PB94-964504
EPA/ROD/R09-94/108
July 1994
EPA Superfund
Record of Decision:
Brown and Bryant Arvin Facility
Site, Arvin, CA
-------
BROWN & BRYANT, ARVIN FACILITY
FIRST OPERABLE UKIT
RECORD OF DECISION
PART I: DECLARATION
PART II: DECISION SUMMARY
PART III: RESPONSIVENESS SUMMARY
BROWN & BRYANT, ARVIN FACILITY SUPERFUND SITE
ARVIN, CALIFORNIA
United States Environmental Protection Agency
Region 9 - San Francisco, California
-------
TABLE OF CONTENTS
Page No.
Part I. Declaration 1
Part II. Decision Summary 3
I. Site Name, Location and Description 3
II. Site History and Enforcement Activities 3
III. Highlights of Community Participation 4
IV. Scope and Role of the Operable Unit 5
within the Site Strategy
V. Summary of Site Characteristics 6
VI. Summary of Site Risks 9
VII. Description of Alternatives 10
VIII. Summary of Comparative Analysis 17
of Alternatives
IX. The Selected Remedy 21
X. Statutory Determinations 23
XI. Documentation of Significant Changes 25
Part III. Responsiveness Summary 26
Summary of Response To Commenter's 26
Major Issues and Concerns
Detailed Response to Comments 29
-------
PART I - DECLARATION
BROWN & BRYANT, ARVIN FACILITY
ARVIN, CALIFORNIA
CAD052384021
Statement of Basis and Purpose
This decision document presents the selected remedial action for the
Brown & Bryant, Arvin facility in Arvin, California, which was chosen in
accordance with CERCLA, as amended by SARA, and, to the extent practicable,
the National Oil and Hazardous Substances Pollution Contingency Plan . (NCP).
This decision is based on the administrative record for the site.
The State of California concurs with •the selected remedy.
Assessment of the Site
Actual or threatened releases of hazardous substances from the site,
if not addressed by implementing the response action selected in the Record
of Decision (ROD), may present an imminent and substantial endangerment to
public health, welfare or the environment.
Description of the Selected Remedy
This operable unit is the first of two planned operable units for this
site. The first operable unit addresses the surface soil, •the subsurface
soil and the shallowest groundwater unit, the A-zone groundwater. The
function of this operable unit is to address the principal threat at the
site, the A-zone groundwater, and to address the surface soil exposure
threat.
The major components of the selected remedy include:
Extraction, treatment and reinjection of the shallowest groundwater
unit;
Consolidating contaminated surface soil on a 1.2 acre portion of the
site and constructing a Resource Conservation and Recovery Act (RCRA)
Subtitle C cap over it; and
Capping the remaining portion of the site with a basic cap.
statutory Determinations
The selected remedy is protective of human health and the environment,
complies with Federal and State requirements that are legally applicable or
relevant and appropriate to the remedial action, and is cost-effective.
This remedy uses permanent solutions and alternative treatment technology
to the maximum extent practicable, and satisfies the statutory preference
for remedies that employ treatment that reduces toxicity, mobility, or
volume as a principal element.
-------
Because this remedy will result in hazardous substances remaining on-site
above health-based levels, a review will be conducted every five years
after commencement of the remedial action to ensure that the remedy
continues to provide adequate protection of human health and the
environment.
//- & -
Date
gional Administrator
-------
PART XI - THE DECISION SUMMARY
BROWN i BRYANT, ARVIN FACILITY
ARVIN, CALIFORNIA
I. Site Name, Location and Description
The Brown & Bryant (B&B) Arvin facility Superfund Site is
located at 600 South Derby Street in Arvin California,
approximately 30 miles southeast of Bakersfield, California. A
location nap is shown on figure 1. Brown & Bryant was a pesticide
refemulator and custom applicator facility from 1960 to 1989.
Arvin is an agricultural community with a population of
approximately 9,300 people. The site is located in a light
industrial and commercial area, with a residential area located
across the street.
The site includes a former waste pond in the southeast corner
and a former sump area that has been excavated and replaced with
clean fill. Run-off drains to the southern portion of the site
between the waste pond and the sump and to a lesser extent to the
southwest corner. The site is enclosed by a fence. A site map is
shown,on figure 2.
The site geology has been divided into two zones: the A-zone
and the B-zone. The A-zone includes the unsaturated soil between
the surface soil and the first groundwater, and the first
groundwater unit, referred to as the A-zone groundwater. The B-
zone includes unsaturated soil below the A-zone and the second
groundwater unit, the B-zone groundwater. The B-zone extends to a
depth of at least 250 feet and ends at a clay layer known as the
Corcoran Clay which confines the drinking water aquifer below it.
A conceptual site cross-section is shown on figure 3.
There are no wetlands or surface water within a half-mile of
the site. Contamination at the site does not impact any surface
water feature.
II. Site History and Enforcement Activities
Brown & Bryant was a pesticide mixer and custom applicator
facility from 1960 to 1989. Contamination of the soil and
groundwater resulted primarily from poor housekeeping, spills and
leaks from a surface pond and sumps. In 1981, the facility was
licensed under RCRA as a hazardous waste transporter. In 1983, the
State of California required Brown & Bryant to conduct a site
investigation and dispose of contaminated soil.
In October 1989, the Brown & Bryant site was listed on the
National Priorities List. EPA immediately conducted an emergency
-------
response assessment and identified two areas needing immediate
attention, a dinoseb spill area, and the groundwater which appeared
to pose an imminent and substantial endangerment to the municipal
drinking water. EPA treated the dinoseb-contaminated soil in the
winter of 1991 under its emergency response authorities. The
investigation of the threat to the municipal water well. was
conducted in the sprinq of 1991 by the Railroads, as descrl.bed
below.
At the time EPA became involved at the site, the state of
California was evaluatinq a settlement with Brown & Bryant and its
insurers to pay for past costs incurred by the State and the future
costs of clean-up. EPA entered the neqotiations to recover federal
costs. The negotiations ceased in 1991, when the civil complaint
filed by Brown & Bryant against its insurers in state court was
dismissed for failure to prosecute.
In October 1990, EPA issued general notice letters to two
other site property owners, Atchison, Topeka and Santa Fe Railway
and Southern Pacific Transportation Company (the Railroads). In
January 1991, EPA issued the Railroads an administrative order to
conduct certain investiqations of the groundwater at the site. The
work was completed in August of 1992. While the emergency removal
work and Railroad-conducted work were underway, EPA conducted a
Remedial Investiqation and Feasibility Study of the surface soil,
subsurface soil and the first groundwater 'zone. The RIfFS report
was completed in May 1993.
A final investiqation coverinq the remaininq groundwater zones
was initiated in the fall of 1992 and is still in proqress.
III. Hiqh1iqhts of Community Participation
The CERCLA requirements for public participation include
releasing the RIfFS report and the proposed plan to the public and
providing a public comment period for those documents. EPA
satisfied these requirements by placing these documents, along with
other administrative record documents, 'in the pUJ:)lic information
repositories at the Beale Library in Bakersfield and the EPA
document center in San Francisco. EPA also mailed the proposed plan
to interested individuals on the mailinq list. Furthermore, EPA
conducted a public meetinq on July 6th, 1993 in Arvin where the
proposed plan was presented and comments were accepted from the
public. The notice of the public meeting was published in the
Arvin Tiller on June 23, 1993 and El Mexicalo on June 24, 1993.
In addition to the statutory community participation
requirements, EPA also sponsored several community outreach events
to keep the community informed. This included three community
interviews in February 1990, September 1991, and April 1992. The
first and third community interviews were conducted to distribute
4
. .'.- ~-"-----,,-------------_.._. _.-.
-'---"-
- - ..-- - -'-----------..,- . ... -
"----".---'''''-- -
-------
. - ...- _.. - ..... .~ o.
..~._.__._---~-_.- .--..
. --- ._..~.. "0' _....__._-'~'... ~----~'-_._'-----'_.--------_u_-----~-----'..'.'.
fact sheets and answer questions from the adjacent neighbors of the
'Brown & Bryant facility. The second community interview reached
out to the larger Arvin community in order to determine the
community's needs and interest. EPA also distributed four fact
sheets to the community. The first three sent in February 1990,
April 1991 and March 1992, explained the clean-up activities
currently underway or expected to be started soon at the site.
The final fact sheet ,included the proposed plan and was sent June
25th, 1993.
, This decision document presents the selected remedial action
for the first ,operable unit at the Brown & Bryant, Arvin facility
site in Arvin California, chosen in accordance with CERCLA, as
amended by SARA, and, to the extent practicable, the National
Contingency Plan. The decision for the site is based on the
administrative record.
IV. Scope and Role of Operable Uni~ wi~hin ~he Si~e Strategy
EPA has divided the site into two operable units., The first
operable unit includes the current source of contamination, the A-
zone groundwater, and the surface and sub-surface soils. The
second operable unit includes the deeper groundwater units.
, ,
, The response actions selected in this ROD address the first
operable unit. Response actions for the surface soils constitute
a final, remedy for the surface soils. The actions for the
subsurface soil and the first groundwater are interim actions.
The primary objective for the subsurface soils and the A-zone
groundwater response action is to control migration of the
contamination in this zone to deeper groundwater. Based on the
water production rates I the A-zone groundwater is not legally
classified as' a potential drinking water source. However, the B-
zone groundwater is classified as' a potential drinking water
source. Therefore, the clean-up goal is to reduce the
contamination levels in the A-zone to levels that would, protect the
B-zone groundwater. The A-zone,has caused chemical levels in the
B-zone groundwater to exceed maximum contamination levels set by
EPA.
The A-zone groundwater is classified a~ a principal threat at
the site. A principal threat is characterized as a waste that
cannot be reliably controlled in place, such as liquids and high
concentrations of toxic compounds (e.g. several orders of magnitude
above health based levels),. The response action for the A-zone
groundwater'satisfies the statutory preference for remedies
employing treatment that reduces toxicity, mobility, or volume as
a principal, element.
The primary objective of the surface soils response action' is
5
-------
to prevent human and ecological exposure to the contaminated soil.
The most contaminated soil was addressed in an emergency response
removal in 1991. The remaining surface contamination is not
considered a principal threat because it is not highly mobile, is
not several orders of maqnitude above health based levels and can
be effectively controlled in. place. The response action for the
surface soils includes consolidation of soils exceeding health-
based levels onto the southern portion of the site, containment
(capping) and institutional controls.
v. summary of site Characteristics
The geoloqy'at the.site is an alluvial deposit of alternating
layers and mixtures of unconsolidated sands, silts and clay. The
stratiqraphy is very heterogeneous and layers tend to be
discontinuous. The site geoloqy has been divided into two zones.
The A-zone includes unsaturated soil to 65 to 75 feet.below ground
surface (bgs) and includes the first groundwater unit, the A-zone
groundwater. The base of the A-zone is a thin sandy clay layer
from 75 to 85 feet bgs. The clay .layer and the A-zone groundwater
occur under the entire site but disappear within 900 feet south of
the site. The B-zone includes unsaturated soil below the A-zone
and the second qroundwater or the B-zone groundwater at 150 to 155
feet bgs. The B-zone extends to at least 250 feet bgs and ends at
a clay layer known as the Corcoran Clay which confines the drinking
water aquifer ~elow it. The thickness of this clay layer at the
site is unknown. (See figure 3 for Conceptual Cross-section)
Surface Soil.
Surface soil is defined to include the upper seven feet of
soil. This depth includes a "construction zone", a depth where
excavation might occur in the future for utility work. Sampling"
results from. the surface soil identified dinoseb as the only
contaminant of concern. The principal hqt spot of dinoseb
contamination,occurs in the location of a former spill, along the
east fence-line. High concentrations of dinoseb in surface soils
were also found scattered in three other locations on-site and low
concentrations were found over much of the site. The area of
highest dinoseb contamination in the dinoseb spill area was cleaned
in 1991; however, some soil contamination exceeding health-based
levels still remains in this area. .
Subsurface Soil
Soil contamination from a depth of seven feet down to the A-
zone groundwater was found over much of the site, but was primarily
concentrated under three areas: the sump area, the dinoseb spill
area, and the waste pond and a topographic low area between the
.6
- . - .-.-- - - - -.~-_._-,----_._~.. -. -... - -_...
. -"'" -...'" -_..'n - - - "" .
