stripping with vapor-phased
granular activated carbon prior to chlorination and
introduction into the City supply sy.tea. Essentially, HSI
proposes that the City switch its water supply source to
wells extracting the most contaminated groundwater in the
plume, with treatment to potable standards.
This approach has technical merit and looks proaising, in
that the most contaminated water in the aquifer is
extracted, and withdrawal from the periphery. of. the
contaminated plume is reduced. This flow field could flush
the aquifer aore rapidly than would occur if substantial
pumping is continued froa wells PW4, 5 and 6. However,
aquifer restoration issues to be resolved include:
.
The infiltration of excess withdrawal water at the
Kirsch Plant No.2 could cause the further spread of
contamination south and west of the infiltration area.
This issue will have to be considered further during
RD/RA.
.
Areas of the contaminated aquifer at the southern margin
of the plume may not be effected by extraction wells
placed in the center of the plume. Well EXW6 in
Groundwater Alternatives 4A and 48 was placed to contain
and extract the southern extent of contamination.
.
More than two wells may be required to achieve the
extraction rates proposed, which HSI recognized in the
text of their proposal, but not in their cost estimate.
In addition, the proposed alternative restoration plan
raises substantial issues relating to public policy and
community acceptance. Proposing that a co_unity
deliberately utilize the most contaminated portion of an
aquifer for its primary potable supply, while not violating
ARARa, certainly has the potential to generate public
controversy. It is for that reason that the groundwater
100
-------
6
remediation al ternati ves developed in the FS di4 not
include new wells for extraction ot contuainated water tor
potable use. However, u.s. EPA has long recognized (and
did recognize in the Proposed Plan) that discharge of
treated water into the municipal system is a possibility
under Al ternati ves GW3 and GW4. It is felt. that this iS8ue
should be left open until remedial design stage when
modeling and potential public policy issues can be worked
out.
101
-------
SBC'fIO. Z
Cooper Industries, Inc. provided a copy of their CERCIA I 104 (e)
response originally submitted to U.8. EPA on December 13, 1989.
ResDonse
This information is not relevant to the remedy selection
process and the appropriateness of the proposed remedy and
therefore will not be responded to in this document.
B!':-C'TrC~ %1
As part of their comment package, Cooper Industries, Inc.
provided a transcript of public meeting at 8turgis Public
Library, June 20, 1991.
ResDonse
U.8. EPA appreciates the extra copy of the public aeeting
transcript provided by Cooper Industries. As is Agency
pOlicy, this will be added to the administrative record.
SBC'f'I08 SII
Robert W. .Teets, Director of Risk Management and Environaenta1
Affairs for Cooper Industries, provided copies of correspondence
with Terese A. Van Danse1, Remedial Project Manager, U.8. EPA.
ResDonse
This informa~ion is not relevant to the remedy selection
process and the appropriateness of the proposed remedy and
therefore will not be responded to in this document.
However, correspondence that took place during the public
comment period will be added to the Administrative Record.
SBC'f'IOR SIll
Correspondence dated August 12, 1991, from Mark E. Mulhollam,
Attorney for cooper Industries, Inc., to Terese A. Van Dansel,
Remedial Proj ect Manager was submitted. The correspondence
included a summary of TCE and PCE toxicological literature.
ReSDonse
U. 8. EPA notes the submission of toxicity information
concerning TCE and PCE. U.8. EPA currently considers TCE
and PCE to be probable human carcinogens (EPA Group B2).
102
-------
<>
SBc-rXOB xxv
Robert W. Teets, Director of Risk Management and Environmental
Affairs for Cooper Industries, submitted correspondence dated
September 12, 1991, to Terese A. Van Donsel, Remedial Project
Manager for U.S. EPA. Pump test data for active city wells were
enclosed.
FeS'Donse
This information was submitted well after the close of
public comment period, but was, as req-Je:;ted by Cooper,
added to the administrative record. Cooper reiterated its
concerns about the RI/FS and requested that the public
comment period be reopened to allow consideration of the
pump test data enclosed with the correspondence. After
consideration of Cooper's arC)U8ents, U. S. EPA has
determined that the reopening of the public comment period
is not necessary.
Cooper also proposes additional studies to be conducted ~y
its contractor, Hydro-Search, Inc. According to Mr. Teets,
the purpose of these efforts would be to, "design a well
field that will achieve capture of the contaminant plume in
the lower unit while minimizing the volume of impacted
water extracted from the fewest number of wells." As
stated many times in this Responsiveness Summary, the model
put forward in the FS to represent alternative GW4, Aquifer
Restoration, is not the final design for remediation.