-------
"'~--'~'-_.,.~ --_._~,...... ~"-"
.,~ ~. -. .-- -..----...----...-.--..--- - -_.- - ,.--. ~. '---'._+--'~'~--~'--
pond and the larqe storaqe tank in the southwest corner of the
site. Within these three areas and over the entire site, six
chemicals were identified as occurrinq at hiqhest concentrations
and to the qreatest extent wi thin the A-zone soils . These
chemicals are 1,2-dichloropropane (1,2-DCP), 1,3-dichloropropane
(1,3-DCP), dibromochloropropane (DBCP), 1,2,3-trichloropropene
(1,2,3-TCP), ethylene d~bromide (EDB), and dinoseb. All of these
chemicals except for dinoseb are volatile orqanic chemicals.
Dinoseb was found concentrated in the top 30 feet of the' spill
area and then declined siqnificantly in concentration down. to the
A-zone qroundwater. In the pond and sump areas, the concentrations
were siqnificantly less than in the spill area.
Volatile orqanic contaminants were found in the subsurface
over the entire site but were found in hiqhest concentrations in
the sump area.. One borinq in particular, borinq I (located in the
center of the sump), stands out for its exceptionally hiqh
concentrations. These contaminants were also found at siqnificant
levels in the area of the waste pond, and. then were found in only
relatively small concentrations elsewhere at the site. In the sump
area, concentrations were hiqhest from 20 and 30 ft bqs, but were
also found at concentrations greater than 1,000 uq/kq over most of
the A-zone within this area. 1,2-DCP was the volatile contaminant
found at hiqhest concentrations, followed by DBCP, TCP, EDB, and
l,3-DCP. In. the area of the pond, concentrations were hiqhest from
30 to 40 ft bqs, but in qeneral. were found fairly evenly
distributed over the A-zone.
A-zone Groundwater
The same six chemicals found in the subsurface soils plus
chloroform, were found in hiqh concentrations in the groundwater.
EPA's investiqation determined that the total mass of contamination
in the A-zone'qroundwater is siqnificantly larqer than was found in
any other contaminated media at the site. Concentrations for each
of the seven contaminants, except for 1,3-DCP,were found at levels
as hiqh as 1,000 to 100,000 uq/l. The hiqhest concentrations. were
consistently observed in well AMW-2P, located near the sump, and at
well WA-6, which is directly west of the sump, and' at wells AMW-1P,
EPAS-2 and EPAS-3, which are all located near the pond. The
distribution of contaminants was consistent with the locations of
the major source areas and follow a pattern consistent with the
qroundwater flow in the A-zone. In qeneral, contamination was
observed at sliqhtly hiqher levels at wells near the pond when
compared with the wells near the sump; 1,2~DCP was a notable
exception.
l,2-DCP was found to be the 'most wide ranqinq contaminant in
the A-zone qroundwater and was at hiqher concentrations than any
other contaminant. It was found over an area of approximately 5
acres at concentrations.qreater than or equal.to 50 uq/l, or ten
7
-------
times the maximum contaminant level (MCL), and was detected at
concentrations as high as 100,000 ug/l in well WA-6. The other six
contaminants were also found over large portions of the A-zone
groundwater unit, though to lesser extent than 1,2-DCP.
Groundwater in theA-zone flows in a generally southern
direction, with some mounding of the water table observed from the
southwest corner of the site extending south. Water levels
measured during the RI have shown a steady decline in the water
table, probably as a result of the long drought in california. The
saturated thickness of the A-zone groundwater is from 0 to 10 feet.
The hydraulic conductivity in this zone was measured at low levels
. of 10-4 to 10-6 em/s, and from a slug test the groundwater velocity
was estimated at 53 feet/year. Extraction of contaminated A-zone
qroundwater for site remediation is expected to be difficult due to
its low permeability and thinness. Slug test results suggest that
a yield of less than 100 gallons per day can be expected for wells
in this qroundwater unit.
B-zone Groundwater
The B-zone qroundwater is actually composed of a series of
groundwater units. All of the new wells in the B-zone were
installed in the B-2 groundwater unit, located at approximately 170 .
feet bqs. The direction of flow in this unit is to the south, and
the gradient is very flat (0.0004). Permeabilities are much higher
than for the A-zone groundwater. The pump test indicted that wells
could be pumped at 7 gpm for an extended period.
In the B-zone, 1,2-DCP was also observed at levels
significantly higher than any other contaminant and was observed at
least once in. every well. The highest observed concentration of
..1,2-DCP in the B-zone was 1,700 ug/l in well WB2-1, which is
directly south of the site (the MCL for 1,2-DCP is 5 ug/l). Except
for chloroform, the other principal contaminants from the A-zone
groundwater were also' observed in the B-zone, though all at
concentrations below 100 ug/l.
Fate and Mobility
The fate and transport of contaminants at the site are
co~trolled by chemical specific properties and environmental
characteristics and the interaction of these factors. Except for
dinoseb, which is non-volatile, the key site contaminants are all
volatile organic chemicals. All of the contaminants are relatively
mobile in the environment. The volatile contaminants are
transported in the environment as gases or in solution, whereas
dinoseb is transported primarily in solution in the subsurface and
in either solution or adsorbed to soil at the surface. All of the
chemicals are weakly absorbed in soil, although the adsorption of
dinoseb is pH dependent.
8
"- "-----.-..--------------.. -..
. '~-'-'-'-' -~'~_._--'--'_."--'"
-------
. .-..--- -- - _.- -- ,""~_.'--'-' .._..._-
.. - -- '.-' -.-..--...,---. .-..------.-----..----..-.--.- "..
Probably the most important environmental factors influencing
the fate and transport of contaminants at the site are the geology
and the amount of water infiltrating into the A-zone. The site
geology is a heterogeneous mixture of different soil types
characteristic of an alluvial geology typical of that region. This
type of geology results in a high deqree of variability both
vertically and laterally in the permeability of the soil material,
which in turn results in spacial variability in the rate of
contaminant transport at the site. Within the A-zone it was
generally observed that finer grained sediments are more common
below 30 feet until the A-zone qroundwater is encountered. The
base of the A-zone is a thin, mostiy sandy clay unit that retards
downward water movement.
Groundwater flow within the A-zone is very slow as a result of
a low hydraulic conductivity. However, local variations in flow
are expected due to difference in the li tholoqy of this water
bearing unit over the site; higher hydraulic conductivities are
expected at the south-east side of the site where more sand was
observed wi thin this unit. Patterns of contaminant distribution in
the A-zone groundwater are generally consistent with-the direction
of groundwater flow. The exact nature of water movement between
the A-and B-zone is not known. The A-zone is expected to be leaky
and it may be that there are preferential downward flow paths where
the clay layer at the base of the A-zone thins out. At a soil
boring located 900 feet south of the site this clay layer and the
A-zone g~oundwater were not observed.
The infiltration of water into the A-zone is important because
of its impact on contaminant movement in the vadose zone and as a
source for the groundwater in the A-zone. The transport of dinoseb
in particular is directly related to the amount of water
infiltration-because of its high solubility and low volatility.
VI. summary of site Risks
Site risks were formally characterized for the surface soil.
- A screening -risk assessment .was conducted for- these areas to
analyze only the dominant pathways and contaminants that may
significantly contribute to site risk. Risks from ingestion of
contaminated surface soil were characterized for a child and young
adult, and risk from ingestion of contaminated soil in the
construction zone was characterized for an adult worker. Each of
these exposure scenarios exceeded the threshold for deleterious
effects to human health for the maximum detected concentration and
only the child exposure scenario exceeded the threshold for the
average detected concentration.
The other dominant pathway of concern at B&B is potential
exposure from ingestion of- contaminated groundwater either as a
9
-------
result of contamination reaching the city well or .from future use
of the B-zone groundwater; there is no current exposure to
contaminated groundwater above health levels. The screening risk
assessment did not characterize this risk. Instead, concentrations
in groundwater and predicted impacts from the modeling results were
compared to drinking water maximum contaminant levels (MCLs) or
other published health-based levels where MCLs are not available.
contaminant levels in the B-zone groundwater exceeded MCLs in two
wells for both 1, 2-DCP and DBCP. Concentrations in the A-zone
groundwater exceeded MCLs by orders of magnitude; however, because
the A-zone groundwater is not a potential drinking water source,
the concentrations are more important for characterizing the A-zone
groundwater as a contaminant source that threatens the B-zone
groundwater. .
Based on data from the city well closest to the site, B-zone
contamination is not currently impacting drinking water above
health-based levels.
There is no significant ecological risks associated with the
site.
VII. Description of Alternatives
This section provides the specific components of each alternative
and explains the remediation goals and Applicable or Relevant and
Appropriate Requirements (ARARs) as they apply to the specific
alternative.
ARARs
The specific requirements that are applicable or relevant and
appropriate for the Brown & Bryant site can be classified into
chemical-specific regulations and action-spe~ific regulations.
There are no location-specific ARARs at this site. The ARARs at
Brown. & Bryant are: .
state Water Resources Control Board, Resolution 68-16
(Anti-degradation policy).
Environmental Health Standards for the Management of Hazardous
Waste, CCR Title 22, Div. 4.5, Chptr. 15
Article 9, Section 66265.170 - 66265.177 (Containers)
Article 10, Sections 66265.190 - 66265.200 (Tanks)
Article 11, Sections 66265.228 (Surface Impoundments,
Closure)
Underground Injection Control Regulation, 40 CFR Parts 144-147
10
-. -----. -------... ------------.--.------". -~--~--'"
- -..-.. - _.~.. .
-..---. U_-..----.u_.-.--. ..
.. - ... '.--.
-------
,.... .-._....
.. -.-.. --.- ..'- .. ....
-. ...- ,'._---'- ._~..._-------- ..--~- ...--_...-
Althouqh the A-zone is not a potential drinkinq water source,
water re-injected into the A-zone should be treated to be
protective, as required by state Board Resolution 68-16. This'
resolution offers a narrative description of anti-deqradation
policy. EPA believes that reinjection of water containinq
pesticides at the Maximum Contamination Levels (MCLs) under the
Safe Drinkinq Water Act would comply with Resolution 68-16.
Title 22 of the California Code of Regulations (CCR) contains
the state's RCRA-equivalent regulations. Althouqh Brown & Bryant
did not apply for interim status under RCRA, disposal of waste
water into the sump and waste pond at the site constituted RCRA
activities. Therefore, Brown & Bryant should have been classified
as an interim status facility and the State RCRA regulations would
be applicable. Specifically, the waste pond and the sump area are
considered RCRA surface impoundment units and must be closed and
monitored pursuant to 22 CCR 566265.228.
Other RCRA-equivalent requirements for specific treatment
units such as tanks, containers, etc. would be applicable, if
used. The UV/Oxidation (UltraViolet/oxidation) system, and the
Granulated Activated Carbon (GAC) system if used, would be
considered tanks. A variance for the secondary containment
requirements in Title 22 CCR 566266, will be invoked when design'
and placement of the tanks do not pose a substantial hazard to
human health and the environment.
Underqround Injection Control Regulations under the Safe
Drinking Water Act regulate operation of underground injection
wells. 40 CFR 5144.13 exempts actions under a CERCLA response from
the prohibition against reinjection of treated hazardous waste into
or above underqround sources of drinkinq water. Therefore,
. reinjection into the A-zone is permitted. The part of the
regulations '(40 CFR 5144.12) that discuss well construction,
operation and abandonment are relevant and appropriate.
Land disposai restrictions (LDRs) in 22 CCR 566268 et sea are
applicable in certain circumstances whenever there is placement of
soil containinq listed waste on the land. At Brown & Bryant, the
. soil contains listed waste. . However, LDRs are not applicable if
contamination is consolidated within one area of contiguous
contamination. The Brown & Bryant facility is considered one area
of contiquous contamination because the dinoseb surface
contamination is prevalent allover the site without any specific
operational boundaries. Therefore, the surface soil can be
consolidated within the facilit~ without triqqerinq LDRs.
All the proposed action alternatives comply with the ARARs.
11
-------
Remediation Goals
The A-zone groundwater is not a potential drinking water
source; nor is the A-zone soil (excluding surface soil) a direct
ingestion threat. Clean-up standards for these zones are developed
by weighing the cost-effectiveness of cleaning up the zones to
levels where they will no longer be a threat to the B-zone-
groundwater as compared to treating the contamination when it
reaches the B-zone groundwater.
The strictest goal for the A-zone groundwater would be under
the scenario where most of the contamination is captured in the A-
zone and the remaining contamination would not be a threat to the
B-zone groundwater. Two vadose models were run, one to model the
volatile movement through the A-zone groUndwater to the B-zone
groundwater, and one to model the movement of dinoseb. A different
model was chosen for dinoseb because it is non-volatile and water-
soluble, and therefore has different transport characteristics
(refer to the Remedial Investigation Report). Based on these
models, clean-up goals for the A-zone groundwater have been set at
ten and one hundred times the respective MCLs in order. to keep
contamination levels in the B-zone at or below MCLs.