Neqotiations will be conducted with potentially Responsible
Parties to afford them the opportuni ty to conduct the
Remedial Design/Remedial Action. Thus, since this
opportunity will be provided in the future before
remediation begins, it is not necessary to debate the
merits of one design over another or afford the opportunity
for the development of a competing design. . .
Cooper asks that copies of the results from September 1990
and September 1991 sampling be made available and added to
the administrative. record. Volatile sampling from
Septellber 1990 was unusable due to holding time exceedances
by the contract laboratory. However, the data may be
reviewed at MOHR offices. In August 1991, the wells were
re-sampled, but the validated data is not yet available.
Once this data is validated, copies will be provided to
Kirsch Co. and cooper Industries.
a...
103
-------
P.
SICDn:PICAII'J.' STAR COIIIIBII'1'S
The state of Michigan has reviewed the ROD and concurs on
the selected alternative.
104
-------
"
~
PDge No.
09/19/91
ADMINISTRATIVE RECORD INDEX
UPOATE "
STURGIS WELL FIELD SITE
STURGIS, MICHIGAM
flCIIEtrR/lME PAGES DATE
TITLE
AUTHOR
RECIPIENT
DOCUMENT TYPE
-
DOC NUMBER
"BOrn RAW AND VAUDA'TlW DATA CAN BE MADE AVAIlABLE .
mROUGH mE u.s. EPA AND TRE MICHIGAN DEPARTMENT OF
NATVRAl. RESOURCES (MDHR). INTERESrm PAR7JES SHOUlD
CONTACT MS. 1ERESE VANDONSEl. OF 77lE U.S. EPA AT .
(JI2) JjJ~j64 OR MR. ROB FRANKS OF MDHR AT
(517) JJj.JJ92"
7 91/06/12 ll!t ~.r re: Sturgis Fren~ J. Keleaky, T. VenDoneel, USEPA Correspondence
F~"V Corporation'. Sturgis Foundrv Corp.
options to remedial
action plan
(Attached EP Toxicity
and TLOP results)
91/06/19 letter re: Thenlt you Robert V. TeeU, T. v...oonsel, USEPA Correspondence 2
for 30 day ext-Ion Cooper IrOatr'"
to comnent on
feasibility Study Reports'
and request for IIIHtlng
91/07/11 letter re: Cooperation Paul Vartner, State T. VenDonsel, USEPA Correspondence 3
of USEPA with Kirsch Senetor, State of
Division of cooper Michigan
Industries and their
concerns about
completeness of RIIFS
2 91/07/11 letter re: Installition lelson M. OI8V8r,I., T" VarDonsel, USEPA Correspondence "
of fence It rl rICh Plant cooper Indultrl..
110. 1 end enclosed
dlagr..
91/01123 letter re: Copy of Terese A. VenDonsel, It. 'erod, Gradfent Correspondence 5
Michigan Risk Assessment USEPA Corp.
Guidelines
B 91/07/24 letter re: CII,Iflcltion Itobert V"Teets, T" VarD_el, USEPA Correspondence 6
concerning dati or Cooper I~trles ..!
docuaents available In -,
the administrative record
for Sturgis Veil Field
91/07/25 letter re: Copy of Terese A. v...oonsel, R. 'erod, Gradient Correspondence 1
Append I x A and Append I x USEPA Corp.
0 of the Quality
Assurance Project Plan
dated August 1981
91/01130 letter re: Request Nelson N. Ollvarrll, T. VerClonsel, USEPA Correspondence 8
-------
P
-------
"
Page No.
09/19/91
3
FICHE/FRAME PAGES DATE
74
91/09/12
2
91/08/30
.'...,
ADMINISTRATIVE RECORD INDEX
UPDATE '1
STURGIS WEll FIELD SITE
STURGIS, MICHIGAN
TITLE
AUTHOR
RECIPIENT
DOCUMENT TYPE
DOCNUMBER
,
letter re: Cooper' I
concerns about denial
of ~etlng with USEPA
and the remedial
Investigation
Robert W. Teets,
Cooper Industries
T. VenDonsel, USEPA Correspondence
17
News Article Entitled:
Coooer presents city
welt cleanuc plan
Robin Johnson, Sturgis
Journal
Newspaper Articl~
18
24 00/00/00 Cooper Industries, Inc:. 'I Cooper Industrlel, Inc:. USEPA, MONR Other 19
COII8ent In Regard to
USEPA/MDNI Identification
of Potentially V
Responsible Parties
vJ(L\~5
275 00/00/00 Cooper I ncbItr les , Inc.'1 Cooper Incbltrles, Inc:. USEPA and MONR Other
Cannents on the Remedial \'? 1
Investigation/Feasibility ~~~1
Study with attached
CERClA Cclq)lIence with ~~I( 'I r (;'
Other laws Manual)
37 89/12/13 Letter re: Enclosed Villette A. Lemelle, . f1.~ ~ (.
G. Vatts. USEPA Ott .(\.~
response of Cooper Cooper Incbltrles
Irdatrles, on beIIal f
of Kirsch, to Infol'88tlon
Request letter dated
lIoveneer 7, 1989
43 91/06/20 Trenscrlpt of Public Other 22
Hearing In the ..tter
of Sturgis Munlclpel
Wel l Field S~rfU1d
Site
1160 91/08/07 Letter re: LI terature Mark J. Simonett, II. Olavarria, Cooper Other 23
review of effectiveness Gray, Plant, Moody, Ind.
of ~ end Treat Moody & Bennett,
Re8edy to Attain ARARs Attorneys for Cooper
(Attached copies of Incbltrles
all publlcatlons
referenced In letter
and additional
pertinent articles
listed In the
bibliography)
-------
Page No.
09/19/91
4
FICHE/FRAME PAGES DATE
97
90/12/03
869
91;08/00
158
91/08/09
30
91/08/09
321
91/08/09
52
91/08/14
TITLE
Draft, SUte of
Michigan Risk Assessment
Guidelines
Critique of Warzyn's
Baseline Risk As~essment
for the Sturgis Upl!
Field Superfund Site
and attached referenced
docunents
letter re: Nearby
municipalities which
cons..- treated
contaminated groundwater
(Attached Record of
Decision, Selected
Remedial Alternative,
Vest Kl Avenue landfill
site)
Report Entitled:
Alternative Ground-
Vater Remediation
Concept, Sturgis Well
Fleld"Sturgla, Michigan
Revi ew C08IIIerIts,
Sturgis VeIl Field
Renedial Investigation!
Feasibility Study
letter re: Copies of
the AlterNltlve
Remediation Concept,
Review Comments for
Remedial Investigation!
Feasibility Study and
Curriculum Vitae for
the professional staff
at Hydro-Search, Inc.
ADMINISTRATIVE RECORD INDEX
UPDATE '1
STURGIS WEll 'FIELD SITE
STURGIS, MICHIGAN
AUTHOR
COU'IC i I on
Environmental Quality
Committee on Risk
Assessment'
Gradient Corporation
Chri stopher L., SIIi th
Cooper Industries
Hydro-Search, Inc.
Hydro-Search Inc.
Mactay R. Hyde,
Gray, Plant, Moody,
Moody & Bennett, P.A.
Attorneys for Cooper
Industries, Inc.
RECIPIENT
DOCUMENT TYPE
,
Reports/Studies
Attorneys for Cooper Reports/Studies
Ind.
M. Hyde, G.P.M.M.&
B.
Reports/Studies
Attorneys for Cooper Reports/Studies
Ind.
USEPA
Reports/Studies
T. YanDonsel, USEPA Reports/Studies
4. ~
DOCNUHBER
24
2S
26
27
28
29
-------
(,) Page No.
09/19/91
GUIDANCE DOCUMENTS INDEX, UPDATE "
v STURGIS ~LL FIELD SITE, STURGIS, MICHIGAN
Guidance Documents ere eveileble for review et
USEPA Region V-Chicego IL
T I HE AUTHOR DATE
,
Community Reletions in USEPA 88/06/00
Superfund: A Hancbook
(Interim Guidance)
Guidance for Conducting USEPA 88/10/00
~emedlal Investigations
and Feasibility Studies
(~l/FS) Under CE~CLA
Policy for Superfund RCRA/HSUA 89/01.117
Complience with the
RCRA Lend Disposal
Restrictions
Superfund LOR Guide RCRA/HSWA 89/07100
11: Overview of RCRA
Lend Dlsposel
Restrictions (LORs)
The Feeslbllity Study: USEPA 89/11/00
Development end Screening
of Remedlel Action
AI ternetives
A Guide to Developing USEPA 89/11/00
Superfund Proposed Plena
Risk Assessment Guidance USEPA 89/1Z/00
for Superfund, Volume I:
Human Heelth Eveluetion
Manuel, P8rt A
Basics of Pump end Treat USEPA 90/09/00
Ground Water Remedietion
Technology
.!
-------
Page No.