Maximum A-zone Groundwater
contamination Clean-up Level
Chemical Level Range
(ppb) (ppb)
Chloroform 100 1000 - 10,000
1,2-Dibromo""3- 0.2 2 - 20
chloropropane
1,2-Dichloropropane 5 50 - 500
Dinoseb 7 70 .~ 700
Ethylene Dibromide 0.05 ..
0.5 - 5
1,2,3-Trichloropropane 401 400-4000
1 Chron~c (l~fet~me) Health Adv~sory
Again, the ultimate goal at the site is to protect the B-zone
groundwater in the most cost-effective manner. After the remedial
~nvestigation ~f ~e B-zon~ is co~plete and the extraction system
1n the A-zone ~s ~n operat1on, the final remediation levels for
this zone will be determined within the above-stated range that
takes into account the cost-effectiveness of meeting the strictest
goals in the A-zone groundwater clean-up range. The final
remediation leyels will be set in the final ROD.
12
~----- -'----'-----'--~--'--_._'--' --"
. - - -- '. --".. - --. .". n___._. 4. . _. - .
-------
. .. .._------- - . ---.- --.-_._.. 0._..
The subsurface soil contaminant levels were also evaluated
wi th respect to protectinq the B-zone groundwater. The vadose zone
modelling showed that only one contaminant, 1,2-DCP, would pose a
risk to the B-zone if a cap is installed. This contaminant could
be captured in the A-zone qroundwater prior to reaching the B-zone
qroundwater. EPA determined that it would be more cost-effective
to capture the contamination when it reached the A-zone
qroundwater.
The remediation levels for. the surface soil are based on
health calcul~tions considerinq the human inqestion pathway.
Dinoseb was. the only chemical found in the upper 7 feet in
appreciable amounts. Since dinoseb is a systemic toxicant, the
clean-up level was developed based on the most sensitive subgroup,
younq children. The level for dinoseb, 80 milligrams per kilogram,
was developed assuminq a child inqests 0.2 mq/day of soi1 over a
five-year period using calculations for RCRA no-action (Proposed
Subpart S - Federal Reqister Vol. 55, No. 145, July 1990).
Alternatives
On the basis of the results of the remedial investiqation, EPA
identified six alternatives for addressing the soil and A-zone.
qroundwater at Brown & Bryant. Detailed descriptions of these
alternatives are provided in the RI/FS report which is located in
the information repository. Costs for the alternatives are included
in table 1.
EPA believes that controllinq the A-zone groundwater is
essential to protect aqainst further B-zone groundwater
deqradation. . Therefore, all the alternatives, except the no-action
. one, contain an extraction, treatment and reinjection system in
this zone. Althouqh EPA is confident that extractinq con~aminated
water from the A-zope will be effective, there is uncertainty as to
the. number of wells and time frame required for remediation of the
A-zone qroundwater. Based on the current data, EPA estimates that
it will take ten years to remediate the A-zone to the remediation
qoals and up to a five acre area of groundwater will be treated.
EPA intends to phase in the extraction/treatment/reinjection
system to optimize design and control cost. The initial phase of
the clean-up will include a limited number of wells. The limited
system will be monitored to determine extraction effectiveness and
the impact of reinjection on the formation. Expansion of the
initial system will be made after the evaluation of the initial
phase is complete. The time frame for remediation and area of
attainment may chanqe after investigation of the B-zone and
completion of the initial phase of treatment.
The extracted
A-zone
groundwater
will
be
treated
usinq
13
-------
UV/Oxidation. Based on comments received during the public comment
period, EPA will consider during the remedial design, the use of
Granulated Activated Carbon (GAC) as either a post-treatment to the
UV/Oxidation or as a primary treatment. Preliminary costs
estimates indicate that the two systems are comparable in costs.
A detailed cost estimate will be performed in the remedial design.
If-GAC is significantly cheaper, EPA will re-evaluate its decision
on UV/Oxidation. EPA will also evaluate if it is cost-effective to
treat the majority of the chemicals using UV/oxidation and then use
- GAC for low contaminant concentrations where uv/oxidation becomes
less effective~. - - - - -
The extracted groundwater will be treated until it meets
Maximum Contamination Levels established by state and Federal
Regulations. After treatment, the extracted water will be re-
injected into the contaminated portion of the A-zone to help flush
out the remaining chemicals. Excess treated water will be
discharged into the sewer system. The following table lists the
contamination. levels that the treatment system must meet before
reinjection or discharge into sewer system.
Treatment Level
Chemical (ppb) -
Chloroform 1001
1,2-Dibromo-3-chloropropane 0.21
1,2-Dichloropropane 51
Dinoseb 71
Ethylene Dibromide 0.051
1,2,3-Trichloropropane 402
Max1mum Contam1nat10n Level set b Safe Dr1nk1n Water Act
1 Y
2 Chronic (lifetime) Health Advisory
q
Another element common to all the action alternatives is a
multilayered/basic cap combination. Since the state Hazardous
Waste Control Law (HWCL) is applicable and clean closure is not
technically feasible, the RCRA Subtitle C landfill closure
requirements will be implemented. This includes a RCRA Subtitle C
multilayered cap encompassing the sump area, the waste pond area,
and the dinoseb spill area, which is estimated to be 1.2 acres.
The remainder of the property will be covered with a basic cap,
such as asphalt, to minimize infiltration. To assure that the site
remains safe after EPA completes the clean-up, deed restrictions or
other instit~tional controls will be placed on the portion of the
property.hav1ng a RCRA cap to ensure that the cap remains safely
14
.. -.- - -- '-~-~- ----------- -- -------.. . --- .
. -'-''''-----'''-,----.-- -- - .
-------
. .~-_._-_.~_.
..-.. . ---- ~..
. --.. ._-
-... -...--'--..-.-.-..- ..--.--...--........-..
intact and that the soil under the cap remains undisturbed in the
future.
What differentiates the alternatives considered are the
actions proposed for addressing contaminated surface soil and
subsurface soil. The alternatives include either consolidation of
contaminated surface soil under the RCRA cap, treatment of
contaminated sUrface soil and disposal off-site, or treatment of
contaminated surface soil and disposal on-site. Also, the
subsurface soil may be treated, under alternatives 5 and 6, using
soil vapor extraction depending on the added value and cost of this
additional treatment. . . .
The selected remedy will undergo a review every five years to
insure protection of human health and the environment as required
by EPA when waste is left in place.
Alternative 1 - No Action
Superfund regulations require EPA to include consideration of
a no action alternative for comparison with the other alternatives
(#2 - #6). EPA presumes that even if the no action alternative was
selected, site monitoring would continue. .
The Selected Alternative -Alternative 2 -
Consolidation of contaminated Soil, RCRA/Basic
Cap, Extraction and Treatment of A-zone
Groundwater
. Under this alternative, like all the action alternatives, a
RCRA SubtitleC cap will be placed on the southern 1.2 acres and a
basic cap will be placed on the remaining property. Deed
restrictions will be recorded to assure the cap remains intact.
This alternative varies from the other alternatives in its handling
of soil containing dinoseb in excess of the remediation level of 80
mgfkg. Approximately 70 cubic yards will be consolidated from the
area outside the RCRA Subtitle C cap to that area. In addition,
approximately 570 cubic yarc;ls of contaminated surface soil and
48,000 cubic yards of subsurface contaminated soil currently in the
southern portion of the site will be covered by the RCRA Subtitle
C cap. Included in this alternative, as well as all the other
action alternatives, is an injection and extraction system that
will flush the A-zone groundwater and treat it using UV/Oxidation
prior to reinjection. It is estimated that ten pore volumes
(approximately 35 million gallons) will be needed to reach the
remediation goals.
15
-------
Alternative 3 - Off-site Treatment of Some Surface Soil,
RCRA/Basic Cap, Extraction and Treatment of A-zone
Groundwater
Alternative #3 is similar to alternative #2 because 'it also
includes a,RCRA cap on the southern portion of site containing the
sump and' the waste pond," a, basic cap on the remaining property,
deed restrictions, and an injection and extraction system that will
flush the A-zone groundwater and treat the extracted water prior to
reinjection. This alternative differs from alternative #2 in that
the dinoseb contaminated surface soil in the portion of the site
not covered by a'RCRA subtitle C cap would be excavated and treated
off-site rather than consolidated on-site. "
Alternative 4 - On-site Treatment of all Surface Soils,
RCRA/Basic Cap, Extraction and Treatment of A-zone
Groundwater
Alternative #4 is also similar to alternative #2. However,
instead of consolidation or off-site treatment of some of the soil
as envisioned in the earlier alternatives, alternative #4 will
treat on-site all surface soil with ,dinoseb in excess of
health-based standards by soil washing. The treated soil will then
be replaced back on-site. The volume to be treated is estimated at
570 cubic yards. All other aspects would be the same as described
in Alternative #2, including a RCRA subtitle C cap on the southern
portion of site containing the sump and the waste pond, a basic cap
on the remaining property, deed restrictions, and an injection and
extraction system that will flush the A-zone groundwater and treat
the extracted water prior to reinjection.
Alt~rnative 5 - Off-site Treatment of Some Surface Soil,
RCRA/Basic Cap, In-situ Treatment of Deeper Soils,
Extraction and Treatment of A-zone Groundwater
Alternative #5 is identical to alternative #3, except it
includes an additional treatment technology, Soil Vapor Extraction,
to remove volatile compounds in deeper soil (25 to 40 feet). ,All
other aspects would be the same as described in Al ternati ve #3,
including off-site treatment-and disposal of a small quantity of
contaminated surface soil, a RCRA Subtitle C cap on the southern
portion of site containing the sump and the waste pond, a basic cap
on the remaining property, deed restrictions, and an injection and
extraction system that will flush the A-zone groundwater and treat
the extracted water prior to reinjection.
16
~-, ..-.~_.__... - ~ - -.-------.... ...."
--- ----_..--- ---
. -.-. _.. . . .
-------
... --- .... _h - .
.-.----. ..,-~ 0- ---------_.~-_._-----_..~ .,-, -..
Alternative 6 - On-site Treatment of all Surface Soils,
RCRA/Basic Cap, In-situ Treatment of Deeper Soils,
Extraction and Treatment of A-zone Groundwater
Alternative #6 is identical to alternative #4, except it
includes an additional treatment technology, Soil Vapor Extraction,
to remove volatile compounds in deeper soil (25 to 40 feet). All
other aspects would .be the same as described in Alternative #4,
including on-site treatment of all contaminated surface soil, a
RCRA cap on the southern portion of site containing the sump and
the waste pond, a. basic cap on the remaining property, deed
restrictions, . and an injection and extraction system that. will
flush the A-zone groundwater and treat the extracted water prior to
reinjection.
VIII. summary of Comparative Analysis
Overall Protection of Human Health and the Environment
The overall protection of human health and the environment
criterion assesses each alternative to determine its effectiveness
in reducing risks.at the Site. .
Alternative 1 offers no protection .other than natural
degradation and attenuation. All the other alternatives (2,3,4,5
& 6) contain a technology to remove contamination from the A~zone
groundwater in order to protect the B-zone groundwater. Without
removing the contamination in the A-zone groundwater, the
contamination would need to be captured in the B-zone groundwater.
Alternatives"5 and 6 add an incremental protection by treating the
A-zone soil. This incremental protection can also be achieved in
alternatives "2, 3 and 4 by capturing the contamination when it
reaches the A-zone groundwater. .
Alternative 2 eliminates the exposure to highly contaminated
soil by placing a RCRA cap over the southern t,hird after all
contaminated soil. had been consolidated beneath the cap.
Alternatives 3, 4, 5 and 6 also reduces the potential exposUre to
surface soils by either treating the highly contaminated surface
soils before placing a RCRA/basic cap combination or placing a RCRA
cap over the hot-spots. Alternatives 4 and 6 removes and treats
the most highly contaminated soil.
Therefore, alternative 2 through 6 are protective of human
health and the environment.
17
-------
Comnliance with ARARs
All alternatives, except alternative #1, will comply with the
substantive requirements of the identified ARARs.
Lonq-term Effectiveness and Permanence
Alternatives are assessed for the long-term effectiveness and
permanence they afford, along with the degree of certainty that
the alternative will be successful.. The alternatives that~emove
and treat the greatest amount of contamination will be the most
permanent (alternative 5 & 6). However, all alternatives, except
no action, treat the largest source of continuing contamination,
the A-zone groundwater, and leave some contamination in the soil
behind. Alternatives 2, 3 and 4, also offer a high degree of long-
term effectiveness because the surface soil threat is addressed by
removal and treatment or containment of the contamination. Also in
al ternati ves 2, 3 and 4, the A-zone soils are controlled by
limiting the movement of contamination in this zone. The long-term
effectiveness of the alternatives that leave. . levels of
contamination exceeding health-based levels beneath the RCRA-cap
(alternatives 2, 3 & 5) is determined by the long-term maintenance
of the. cap.
All the alternatives that treat the surface soil (alterna.tives
3, 4, 5, & 6) have a strong probability of success because the
treatment has already been demonstrated on-site to be successful.
It is uncertain how the water-bearing zone will respond to the
horizontal flushing and extraction component of alternatives 2, 3,
4, 5, & 6. It is anticipated that a significant volume of
contamination can be removed. Finally, the effectiveness of the
soil vapor extraction component in alternatives 5 and 6 is
uncertain due. to the heterogeneity of the soil layers.
Reduction of Toxicitv. Mobilitv.or Volume Throuqh Treatment
The alternatives are assessed on the degree to which they
employ recycling or treatment that reduces toxicity, mobility or
volume, especially with respect to the principal threats at the
site. There are two principal threats at the site, the threat the
surface soils pose to human exposure and the threat the A-zone
groundwater poses to the B-zone groundwater, which is the first
potential drinking water source at the site.
All alternatives except no action, alternative 1, actively
address the principal threat to the B-zone groundwater by hori-
zontal flushing and extracting contaminated A-zone groundwater. The
extracted water will be treated to destroy the contaminants. The
~egree of reduction of the contaminants in the A-zone groundwater
~s unknown. due to the uncertainty of the flushing/extraction
18
-. .....-.__._..._..._-~_.__.__._-_._.__...---~.__. - ...
-------
- . ...- -_.. -_...- _. -..
." -" -----.-..------ - .,,--.- -,--- ..---. .....
process in a geologic formation containing clays. It is expected
that the contaminated groundwater and the readily removable
contamination on the soils in the saturated zone can be removed.
Alternatives 3, 4, Sand 6, all actively reduce the volume to
surface soil contamination by excavating the contamination and
either, treating. it on-site and returning the treated soil to the
site, or treating a portion off-site and disposing of it at a
hazardous waste landfill. Alternative 3 & S treat a relatively
small volume of contaminated soil compared to the volume left
beneath the RCRA cap. Alternatives 2, 3 & S reduces the toxicity
of the surface. soil contamination by eliminating potential exposure
by installing a RCRA cap. The cap also reduces mobility of the
contamination ..
Alternatives 2, 3 and 4 reduce the mobility of the contami-
nation in the A-zone soils by reducing infiltration. The A-zone
soil contamination is a minor threat to B-zone groundwater.
A1 ternati ves Sand 6 acti vely reduce the volume of volatiles
contamination in the A-zone soil by removal wi th 50il Vapor
Extraction (SVE). The degree of expected reduction in volume by
active treatment is difficult to judge because of the heterogeneity
of the geological formations in the A-zone soils. SVE will not
remove dinoseb.
Short-term Effectiveness
Short-term effectiveness assesses for each al ternati ve the
short-term.risks to workers and the community during implementation
of an alternative, potential short-term environmental impacts of
the alternative and the time until protection from any short-term
risk is achieved.
The a1 ternati ves that propose excavation of contaminated
surface soils (Alternative 2, 3, 4, Sand 6) may pose a short-term
fugitive dust risk to workers and the community. Dust control
measures should be implemented. Alternatives 3 and S contain off-
site transportation of hazardous waste which pose a.short-term risk
to the communities en route. Alternative 2 requires the minimal
. amount of soil handling; therefore, it poses the least significant
short-term risk.
The flushing/extraction process for the A-zone groundwater
proposed in alternatives 2, 3, 4, Sand 6 poses no short-term risk
to the community and the workers. It is estimated that it will
take ten years to remediate the A-zone groundwater; however,
neither the A-zone groundwater (which is not a potential drinking
water source) nor the A-zone soil pose an immediate risk to the
community or to the workers.
19
-------
~n-situ installation and operation of SVE in alternatives 5
and 6 requires little handling of contaminated soils, and thereby
limits the risk of e~osure to workers and the public.
Imnlementability
The ease or difficulty of implementing the alternatives are
assessed with respect to technical feasibility, administrative
feasibility and. availabiiity of services. All the alternatives
that address surface 50il contamination use either standard, proven
technologies (alternative 2, capping and alternatives 3 & 5, off-
site treatment and disposal, and capping), or an innovative
technology (alternatives 4 & 6, soil washing and alternatives 5 &
6, soil vapor extraction). Soil washing was proven successful at
the site by an EPA removal action. All these technologies are
implementable.
The horizontal flushing/extraction procedure proposed in
alternatives 2, 3, 4, 5 & 6, consists of installing extraction and
injection wells. The installation of these wells are standard
procedures. However, operation details such as recovery of .
injected fluid will. require adjustment during the operation and may
require a longer time for remediation.
The technical feasibility
dependent on the ability to pull
silt layers in the soil profile.
feet in depth contain the highest
of soil vapor extraction is
air through the silty sand and
These layers between 25 and 35
concentrations of 1,2-CCP.
~
Cost estimates for the six alternatives are presented in Table
1. The costs for the action alternative range from $9,193,000 to
$10,923,000. "
state Accentance
The State of California through its Department of' Toxic
Substances Control, has been active participants throughout the
RIfFS process. The State has been interested in protecting all he
waters of California, and as a consequence, has been most
interested in formulation of alternatives that protect the B-zone
groundwater. Only alternative #1 does not actively address the
source of contamination for the B-zone groundwater. The State
would like to see an early action pump and treat on the B-zone
groundwater. .
The State concurs with the selected remedy.
20
.. "--'.. "..
-"-~-" ...- ... "-__H__- --...- .~. . .'. .
.~-- -.-. --.. -.
-------
~--~~-------L_~_.~-_..
. -.... --. - ..-.. -..- ..
-.-.. .-" ---_.. '---"----.'----'.-----'-------'"
Communitv Acceptance
Public comments on the proposed plan are presented in the
"Responsiveness Summary" of this ROD. The comments received from
the Arvin";Edison Water Storaqe District and the city of Arvin
express concern with the cost of the selected alternative. Two
other alternatives addressinq the contamination were suqqested by
the Arvin-Edison Water storaqe District, none of which included
treatinq the water after extraction. Given that 1) the cost for
the Water Storaqe District's proposal would probably not be
siqnificant1y different from EPA's proposal, 2) the proposed
alternative is not protective of human health and the environment,
and 3) there is the statutory preference for treatment, EPA
continues to prefer its selected alternative. The alternative
selected has what EPA believes is the minimum amount of removal and
treatment of contamination necessary to protect human health and
the environment and comply with ARARs.
IX.
Selected Remedy
Alternative #2 is the selected remedy for the first operable
unit at the site. The qoa1 of this remedial action is to prevent
exposure to soil contaminated above health-based levels and to
control the source of contamination to the B-zone qroundwater.
Based on the information obtained durinq the remedial investiqation
and analysis of all the remedial alternatives, EPA believes the
selected remedy will be able to achieve this qoal. Specifically
the selected remedy is as follows: .
- Move contaminated surface soil from the area not included in
the RCRA' Subtitle C cap to the waste pond and adjacent area
where the RCRA Subtitle C cap will be placed. Remaininq soil
will be tested to confirm that all surface soil containinq .
pesticides above heal th-based levels has been moved.. In
addition, the surface soil surroundinq the site will be tested
to assure that levels of contamination off-site do not exceed
health-based levels. If any soil is found exceedinq health-
based levels, that soil will be consolidated under the cap.
- After consolidation of the contaminated soil to the southern
portion of the site, the northern and western portion of the
site will be reqraded and covered with a basic cap, such as
asphalt. The purpose of the basic cap is to control storm
water runoff. . This portion of the site will be considered
clean. .
Institutional controls will be implemented which will
consist of deed restrictions prec1udinq residential use of the
21
-------
site and assuring that the RCRA cap area is maintained.
- ARCRA Subtitle C cap will be installed on the waste pond,
sump area, dinoseb spill area and adjacent areas. The cap
will be designed to prevent exposure and minimize
infiltration.
- All capped.areas will be maintained as appropriate.
- The A-zone groundwater will be extracted. After extraction,
the water will be treated using UV/Oxidation, and/or possibly,
GAC and then reinjected into the A-zone groundwater.
Reinjection will be carefully monitored to ensure control of
the extracted water. If the water required for reinjection is
less than the water produced during treatment, the additional
treated water will be discharged to the sewer system.
- The extraction/reinjection system will be phased in to allow
for optimal design of the system. Reinjection rates will be
monitored to prevent build-up of excess head of water that
might spread contamination further. The number of
extraction/reinjection wells, location of these wells, and
extraction and reinjection rates for the initial phase will be
establishing during the remedial desiqn. Expansion of the
system will be considered after evaluating the effectiveness
of the initial system.
- After completion of the remedial investigation of the second
operable unit and the extraction and treatment system has been
in operation long enough to estimate rate of contamination
removal,. an analysis of the cost-effectiveness of further A-
zone treatment versus capturing in the B-zone shall be made.
- As required by the State Hazardous Substances Control Act,
the period of qroundwater monitoring will not be less than
thirty. years.
22
---.-----.----------------. ---'.'
"--'-".'_._----'-~- ~-.
-------
~ - . .. . .~.... - n. ~
-.-.' --.---. .._.,_...~----~--_. ..-------~... ,..' ~..
x. statutory Determinations
Under CERCLA, EPA's primary responsibility at Superfund sites
is to undertake remedial actions that achieve adequate protection
of human health and the environment. Additionally, the selected
remedial action must comply with ARARs established under federal
and state environmental laws unless a statutory waiver is
justified. The selected remedy also must be cost-effective and
utilize permanent solutions and alternative treatment technologies
to the maximum extent practicable.. ~inally, the statute includes
a preference for remedies that permanently and significantly
reduces the volume, toxicity, or mobility of hazardous waste as
their principal element. The following sections discuss how the
selected remedy meets these statutory requirements and preferences.
Protection of Human Health and the Environment
Threats to human health and the environment include ingestion
and contact with contaminated soil and potential exposure to
contaminated groundwater. The selected remedy partially addresses
the threat of exposure to contaminated groundwater by controlling
the greatest source of contamination, the A-zone groundwater. This
source of contamination will be extracted and treated to maximum
contaminant levels then re-injected into the A-zone groundwater. . .
The A-zone groundwater will be treated to levels that no longer
pose a threat to the deeper groundwater; or if the contamination
can be more cost-effectively extracted from the B-zone, the A-zone
groundwater will be treated to levels that are easily and quickly
achieved.
The selected remedy addresses the threat of exposure to
contaminated. soils by consolidating all contaminated soil in one
. portion of the site, capping this portion with a high-quality, RCRA
Cap, then implementing institutional controls.
Comoliance with Applicable or Relevant and Appropriate Reauirements .
(ARARs) . ..
The selected remedy complies with all federal and state ARARs
identified for the site. ARARs are discussed in more detail in the
Description of Alternatives section. .
Cost-Effective
While compiling the alternatives, EPA determined that source
control was paramount to the overall clean-up of the site. If the
A-zone groundwater could be remediated before the contamination
reaches the B-zone, then the overall site clean-up costs should be
greatly reduced. Extraction of a large JIlass _of the chemicals early
23
-------
in the clean-up process should also reduce the time needed to clean
up the site. In general, the less time needed to clean up, the
less it costs. Because all the action alternatives contain this
approach to treatment, the costs are comparable.
With respect to the subsurface soil contamination, the
selected remedy addresses it by reducing contaminant migration with
a cap and capturing any contamination in the A-zone groundwater if
it miqrates there. Since the treatment would already be in place,
the cost for treating the additional contamination is nominal. It
is the most cost-effective of the subsurface soil alternative~.
The selected remedy for the surface soil provides comparable
overall effectiveness at the lowest cost.
Use of Permanent Solutions and Alternative Treatment Technoloqies
to the Maximum Extent Practicable .
The selected remedy uses permanent solutions and treatment
technologies for the principal source of contamination to the
maximum extent practicable. The selected qroundwater remedy will
result in a reduction of volume, toxicity and mobility through
groundwater extraction, treatment. and reinjection of the
groundwater. Continued monitoring will be conducted to ensure that
the groundwater remedy is protective of human heal th and the
environment. UV/Oxidation system is considered an alternative
treatment technoloqy.
While the selected remedy does not offer as high degree of
long-term effectiveness and permanence for subsurface and surface
soil as some o~ the alternatives that proposed aggressive treatment
of these zones, it will significantly reduce the hazards associated
with the contaminated soi1. Since the remaining soil contamination
will be capped, the impact to huma~ health and the environment
will be minimal as long as the cap is properly maint~ined." .