09/19/91
ACRONYM GUIDE for the Administrative Record
Updete '1
Sturgis Veil Field Site
Sturgis, Michigan
ACRONYM
DEFINITION
ARARs
Applicable or Relevant
and Appropriate
Standards, limitations,
Cd ted II, IInd
j(e~u'remen'::s
EP
fS
Extrllction Procedures
feasibility Study
GPMMB
Gray, Plant, Moody,
Moody & Bennett
Michigan Department
of Natural Resources
Quality Assurance
Project Plan
Remedial Investigation
MONR
QAPP
RI
RIIFS
Remedial Investigation
/Feasibillty Study
Toxicity Characteristic
leachate Procedure
Uni ted States
Environmental Protection
Agency
TClP
USEPA
. .
. .~~;.(.
.
. ..: .
\
.,
-------
-OFFICE COpy
REMEDIAL ACTION
ADMINISTRATIVE RECORD
(Index and Documents)
for the
STURGIS WELL FIELD SITE
STURGIS, MICmGAN
JUNE 1991
United States Environmental Protection Agency
. Region V
130 South Dearbom Street
Chicago, IL 60604
. :;. ;~}~~~.
.
-------
Page No.
06/05/91
~.~;'.~~,
ADMINISTRATIVE RECORD INDEX
STURGIS YELL FIELD SITE
STURGIS, MICHIGAN
FICHE/FRAME PAGES DATE
TITLE
AUTHOR
RECIPIENT
DOCUIEIIT TYPE
DOCNUMBER
"
2
91/04/30
Letter re: Belief that
Sturgis Well should be
cleaned up to Michigan
Environmental Reoonse
Act Type B levels for
both ground wate~ and
James S. Linton, Michlgan.N. Tyson, USEPA
Department of Natural
Resources
Correspondence
~
.:our.:~ aree so; l
contami net i on
130
91/05/20
Copies of Act 307 and
the Administrative Rules
to Act 307 (with cover
letter)
Michigan Department of
Natural Resources
T. Yan Donsel, USEPA Ple8dings/Orders
2
87108/19
Sturgis Groundwater
Study to Be Discussed
at Plbllc Meeting
Plbllc Notice
3
88/04I2D Notice, Plbllc Meeting
Sturgis Groundwater
Study
44 00/00100 City of Sturgis. Gove Assoc I ates
Preli..lnar-y Investigation
01 Trichloroethylene
ContMINltion
6 00/00/00 Community Relations Plan
4 87/04/25 Progress Report ", Michigan Department
City of Sturgis, 01 Natural Resources
Ground Water
Cont..inatlon Study
64 81/07/00 Final Work Plan, Warzyn Engineering, Inc. MONR
Remedial Investigation/
Feasibility Study
Plbl'c Notice
4
Reportl/Studles
5
Reports/Studies
6
Reports/Studies
7
Reports/Studies
8
12 87/01/31 Health and Safety Plan, warzyn Engineering, Inc. MONR Reports/Studies 9
Remedial Investigative
Activities
199 87/08/00 Final QAPP, Remedial Warzyn Engineering, Inc. MONR Reports/Studies 10
Investigation!
Feasibility Study
247 88/02100 Phase 1 Technical Warzyn Engineering, Inc. MONR Reports/Studies 11
-------
p Pag~ No. 2
06/05/91
ADMINISTRATIVE RECORD INDEX
STURGIS WELL FIELD SITE
STURGIS, MICHIGAN
FICHE/FRAME PAGES DATE 11 TLE AUTHOR RECIPIENT DOCtJN&NT TYPE DOCNtMBER
Memorandum, Sturgis Well
Field RI/FS
3 88/04/08 Progress Report '2 Michigan Department of Reports/Studies 12
Sturgis Municipal Natural R~sources
~ellfietd Sucerfund
Site
:31 a8/05/.)0 pnase II CAPP, ~arzyn Engineering, Inc. "'IIR Reports/Studies 13
Revision " Remedial
Investigatlon/Feaslbllty
Study
24 88/05/00 Phase II Work Plan, Warzyn Engineering, Inc. IIIQ ReportS/Studies 14
Revision "
Remedial Investlgation/
Feasibllty Study
9 88/07/18 Cover letter with Kenneth J. Quim, S. Luzkow, IIINR ReportS/Studies 15
attached Hellth and Warzyn Engineering, Inc.