Preference for Treatment as a PrincinalElement
By treating the A-zone qroun~water, the selected remedy
addresses the principal threat posed by the site through the use of
treatment technoloqy. Therefore, the statutory preference for
remedies that employ treatment as a principal element. is satisfied.
24
. .'~"-'_.'----.'-'--'-"--'--' -'_.4_._~~..
'- .-_.. - -- ,.-.- - --- ..
-...-,.,-,.- .
-------
-..-. ..... --'--'.~--'_."-'''-_.
. . -- .. .--... - --- .-."--------'. ._--,- ---
-
XI. Documentation of significant Changes
The only siqnificant chanqe. to the Brown & Bryant first
operable unit (OU) interim remedy proposed in the Proposed Plan
fact sheet dated June, 1993, involves the possible use of liquid
phase GAC treatment technoloqy in addi tion to or in instead of
UV/Oxidation.
As a result of comments received durinq the public comment
period, EPA preliminarily evaluated the use of a Granulated
Activated Carbon (GAC) as post-treatment to UV/Oxidation or for
primary treatment. Preliminary costo estimates show that the GAC
and UV/Oxidation are comparable in cost. However, a more detailed
cost comparison will be done in the remedial desiqn. In addition
to cost, EPA is required to consider other factors when selecting
the alternative. CERCLA S121(b) states "Remedial actions in which
treatment permanently and siqnificantly reduces the volume,
toxicity or mobility of the hazardous substances .. as a principal
element, are to be preferred over remedial actions not involving
such treatment." In addition, "The President shall select a
remedial action that is protective of human health and the
environment, that is cost-effective and that utilizes permanent
solutions and alternative treatment technoloqies or resource
recovery technologies to the maximum extent practicable."
A'fter the detailed cost comparison is completed, EPA will'
evaluate the selected technoloqy, UV/Oxidation, with respect to
cost as well as other statutory preferences. EPA may modify the
use of UV/Oxidation to include GAC. .
The impact of this potential change is cost. The time to
complete the project, the clean-up goals and the reinjection levels
will remain unchanged. .
25
-------
.- .... ...-.-.--- -....---. . .
--_.. _.. - ._~~.:.\ 1.\
o
~
<3
<
no
. .....---....-
:z:
a:
w
X
t-
a
o
tn
w
x:
o
:z:
<
~
0-
o .
ROAD
~t-
mUJ
a:UJ
wa:
01-
tn.
ARVIN,CA
Brown & Bryant, Arvin Calif.
Figure
LOCATION MAP
BOULEVARD
....-... .-_.--._u.,._-_.- ,,"""
l
I
I
.- --1---
\
N
+
1
"_'_n..'.--"'''''''' ...
iJ
-------
. ,,-- .. .- -_."--_._'~'._~ ..
.-- .. ~.-_.
Brown
o.nd
Bryant
\Jare.
house
4J
Q.)
.Q.)
.~
4J
CI)
:>..
.Q
~
(l)
o
.
Tanks
. .
. .
.
C!J
\Jash Pads
..c::L- --
rt1d SUI"p Arf:;)
Lh /
, "'_/
/-,
,
/ I
( I
,_/
~Dinoseb .a.rea.
o . 100
1 . I
Scale In Feet
.~
Brown & Bryant, Arvin Calif.
Fiqu%;"e 2
SITE MAP
-------
7 Ft.
below
ground
surface
65 Ft.
85Ft
150 Fl
? Ft.
.350 Fl
(aPprox.)
.-.--."--.-. -.4.-,..._.
- -.---------.-.---------.---.----.-----.
.I
B-Zone Groundwater
-------
plge 1 00
COST OF ALTERNA11V1'.S
A". lIem Up-front Coeb (S 0(0) ANNal Coli (S (00) ....... Wodh (SOOO)
1 On.going Monitoring (30 yn) ..- 50 610
2 RCRA/Ba.ic Clp (meintenence - 30 yrs) 1,072 50 1,632
Consolidation of Hoe-spotl .2 10
Horizonta' Flulhing &: Extl8ction of A-zone Oroundwlter 1,471 330 3,67'
UV/Oxidltion of Extl8cted Wlter 492 4711 3,700
limited On-going Monitorin, (30 yrs). -. 16 1110
rorAL 3,047 814 9,193
3 Exclvllion of Surf Ice Soil Hoe.spotl in Northern 2/3 lite .2 - 12
Off-lite Treltment and Diapoll' 225 - 225
RCRAlBa.ic Cap (meintenence . 30 yn) 1,072 50 1,632
Horizonta' flushing &: BJI1l8ction of A-zone aroundwater ',471 330 3,671
UVfOxidltion of BJI1racled Wlter 492 4711 3,700
limited On-golng Moniloring (30 1n) -- 16 1110
rorAL 3,2n 174 9,420
-------
plge 2 of 3
COST OF ALTERNATIVES
Ah. hem Up-noN COIb (S (00) ~ ColI (S 0(0) ..... Woc\h (S 0(0)
4 Excavalion of All Surf Ice Soil "01-'1'011 140 --. 140
Soil Wllhing 870 -- 870
Additiollli UVIOxidalion 80 - 80
RCRAlBalic Cap (maintelllnce - 30 yn) 1,072 50 1,632
Horironlal Flushing &. Exlnclion of A-zone Oroundwller 1.471 330 3,671
UVlOxidllion"Of A.zone Groundwaler 492 478 3,700
Umiled On-,oing Moniloring (3O-year) ..- 16 180
1UfAL 4,12.5 874 10,273
5 Excavalion of Surface Soil Hot-1poI1 in Northern 213 Ii Ie 12 - 12
Off-lile Trealment and DiIpOlllI 225 - 215
RCRAlBallc Cap (mainlelllnce . 30 yn) 1,072 50 1,632
Horizontal flushing & Extnction of A-zone Groundwaler 1,471 330 3,671
UVIOxidalion of Extracted Waler 492 478 3,700
SVB orsump and Waite Pond Area. (I ,ear) 550 100 650
Umiled On'SOlns Monitorln, (30 ,n) -. 16 180
rofAL 3.821 974 10,010
-------
p.ge 30f3
COST OF ALTERNATIVES
Alt. lIem Up-front Coe-. (S 0(0) Alnlal COlt (S 000) ...... Wooh (S 000)
6 Exc,v'lion of All Surf.ce Soil HoHpoIs 140 ..- I~
Soil Washing 870 ... 870
Addilional UV/Oxidllion 80 - 80
RCRAlBasic C.p (maintenance. 30 yn) 1,072 50 1,632
Horizonl.1 Flushing &; Exlnclion of A-zone Oroundwller 1,471 330 3,671
UVlOxidalion of EXlncled A-zone Oroundwller 492 478 3,700
SVE of Sump .nd Waste Pond Arc.. ( I year) 550 100 650
Umiled On-going Monitoring (IO-year) .00 16 180
T(JJ'AL 4,67~ 974 10,m
-------
... .- ..._...~, - ,_-_,.~,_'J--,~..- _.._----_.._.._._..._----_..._~.._.,....
PART III - RESPONSrvENESS SUMMARY
BR~WN , BRYANT, ARVIN FACILITY
ARVIN, CALIFORNIA
This section provides EPA's response to comments received on
the proposed plan for cleanup of the first operable. unit at the
Brown & Bryant site. The responsiveness summary includes two
parts. The first part is a summary of major issues and concerns
raised by the comments and a summary of EPA/s response. The
second part includes verbatim each comment received and EPA's
detailed response to each comment. Comments from the local.
community are included from paqe 29 to 34, comments from the
State of California are included from paqe 34 to 39, and comments
from the Potentially Responsible Parties (or their
representatives) are included from paqe 39 to 83. In the event
of ~ny conflict or ambiquity between the two parts, refer to " the
detailed analysis in the second part.
SUMMARY OF RESPONSE TO COMMENTERS' MAJOR ISSUES AND CONCERNS
Summarized below are EPA'sresponse to major issues and
concerns from the public in and around Arvin, the state of
California, and Potentially Responsible Parties for the site.
An alternative cleanup proposal was received from the Arvin-
Edison Water Storaqe District, and was supported by the Arvin
City Council and State Assemblyman Jim Costa. The proposal calls
for dilutinq the contaminated A-zone qroundwater with irriqation .
water used for local irriqation or with irriqation water two and
a half miles away in the Arvin-Edison South Canal, a larqer water
volume than available in local irriqation canals. The commenter
estimates that its proposal would cost approximately $100,000;
the letter qives special emphasis to the cost savinq, which, if
true, are substantial. The proposal emphasizes that the
contaminants in question have for years been used at much'hiqher
concentrations on local farmlands. . Also, the proposal requests a
variance from any requlations that may preclude its approval.
EPA's response to this proposal focused on three concerns:
the actual cost of the proposal, the possibility for dilution of
the contaminants to safe levels, and specific leqal.issues
regarding the proposal.
As to the actual cost of this proposal, EPA believes that
the estimated cost does not take into account some important cost
factors. Specifically, the cost estimate does not include the
cost of extractinq the contaminated qroundwater, cappinq the
site, and monitorinq and maintenance for these two remedy
components; all of these costs are still required under the
commenter's proposal. EPA estimates these costs at approximately
$5.5 million." EPA also calculated rouqhcost estimates for'
26
-------
construction of a two and a half mile pipeline and maintenance of
the pipeline for ten years. EPA's cost estimate used a simple
PVC pipe which would not meet hazardous waste handling laws and
it excluded right-of-way costs. With these cost limitations, EPA
estimated the commenter's proposal would cost at a minimum, $6.2
million,.. or approximately $3 million less than the EPA preferred
alternative. . .
EPA also evaluated whether the flow in the south canal would
provide enough water to dilute the contamination to levels equal
to the drinking. water maximum contaminant levels (MCLs). EPA's
analysis found that at high flow rates the canal does not provide
enough water to meet the MCL for ethylene dibromide (EDB), and at
low flow rates, such as during the winter, other contaminants
would not be diluted enough.
Finally, there are requlatory requirements in Federal and
state laws that would prohibit this disposal option. EPA would
not be able to waive these requlations.
EPA also received a request from the Arvin community
services District, which supplies the city's drinking water, for
EPA to consider replacing city well #1, located near the site.
EPA will consider this proposal as part of the second operable
unit RI/FS for the site. This investigation will give a better
picture as to the threat contamination may pose to the city well.
However, EPA's current policy prefers treatment of contaminated
drinking water at the well head rather than relocating the well.
Finally, EPA received comments from the Sierra Club, Kern-
Kaweah Chapter. Its comments consisted of three questions. One
question regarding the potential impact of storms or earthquakes
on the cap, a second regarding the future uses available for the
site, and a third regarding the safety of diluted dinoseb. EPA
responded to each of these questions in the detailed response to
comments. .
Comments were received from one State agency, the Department
of Health Services, which is ~urrently preparing a fOllow-up to a
1989 health assessment. Most of its comments concern the
possibility that EPA did not adequately evaluate other areas of
the site or other chemicals in determining the risks associated
with the site. Its concerns involve off-site contamination at
the surface and subsurface, on-site waste piles, and the
evaluation of risks from probable human carcinogens at the site.
EPA believes that a significant risk at the site has not been
overlooked; however, detailed analysis of all potential
contaminant pathways was intentionally limited in the Remedial
Investigation/Feasibility Study (RI/FS) Report. EPA focused on
~e principal threat~ at ~h7 site in its RI/FS Report. Largely
J.n an effort to. use J.tslJ.mJ.ted resources as efficiently as
possible, EPA limited the level of investigation it put into the
27
. - ---.._- -- -" - .---.-. .~~_.--._-- _.. -'---- ,. ..._-_. ----.- - - -.---._'
u -_. .-. .....-
- _..~_.,. -~'.-.._-~ '"--_.
-------
-. ~... -. - ...,. ... .-. "--
.... ._-~._.....-~-_...__.--.._----_._.....__.._._-_. -. ~...
-'
pathways that showed little or no probable concern, or when the
pathway did not alter the proposed remedy or the cleanup
standards. .
The majority of comments were received from Potentially
Responsible Parties (PRPs) for the site or their representatives.
Comments on the Remedial Investigation (RI) Report focused in
particular on EPA's modeling of contamination in the subsurface
soil, and EPA's analysis of si~egeoloqy and hydroloqy.
. Both comment letters contend that EPA overstated the
potential impact from soil contamination as a result of .
conservative and/or oversimplified assumptions made in modeling
the contaminant transport. The comments contend that EPA arrives
at more stringent cleanup requirements than are necessary as a
result of its analysis. In response, EPA points out that the
remedy EPA selected does not call for treatment of contaminated
soil; therefore, the conclusion that EPA established overly
stringent cleanup requirements for soil is false. EPA's only
treatment remedy for the soil is to cap the contaminated soil.
The cap is also required by RCRA requlations.
Comments on the site geoloqy and hydroloqy generally contend
that. EPA's analysis is deficient or incomplete. EPA recoqnizes
that the geoloqy and hYdroloqy at Brown & Bryant is complicated.
However, with data available from approximately 100 soil borings
and 25 groundwater wells, EPA believes that the data and analysis
in the RI Report is adequate for the purpose of selecting a
remedy. EPA recognizes that additional information or analysis
is needed in order to complete the desiqn of the remedy. This
analysis will occur in the design phase of remediation.
Comments on the Feasibility Study (FS) Report focused on a
number of potential factors that may come into play in the
ultimate design of the remedy. The comments generally raise
specific concerns that will be addressed during the remedial
design, but do not need detailed analysis at this point.
However, in recognition of the number of complicated variables
that will go into the design of. the remedy, EPA has proposed to
phase in the remedy so that the design can be perfected as it is
implemented.
One comment on the FS Report questioned whether granulated
activated carbon (GAC) might be a more cost effective technoloqy
for treatment of the A-zone groundwater. In respon~e, EPA
reviewed cost estimates for GAC and determined that the costs are
comparable to UV/Oxidation. As a result; EPA will conduct a more
detailed analysis of GAC during the remedial desiqn phase and may
also consider the use of GAC as a post-treatment to UV/Oxidation
Finally, the PRPs commented that action on the A-zone should
be delayed until the B-zone RI/FS is complete so that the entire
28
-------
site can be evaluated together. EPA is opposed to delaying
action on this first operable unit because it addresses the
portions of the site were the most contamination occur~. ~PA has
given priority to this action in order that the contam~nat~on
problem does not spread to a wider area and potentially result in
even greater cleanup costs. .
DETAILED RESPONSE TO COMMENTS
Comments from the Arvin-Edison Water storaae District:
1. The Arvin-Edison Water Storage District consists of 132,000
acres in the Southeastern portion of the San Joaquin Valley. The
community of Arvin lies within the District.
We have followed EPA's activities related to the Brown and Bryant
Superfund site in Arvin, the last of which was a public meeting
held in Arvin, July 6, 1993. In addition, we have reviewed in
detail the material handed out at that meeting and submit the
following as a much less excensive means of disposinq of the
contaminated water. We believe that our suggested alternatives
could be accomplished by an expenditure of approximately one
hundred thousand dollars, a fraction of the ten million dollars
estimated in your outline.
ALTERNATIVE A.
Pump the contaminated water, transport it east and south to the
adjacent farming areas and commingle it with irrigation water
applied to crops at dilution rates which will provide assurances
that it will~ot present a health hazard or pontaminate the
underlying aquifers. .
ALTERNATIVE B.
Pump the contaminated water, transport it east approximately two
and a half miles to the Arvin-Edison South Canal and commingle it
with canal water at much greater dilution rates than possible
under Alternative A. ..
We understand from discussions with EPA staff that dilution as a
means of dissipating contaminants is not acceptable to EPA. In
this regard, we would like to emphasize that for many years the
materials listed as contaminants at the Brown and Bryant site
were widely used in all farming areas at much higher
concentrations than would be present in a diluted form if one or
both of the methods described herein are used, yet they cannot be
detected today. Further, we believe that if EPA regulations
preclude the use of the dilution method, it is logical that some
sort of variance could be obtained.
2'
.._..._-.~.._---~-.._-_._-------- _._. .._. - --, .... ._-
-------
. -~. . --_.. .- .0___...
.. . "'-.'-- -.---....----.
.. -_._--~----- ----.-..
What is outlined herein is intended to be conceptual and will
require more study before it can be properly evaluated. Among
other things, studies will need to confirm that the commingling
supply would not be harmful to crops which we are confident can
be demonstrated. In addition, EPA or others would be required to
indemnify those involved in the disposal process from any
liability arising out of. the use of the commingled water.
If the above conditions can be met, Arvin-Edison, would like to
ass'ist in restoring the Brown and Bryant superfund site if one or
both of the concepts outlined above would be acceptable to EPA.
District activity could also possibly include an alternative to
RCRA's cap.. .
In summary, we believe that Arvin-Edison has a vested interest in
this activity because of its responsibility in protecting the
groundwater underlying the District. Further, we believe that
the District Engineering and Field Staff, by virtue of its
training experience, and local knowledge has the ability to
assist in the. clean-up of the Brown and Bryant Superfund Site a
substantiallv less cost than that estimated by EPA.
Response: This "proposal vas evaluated vith respect to technical
and legal considerations. The technical considerations were cost
and protection of human health and theenvironment~ BPA
evaluated whether the flow in the canal would ~e sufficient to
dilute the contaminated water to or below the maximum
contamination levels. Even under the most favorable scenario,
high summer flows and expected average concentrations, £DB would
never be diluted to KCLs. Two other chemicals, DBCP and 1,2~DCP,
would ~e sufficiently diluted in high flows ~ut not during the
low flows in the winter.
The aspects of EPA's alternative that would change under the
commenter's proposal is the treatment system (cost $650,000) and
the operation and maintenance of such system (cost $3.62
million). The costs in common to ~oth EPA'. alternative and the
commenter's proposal, the extraction system, the cap, monitoring
and 0&., would remain at $5.96 million. EPA estimates that a
pipeline would cost $370,000 and have a present worth value of
annual costs of $350,000. Therefore, the total cost for the.
commenter'B proposal is $6.68 million, approximately $2.5 million
less than EPA's selected alternative (8ee comments 59 , 62"
concerning revisions to EPA's alternative original. cost
estimate). The pipeline envisioned during EPA's cost estimate
would not comply with state or Federal requirements for pipelines
transporting hazardous sUbstances, nor d~es it include right-of-
way costs. Compliance with such regulations and acquiring right-
of-way would drastically increase the cost.
There are also many legal implications that would arise as a
result of the proposal to commingle contaminated, untreated water
30
-------
f'
extracted from the A-zone groundwater into the ArVin-Edison South
Canal for irrigation uses. For example, regulations under the
Resource Conservation and'Recovery Act (RCRA) would apply
because: 1) dinoseb is a listed hazardous waste UDder RCRA (40
CFR Part 261.33) aDd 2) the activity at the site would
constitute disposal as defined by RCRA. Disposal of dinoseb
contaminated water, in particular, triggers a number of
significant RCRA requirements. As a result, EPA would be
prevented from simply dumping the untreated water into the canal
as a meaDS of dilution. At the very least, EPA would be required
to haul off-site the dinoseb contaminated water to a proper
treatment, storage or disposal facility.
Additionally, the state and Regional Water Boards are authorized
to take enforcement action to protect the quality of waters of
the state, such as the Arvin-Edison South Canal, based on various
environmental reguiations. For example, the state Board has
established Regional Water Quality Control Pians that set forth
beneficial.uses and numerical and narrative standards to protect
surface water quality. Additionally, State Board Resolution No.
68-16 (ADtidegradation policy) states that the disposal of wastes
into the waters of the state shall be so regulated to achieve
highest water quality consistent with maximum benefit to the
people of california. This policy generally serves to restrict
dischargers from reducing the water quality of surface waters
even though such a quality reduction might still allow the
protection of the beneficial uses associated with the water prior
to the quality reduction.
Secondly, with respect to the concept of disposing of the
contaminated water in the canal as a selected remedy, EPA is
guided by the.. statutory preference that treatment of
,contamination rather than non-treatment be aChieved. section 121
of CERCLA states: "Remedial actions in which treatment which
permanently and significantly reduces the volume, toxicity or
mobility of the hazardous substances, pollutants, and
contaminants is a principal element, are to be preferred. over
remedial actions not involving such treatment." 44 u.s.c 5 9621.
Because treatment of dinoseb i8 available for the Brown' Bryant
site, dilution would not be the preferred remedy.
Finally, the proposal notes that EPA may be required to indemnify
those involved in the disposal process from any liability arising
out of the use of the commingled water. EPA does not enter into
indemnification agreements for reasons including that such an
agreement would constitute an unauthorized appropriation of
federal funds. .
31
"-..--. ----'- ._-~._-... "---'-------.---......
-'. -'-'--'--"-.-'-".--. "-...'
... .".'...'-' -.-. .
-------
--- ...,_.. .---...
---. "." --.------.-'-------'---.----"---"'-."
Comments from Arvin Community Services District:
2. The Arvin community Services District's well #1 has been a
concern of the U.S. EPA, the California state Health Department,
and the Kern County Health Department, as well as the Arvin
Community Services District, for a number of years because of, the
close location of the Brown. and Bryant site to the well. The
well has been sampled by the Arvin C.S.D. and EPA for a number of
years for contaminants that were found at the Brown and Bry~nt
site as well as other pesticides, organics and inorganics. I
first began testing well #1 in 1985 at the request of the Kern
County Health Department for contaminants found at the' Brown and
Bryant site. Although the drinking water produced by well #1 has
not currently been affected by the contamination at the Brown and
Bryant site, there is still a concern that the drinking water
could some day be affected by that contamination.
Because of this concern the Arvin Community Services District
would like to request that some of'the funds that have been
appropriated for the clean up of the contamination of, the Brown
and Bryant site be used to relocate well #1 to another location
in Arvin that would be further from the Brown and Bryant site. A
map has been enclosed to show one possible location as to where
to relocate the well. Other locations may also need to be looked
at as possible relocation sites because of the travel effects of
the contaminants found at the'Brown and Bryant site. In any
effect another location to replace well #1 would help eliminate a
lot of the present concerns about the close proximity of well #1
to the Brown and Bryant contamination.
Response: The purpose for this proposed remedial action is to
address the source of contamination, the A-zone groundwater, and
the surface soils. The deeper groundwa~er layers will be
addressed in the next proposed plan. Well #1 draws indirectly
from the deeper groundwater; therefore, it will be addressed in
the next proposed plan. ' .
In the past, EPA has selected remedial actions that require
relocation of municipal wells that were impacted by ,
contamination. Bowever, EPAhas written a new policy with regard
to municipal wells where EPA would pay for cost associated with
treating the contaminated water from the municipal well, not
relocate the well. Since well #1 is not currently impacted from
the contamination, EPA presently would not relocate the well.
32
-------
----------------------
. ----
Comments from the Sierra Club. Kern-Kaweah Chapter:
3. Could the cap be disturbed by windstorms, like the one we
had here 12/77, or an occasional earthquake, or another Caliente
stream Group flood? If so, would there be any danger? Would
costly rep~irs ~ needed? WhQ would pay for the repairs?
Response: The RCRA cap proposed ror the site will be a multi-
layer cap designed to meet the Federal and state requirements for
such caps. A RCRA cap is commonly used at hazardous ~aste sites
around the state and country. Damage to the cap may 9ccur during
a very unusual storm event or large earthquake; however, severe
damage would be very unlikely. Any damage to the cap would be
repaired in accordance with stat. RCRA regulations; the costs
associated with such repairs will vary depending on the type of
repair. The State has requested that the cap be designed to
minimize as much as possible the long term maintenance and
monitoring costs. Any damage to the cap is not expected to pose
any short term danger to the public and prompt repair will ensure
that there is no long term impact.
4. After clean up, is there any agricultural, industrial,
residential or recreational or wildlife use for this area?
usefulness vary with the alternative chosen?
Does
Response: The ruture use of the property will be different for
the portion of the site under the RCRA cap as compared to the
rest of the site. Hore than likely, the area with the RCRA cap
vill not be available ror any ruture use ror at least 100 years,
after the required'cap maintenance period is over. The remainder
or the site should be available ror any type of use arter clean
up. Host likely, the area will be used ror light industrial or
commercial purposes, vhich is consistent with current land use in
the area of the site. The different alternatives do not affect
the ruture land use because all alternatives ,include the capping
requirements. '
5. A recent Bakersfield Californian article says the Arvin-
Edison Water Storage District' would dilute the contaminated
water. Has exposure to diluted dinoseb been shown to be safe for
pregnant women, infants, and all genetic varieties of other
people, including those exposed to unknown amounts of other
substances humans did not evolve with? .