Safety Plan Addendul,
Remedial Investigative
Activities
36 89/03/00 Phase 118 Work PllnS, Warzyn Engineering, Inc. III... Reports/Studles 16
RifFS
11 89/03/10 Preliminary Hellth Center for EnvfronRental ATSOR ReportS/Studles 17
Assessment Health Services (CEHS),
Michigan Depertl8lt
of Publfc Health (MOPH)
19 89/04/00 Work Plan Adderan, Warzyn Engineering, Inc. IIIQ Reports/Studles 18
Feasbilfty Study
5. 89/04/05 Progress Report 13 Michigan DePirteent Reports/Studies 19
. Sturgis Municipal of Natural Resources
Well field Site
228 89/05/00 Pret iminary Findings and lIarzyn Engineering, Inc. IIINR Reports/Studies 20
Interpretations, Phase II
Technicat Memorendua,
Sturgis Well Field RI/FS
86 90/02/00 Phase 118 Technical Warzyn Engineering, Inc. IIINR ReportS/Studies 21
Memorandum, RI/FS
22 90/09/12 Qual i ty ~Iurance Terese Van Donsel, USEPA, "'NR Repol'ts/Studies 22
Pl'oject Plan, Addendum USEPA
-------
Page No.
06/05/91
3
ADMINISTRATIVE RECORD INDEX
STURGIS YELL FIELD SITE
STURGIS, MICHIGAN
FICHE/FRAME PAGES DATE
TITLE
AUTHOR
RECIPIENT
DOCUMENT TYPE
DOCNUMBER
..
248 91/03/00 Remedial Investigation \larzyn Engineering, Inc. MONR Reports/Studies 2J
Report, Remedial
(nves' on/
Fea~'- : StudY
Vo: .,'" 5
42 91/03/00 Remedial Investigation \larzyn Engi~ring, Inc. MONR Reports/Studies 24
Report, Remedial
Investigation/ ~
Feasibility Study
VollJIIe 2 of 5
364 91/03/00 Remedial Investigation \larzyn Engineering, Inc. MONR Reports/Studies 25
\ Report, Remedial
Investigation!
Feasibllty Study
VollJIIe 3 of 5
295 91/03/00 Remedial Investigation \larzyn Engineering, Inc. MONR Reports/Studies 26
Report, R~ial
Investigation!
Feasibility Study
VollJIIe 4 of 5
165 91/03/00 Remedial Investigation Warzyn Engineering, Inc. MOIIR Reports/Studies 27
Report, Remedial
Investigation!
Feasibllty Study
VollJIIe 5 of 5
189 91/05/00 Feasibility Study, \larzyn Engineering, Inc MOIIR Reports/Studies 28
Sturgis Well Field,
Volune 1
179 91/05/00 Feasibility Study, \larzyn Engineering, Inc. MONR Reports/Studies 29
Sturgis Well Field,
VollJIIe 2
-------
Ii'
?
Page No.
06/05/91
TITLE
Risk Assessment Guidance
for Superfund, Volume I:
Human Health Evaluation
Hanual. Part A
8asl~s of Pumc and Tr~at
Ground ~ater RemedIatIon
Technology
Community Relations in
Superfund: A Handbook
(Interim Guidance)
The Feasibility Study:
Development and Screening
of Remedi.1 Action
A I ternet i ves
A Guide to Developing
Superfund Proposed Pllns
Policy for Superfund
Compllince with the
RCRA Lind Disposil
Restrictions
Superfund LOR Guide
'1: Overview of RellA
land Disposil
Restrictions (LDRs)
"*
CUIDANCf DOCUMENTS INDEX
STURGIS VElL FIELD SITE, STURGIS, MICHIGAN
Guidance Documents Ire Ivaillble for review It
USEPA Region V-Chicago Il
AUTHOR
DATE
,
USEPA
89/12/00
USE?A
'J()/09/00
USEPA
88/06100.
USEPA
89/11/00
USEPA
89/11/00
RCRA/HSVA
89/04/17
RellA/HSVA
89107100
-------
PlIge No.
06/05/91
ACRONYM GUIDE for the Administrltive Record
Sturgis Well field Site
Sturgis, Michlg8n
ACRONYM
DefiNITION
..
ATSDR
Agency for Toxic
Slbstlnces lnet
Diselse Registry
Center for Env;ro~tal
Health services
Fellsib,iity Study
,
CE"S
r~
MONR
Michigan Deplrtment
of Natural Resources
Michigln Deplrtment of
Plblic Hellth
Quality Assurence
Project Plen
Remedl.1 Investig.tlon
MOPH
QAPP
RI
Rl/fS
Remedial Investig.tion
/feasibility Study
Uni ted States
Environment.l Protection
Agency
USEPA
"".~::T~~;
.,,;:
------- |