Response: A discharge or dinoseb to the environment, such as was
proposed by the Arvin-Edison Water Storage District, would
require that dinoseb concentrations be reduced to at least the
drinking vater Hax~um contaminant Level of 7 ug/l. This
concentration is the liretime health advisory determined ror
dinoseb. The lifetime health advisory is considered protective
of noncarcinogenic adverse health effects over a lifetime'
exposure. It is derived rrom the no-observed-adverse-erfect-
33
.._._-------------._-~.._.__..._... ... --
- ~- -. -.... ----.--.--- -..-. ..--
-------
. ."-'--'- ~_-'M'_"__---'--'-'-"'--"--"'-"-'----.' ._---
level (NOAEL) for dinoseb developed from animal studies and
assumes a lifetime exposure. The level is rurther reduced by
takinq into account uncertainty factors and assuminq that
exposure from otber sources may also occur durinq a lifetime.
Lonqterm bealtb effects on bumans from exposure to dinoseb have
. not been documented. Tbe health effects data used to develop the
lifetime bealtb advisory vas based on animal studies. At hiqh
levels, dinoseb is toxic to bumans. Animal studies also suqqest
that dinoseb, at concentrations vell above the lifetime health
advisory level, affects liver and kidney functions and fertility,
and may cause birtb defects in preqnant vomen. Dinoseb has not
been sbown to be a possible carcinoqen.
Please also see our response to the Arvin-Edison Water storaqe
District's proposal.
Comments from the California Denartment of Health Services:
6. During our site visit, we observed that the north gate had
become removed from its hinges and a portion of the fence on the
south border of the site was bent so that trespassers could
easily enter the site from either side. Graffiti was observed on
the walls of the warehouse office, attesting to the presence of
trespassers. Better security measures are needed to ensure that
trespassers, especially children, are not exposed to surface soil
contamination from the site.
Response: EPA bas already taken measures to better secure the
gate and will inspect the rest of the fence and make additional
repairs as needed. We will also continue to monitor for
evidence of trespassers and openings in the fence so that
appropriate measures can be taken to ~elp prevent trespassinq.
7. A large pile of soil was observed at the northeast corner of
the site. No vegetation was growing on it, in contrast to a
similarly large soil pile at the center of the site. The pile
was not marked or covered and it was not clear whether it had
been sampled for contaminants. If this pile contains high levels
of Dinoseb (or other contaminants), it could create an exposure
pathway of fugitive dust inhalation exposure to neiqhboring
workers or children playing near the site. We recommend that the
pile be covered with a tarp until sampling shows that contaminant
levels are low, or it is remediated.
Response: Tbe soil pile in the northeast corner of the site is
from drill cuttinqs that bad low levels of contamination. This
pile was treated with a soil sealant to minimize dust dispersion.
Tbe second pile is from berm material that vas used durinq the
emergency response clean up activities; it is nonhazardous.
Since tbe first pile has been previously sampled and also treated
34
-------
with a sealant, EPA does not plan to place a tarp over it.
Bowever, these piles will be put under the on-site RCRA cap for
final disposal.
8. No off-site surface soil sample results were analyzed in
this ~I/FS, even though the two highest surface soil results for
dinoseb on the site were each located within approximately one
foot of the south and east fence line (figure 4.1). These
samples (locations 1110 and IC) showed concentrations of 5200
mg/kg (Table 6.1) and 7400 mg/kg (Table 6.2) respectively. The
only indication of off-site soil sampling is Figure 2.3, the
Kennedy Jenks Soil Boring Location Map. However, the RI/FS
states that U.S. EPA chose not to incorporate these results "due
to resource limitations and because the data K/J collected was
not within the scope of EPA's RI/FS" (p. RI2-7).
Given these high surface soil levels so close to the fence line,
a critical data gap exists as to the levels of surface soil
contaminants outside the fence lines. Additional off-site ..
surface and subsurface soil sampling needs to be done in order
to determine whether a current completed exposure p~thway exists
for children or other residents and neighboring workers to off-
site surface soil contaminants. This data is essential to
determining whether the proposed RCRA cap needs to. be extended
off-site or other types of remediation need to be considered.
Response: EPA did collect soil boring samples from the top one
foot at off-site locations. These samples are from soil borings
v, Z and AA, located near the dinoseb spill area, and 80il
borings T, RB, KK and HN, located south of the site (see Table
4.3 and Figure 2.1 of the RI Report). Concentrations from these
samples were all well below the cleanup standard of 80 mg/kg for
dinoseb proposed in the Feasibility study. . Bowever, to ensure
that off-site 'areas located near to dinoseb hot spots are at
acceptable concentrations, EPA will collect additional off~8ite
samples during the remedial design phase for the remedy. .
9. The RI/FS does not evaluate miqration patterns of soil
contaminants .. off-site. Based on Figures 2.1 through 2.5 , it"
appears that essentially no off-site subsurface soil samples were
taken by U.S. EPA since the Emergency Response sampling event
ending in May 1990. Kennedy Jenks off-site soil samples taken
between June 1991 and April 1992 were not evaluated as part of
this RI/FS. This data gap is of particular importance in
determining whether a potential future exposure pathway exists to
nearby residences. In particular, soil gas emission from .
volatile organic chemicals in subsurface soils should be
considered. .
The U:S. EPA's preferre~ Altern~tive 12 does not directly
remed1ate subsurface s011, stat1ng that "horizontal soil flushing
will also treat the readily removable contamination from the s011
35
. . --..--.--. ..-------.--.------ -.-------."----'-" .
-------
. . - - -. ~~'''.''
. .-.-.---. --....--.-- --.---. -- . - ._._.~
. -_......__..~... ..
in the saturated zone" (pg. FS-5-2) If off-site subsurface soil
migration of chemicals, particularly suspect human carcinogens,
is moving southwesterly toward the nearby residences (as is the
pattern for A-zone. groundwater contaminants), direct soil
remediation, such as the soil vapor extraction method included in
, Alternatives #5 and #6 may be required. In addition, evaluation
of off-site subsurface soil contamination is needed to determine
whether the proposed RCRA cap should be extended off-site.
Response: Based on 8Zis~ing da~a from BPA's Emergency Response
sampling, EPA believes ~ha~ ~here is ~o ~hrea~ to 'nearby
residences from soil gas emissions and ~ha~ additional da~a or
analysis is unnecessary. Four soil borings (T, HH, KM & HN) were
sampled by EPA be~ween the nearby residences and the site (see
figure 2.i in the RI Report). Hone of the target vola~il.
chemicals were de~ec~ed in soil a~ a depth less ~han 60 fee~. A
similar pattern is found in the Kennedy-Jenks data. Based on the
absence of contamina~ion from ~he EPA soil borings, addi~ional
da~a collection or analysis of ~his pa~hway was de~ermined ~o be
unwarran~ed. SUbsurface soil con~amina~ion also does no~ migra~e
la~erally a~ ~he speed and to ~he exten~ that contamina~ion in
groundwa~er migra~es la~erally. Some contamina~ion~will migrate
upwards from ~he groundwa~er, bu~ because of the dep~h of ~he A-
zone groundwa~er, i~ would no~ be expec~ed ~o pose a ~hrea~ ~o
nearby residences; fur~hermore, ~h. A-zone groundwa~er
remedia~ion would address such a threa~ by reducing the source
for ~he vapor emissions.
10. Dinoseb, a non-carcinogen, was the only chemical of concern
evaluated in this RIfFS. Several contaminants found in surface
and subsurface soil are classified by U.S. EPA as B2 carcinogens
(probably human carcinogens).,' These include 1,2-dichloropropane,
chloroform, hexachlorobenzene, DBCP and ethylene dibromide.
, Risks from these chemicals were not evaluated in the risk
assessment portion of the RIfFS. Inclusion of these carcinogens
in ther1sk assessment would likely increase the overall
potential risks to the public significantly, and would provide a
more accurate basis on which to make remedial decisions.
Response: The risk assessment performed by EPA for B&B includes
elements that are part of an ongoing effort by SPA ~o develop
more streamlined and cost effective investigations a~ Superfund
sites. In evaluating ~he risks from surface soil con~amina~ion
(i.e., the contamination in ~he ~op 7 feet of soil), BPA chose to
analyze only ~he most dominan~ pa~hways and con~amin&D~s~ Two
factors were considered in ~aking this approach. Pirs~, BPA
chose to focus its investiga~ion on ~he pathways and contaminants
that produce a potential risk sufficien~ to mee~ the s~tu~ory
requirement tha~ an unacceptable risk be presen~ in order to ~ke
a remedial action. Additional analysis of ~he risks would no~
change this finding. Second, SPA felt that addi~ional analysis
of the risks would only be appropriate ,if the analysis might
36
-------
change the proposed remedies.
Dinoseb was identified as the only contaminan~ of concern for
surface soils after first screening out all contaminants that
were not surveyed at the site at greater than 5% frequency, and
then screening out. any remaining chemicals that were not detected
in surface soils at concentrations above health-based levels of
concern. Only dinoseb remained after conducting this screening.
Of the chemicals mentioned in the comment, only DBCP,
hexachlorobenzene and 1,2-dichloropropane were detected at
greater than 5%. frequency in s~face soils. Because the highest
concentrations observed for' these chemicals do not pose a
significant health risk in surface soil (see section 6.1 of the
RI Report), no further analysis for these chemicals was
conducted. .
As for the risks associated with contaminants within the
remainder of the A-zone, the only exposure route of concern
identified by SPA is the potential exposure that may result from
the contaminants reaching the B-zone groundwater, a potential
drinking water aquifer, or from contamination of the nearby city
well. .Instead of conducting a risk assessment for this exposure
pathway, EPA relied on drinking water Maximum contaminant Levels
CHCLs) or other health based levels for evaluating the potential
risks associated with these chemicals. By this approach, EPA
does include the potential carcinogens found at the site in its
analysis. The proposed cleanup for the A-zone groundwater is
intended to address the risks from the variety of organic
chemicals found in this groundwater zone and not just dinoseb.- A
groundwater risk assessment may be a component of the second
operable unit RIfFS for the site.
11. . The A-zone aquifer was eliminated as a pathway because water
production was shown to be 90 gal/day, well below the EPA
guidance of a minimum of 200 gal/day to be defined as a public
water system. However, the slug tests for water production were
taken before the rainy season of 1992-1993. and therefore sample
results may be atypically low. In addition, no well survey was
made to determine whether nearby residential irrigation or
drinking water wells were in use.
Response: While the slug test was conducted during the extended
California drought, EPA does not believe that the results would
change significantly based on the following observations: Pirst,
this water bearing unit is characterized by thin bedded clays,
silts and sands, that inhibit the ability of this formation to
produce and sustain a substantial flow of water. Second, the A-
zone has consistently shown evidence of low water production as
indicated by how some A-zone wells were consistently purged dry
during sampling and exhibited poor recovery, with little
difference observed seasonally and over time. Pinally, while
there may be seasonal changes in water production from the A-
37
--..---.-..---. .-----...-----..------... . "'-.
. "---. -""-..- --_..
',-'. --.-. ...... ..
-------
. . ... . - ~ . .
. .. - - -,-_. ,-._.. ...-..
.'.".'-".- --_..~- ..-.-..---.------ --.--.-..---.---... .-. ... .
zone, EPA believes ~hat the 200 gal/day criteria needs to be
attainable year round. Based OD ~hese factors, SPA maintains
that the A-zone is not a drinking vater zone.
As for o~her wells located near
drilling logs from these wells.
actively pump water from either
vou1dnotinf1uence the resu1~s
~he si~e, SPA has compiled
However, none of these wells
of the A or B-zones, so they
of ~he slug test.
12. The B-zone aquifer was not evaluated in this RI/FS because
u.s. EPA has decided to consider this aquifer as a second
operable unit. However, since it is a potential drinking water
source and is potentially connected to the deeper drinking water
aquifer via City Well #1, it needs to be fully characterized and
remediated as soon as possible. A timetable for addressing B-
zone remediation needs to be established and cleanup initiated in
the near future.
Response: EPA has already begun ~he B-zone RX/PS and expects to
begin field work ~his fall. A proposed plan for remediation of
this zone is planned for the end of next year. SPA is also
considering some early action on ~he B-zone to address the area
of highest contamination.. .'
Recent da~a collected on ~he B-zone indica~es a dramatic decrease
in concentrations at the most contaminated well, WB2-1.
Concentrations of 1,2-dich10ropropane have decreased steadily
from a high of 1,700 ug/1 in April 1992 ~o 50 ug/1 in JUDe 1993.
No significant increases in contamination have been observed over
the same time period in any of the B-zone wells.
13. The RIfFS states that "deed restrictions or other
institutional controls would be placed on the property to ensure
that the cap remains safely intact and that the soil under the
cap remains undisturbed in the future" (p. FS-1-2). These deed
restrictions need to be spelled out to ensure that future. .
commercial use of the property does not result in exposure to
surface or subsurface soil contamirtation. For example, under
what conditions will, a future owner be allowed to disturb either
of the caps covering the site? Will buildings with foundations
be allowed? will EPA or other regulatory agencies oversee future
site activities? Who will be responsible for upkeep of the caps,
and for how long?
Response: The details of deed restric~ions for the site will be
developed during the remedial design for the caps. It is
anticipated that construction will be prohibi~ed on ~he RCRA cap
and that this area will be kept off-limits from future use, at
least through the completion of cap maintenance. The cap will be
maintained in accordance with the state RCRA regulations that
require 30 years of monitoring and 100 years of main~enance ~f a
RCRA cap. .
38
-------
I
I
I
The remainder of the site containing' the "b~sic cap" will include
at least short term deed restrictions to ensure that the cap is
maintained to control surface water drainaq.. New structures
will likely be allowed as lonq as the drainaqe control is not
impacted. Followinq the completion of the A-zone qroundwater
cleanup it may not be ne~essary to keep this ca~ in place; such a
decision will probably be deferred until that t~e. Both EPA and
the state will have input to future site activities. EPA is
required to conduct a five year review of the site every five
years from when the remedy construction starts.
Comments from Southern Pacific TransDortation ComDanv and The
Atchison. To~eka and Santa Fe Railway ComDanv:
14. In qenera1, the Draft RI is deficient because it does not
present a sufficiently developed conceptual model of the qeo1oqy
and hydro10qy of the Site, and because the fate and mobility
mode1inq conducted to evaluate the potential effects of chemicals
found in soils and qroundwater is overly conservative and
overstates the potential for additional miqration.
The critique presented herein focuses on factual errors in the
text and tables of the report and on deficiencies in the basic
assumptions utilized in fate and mobility modeling. Because of
these errors and the incorrect assumptions utilized, the
conclusions made by the EPA regardinq potential impacts to
groundwater and consequent soil cleanup criteria are incorrect.
Therefore, the Draft RI does not present a sound or adequate
basis for the recommended remedial activities.
Response: EPA stands by its analysis of site conditions in the
RI Report and has responded to each detailed comment on the RI
Report below. . The comments focus in particular on the vadose
zone modelinq conducted by .EPA and challenqe the assumptions used
by EPA in its modelinq. The comment incorrectly concludes that
EPA draws overly conservative cleanup requirements from the
m04elinq. As explaine4 below, EPA in fact concluded not to
select the soil cleanup remedy for the A-zone'soi~s. Xn
ad4ition, EPA relie4 primarily on data collected in the fiel4,
not the m04elinq results, to determine the appropriate action for
the qroundwater.
15. The RI text and fiqures are inconsistent with regard to
direction of qroundwater flow. For example, in the Executive
Summary (paqe 2) the groundwater in the A-zone is stated to flow
"in a qenerally southern direction". On page 3-5 of the RI, the
direction of groundwater flow is stated to be "to the south and
west". A detailed analysis of the direction of groundwater flow
indicates that there is an apparent mound or limb of groundwater
extendinq from the southwest, corner of the B & B Site.
3t
"..- .---..--- -----'-'_n____'<- '-,-" -'~ .
--'J'
. - -- "" ~ '.
-------
. ~ _. ~ -_... .. -- _.. ---....--.. . .-. -- .'
~---.~_._._~_..._--_._---------
Response: Section 3 of the RI Report does provide a more
detailed analysis of the flow patterns in the A-zone groundwater.
Figures 3-6 and 3-7 include maps of the vater table that show the
varied flow patterns of the site, and the text in 8ection 3.5.2
identifies the possible mounding of groundwater on-8it..
Generalized statements elsewhere in the text are not used as
substitutes for the more detailed analysis presented in section
3. .
16. The groundwater velocity in the A-zone is given as 53
ft/year on page 3-7 of the RI. For the A-zone, utilizing a
hydraulic conductivity of 4 X 1.0-4 em/see (pages 3-5, RI) ~ a .
porosity of 40% (page 3-5, RI) and a gradient of 0.007 (page 1 of
Appendix I), the value for groundwater velocity calculated is 7.2
ft/year. The derivation of the groundwater velocity value of 53
ft/year is. unclear.
Response: The groundwater velocity of 53 ftfyear is based on a
hydraulic conductivity of 4 x 10-4 em/sec, vhich vas obtained
from the slug test, an effective porosity of 26% estimated from
literature values (the laboratory results vere not.available at
the time; however, the effective porosity is typicallY less than
the porosity measured in a laboratory), and a gradient of 0.034,
which is a localized gradient for the wells used .in the pump
.test. The gradient used in tbe modeling (0.007) vas an estimate
over a larger area of the site. The RI report states that the
reported velocity is an estimate for the vells included in the
slug test. OVer the entire site the velocity will vary .
considerably; however, the velocities are all expected to be
relatively slow. For the RIfFS, the general characterization of
the A-zone groundwater as a slow moving, low producing vater.
bearing unit with localized variation in hydrology is the most
important observation.
17. In several instances data presented in Table 4.9 of the RI
present average concentrations for chemicals which are of a
greater magnitude than the highest concentration reported.
Examples of this are the concentration reported for 1,3 DCP.at 41
to 65 feet and CBCP at 0 to 10 and 31 to 41 feet. An average
concentration of 1,3 CCP of 32 ug/kg was reported as contrasted
with a ~iqh concentration of 12 uq/kq for the 41 to 65 foot depth
interval. An average concentration of DBCP of 10 ug/kg was
reported as contrasted with a high concentration of 6 ug/kg for
the 0 to 10 foot depth interval. An average concentration of
CBCP of 183 ug/kg was reported as contrasted with a high .
concentration of 110 ug/kg for the 31 to 40 foot depth interval.
In several instances data .presented in Table 4.10 of the RI
present average concentrations for chemicals are of greater
magnitude than thehiqhest concentration reported. Examples of
40
-------
!
this are the concentration report for 1,3 DCP at 0 to 10 feet and
DBCP at 0 to 10 feet, 41 to 65 feet, and 66 to 85 feet. An
average concentration of 1,3 DCP of 16 ug/kg was reported as
contrasted with a high concentration of 15 ug/kg for the 0 to 10
foot depth interval. An average concentration of DBCP of 16 .
ug/kg was reported as contrasted with a high concentration of 15
ug/kg for the 0 to 10 foot depth interval. An average
concentration of DBCP of 77 ug/kg was reported as contrasted with
a high concentration of 72 ug/kg for the 41 to 65 foot depth
interval. An average concentration of DBCP OF 423 ug/kg was
reported as con~rasted with a high concentration of l20,ug/kg for
the 66 to 85 foot depth interval.
Response: Average concentrations were derived by averaging the
detected values with the quantitation limits for samples that
were non-detect. It is EPA's policy to use the quanti tat ion
limit or half that value instead of 0 for values that are not
detected. In .ome cases, this resulted in an average
concentration greater than the highest obserVed result because of
relatively high quanti tat ion limits among some samples. Where
there were no detected concentrations, the average was put in
parentheses.
18. ~he organic carbon distribution coefficients for 1,2 DCP and
DBCP presented in RI Table,5.3 are incorrect. The organic carbon
distribution coefficient for 1,2 DCP should be 51 ml/g. The
organic carbon distribution coefficient for DBCP should be 129
ml/g.
Response: The'comment is correct that the organic carbon
distribution coefficients were mistakenly switched in Table
for 1,2-DCP and DBCP. Bowever, the correct coefficient was
in the VLEACB modeling.
5.3
used
19. Many of the basic assumptions utilized in the fate and
mobility modeling conducted by and for the EPA utilized incorrect
assumptions regarding conditions at the B & B site or the
characteristics of the chemicals of concern. The net result of
these deficiencies is that the results of the' modeling
misrepresent the potential for further migration of chemical
through the vadose zone into groundwater.
Respon.e: All modelling involves generalizations and assumptions
that ..y result in inaccuracies or biased modeling results. For
this reason, EPA used the modeling results as only one of a
number of factors considered in arriving at appropriate remedial
alternatives. por the Brown' Bryant site, the modeling results
were used, along with other data, to decide what remediation
option to consider for the A-zone .oil. and whether or not to
.elect the A-zone .oil remediation option. BPA'. preferred
alternative doe. not include soil remediation for the deeper A-
zone .oils~ This decision was made by weighing the results of
41
- -_._..__._--..--"-----_._._--~._.---...- - -'.
.. --. ... - .'. -..
-------
.4. '.'- .-.-.----....-- ... ...
.- -.- . .---..-.... -.-,--..-.--.------- ----.--.-.--- -.'-_.".'-".~" .
the screening models used in the RI/~S, the cost of the
treatment, and how best to protect the B-zone groundwater.
concluded that treatment of the A-zone groundwater and the
are a sufficient remedy for protecting the B-zone and that
treatment of the soil with soil vapor extraction is not
warranted.
BPA
cap
The modeling during the RI/~S was conducted primarily for the
above purpose. It was also considered in making 8stimatesof the
interim cleanup level range for the A-zone groundwater. A range
for cleanup levels was established in part because of the general
inaccuracies inherent in modeling~ BPA believes that information
obtained during the implementation of the A-zone groundwater
treatment will ultimately provide the best estimates for a
specific cleanup standard within the range established in the ~S
Report.
The comment contends that BPA misrepresents the potential for
migration of chemicals through the vadose zone into groundwater
by generally overestimating the potential impact. BPA disagrees.
As pointed out above, BPA has chosen not to treat the A-zone
soils. As for the potential impact from A-zone groundwater on
the B-zone groundwater, BPA has based its concern about this
threat primarily on observed contamination in the B-zone in
excess of drinking water standards. This B-zone contamination
demonstrates a connection between the A and B-zones and is a
primary reason for EPA's proposal to treat the A-zone
groundwater. The modeling provides estimates of the potential
magnitude of impacts from the A-zone groundwater over time, which
were considered along with the observed concentrations. .
20. Although descriptions of the lateral and vertical
heterogeneities that exist in vadose zone soils are described in
many sections of the Draft RI (see for example, Executive Summary
page 1, RI pages 3-3 through 3-5, 5-4), all model simulations for
chemical transport utilize one dimensional models. Therefore,
lateral heterogeneities are not accounted for in any water, and
vertical heterogeneities are lumped into gross classifications.
The effect of these gross lateral and vertical str~tiqraphy
simplifications on model results is not discussed anywhere in the
RI. For example, the lateral movement of chemicals due to
stratigraphic barriers (clays and silts) is not assessed in the
modeling. This would result in the results of the modeling
overestimating concentrations of chemicals of concern which may
occur in groundwater. This error then results in the development
of unnecessarily stringent soil cleanup criteria.
Response: A one dimensional model of the type used for this
RI/FS is typically used at many Superfund sites. Although, it
. may result in an overestimate of the potential impact, this
factor was considered in the recommendations made from the.RI/FS.
It is unclear why the commenterbelieves that EPAhas set.
42
-------
I
I
I
unnecessarily stringent soil cleanup criteria, given that EPA did
not set numerical cleanup standards for the deeper soil (see
section 3.1.3.2 of the FS Report), and the final remedy selected
does not involve direct remediation of the vadose zone soil.
21. The Ri (and modeling report, see below) does not use curre~t
deqradation rates available in the literature, correct
assumptions regarding deqradation pathways, and in general, does
not include adequate assessment of the effects of degradation on
chemical fate and mobility.
Response: Deqradation rates used for the modeling were ~ased on
literature va1uesl. Because these values may not reflect site
conditions, SPA used a range of degradation rates and did not
rely on a single value. As noted in the Rl Report, the model
normally assumes no degradation. SPA included a degradation
factor to make the results more realistic. EPA believes that the
assessment of chemical degradation is adequate considering how
the results were used and what other factors were considered in
the remedY,. selection.
22. Ri page 5-6 states that, "relatively low expected oxygen
levels" are expected in the subsurface at the site. Therefore,
the RI continues, conditions at the site do not favor most
degradation processes. The basis of these statements are not
given. In fact, elevated nitrate and sulfate concentrations in
A-zone qroundwater suggest that condit~ons are far from anaerobic
(methane producing). In addition, the critical process which the
RI suggest controls the miqration of volatile organic chemicals
in soils at B&B (i.e., diffusion in soil gas), controls the
movement of oxygen from the atmosphere into the soil. The RI,
therefore,is.not consistent with comments regarding movement of
'gases in vadose zone soils. Assumptions regarding the low
expected oxygen levels lead to overestimated concentrations of
chemicals of concern predicted to occur in groundwater resulting
in unnecessarily stringent soil cleanup criteria. .
Response: The statement regarding oxygen levels was not intended
to imply that site conditions are anaero~ic. The main point of
the comment seems to ~e that SPA has used the assumption of low
oxygen levels to overestimate the impact of vadose zone
contamination resulting in overly stringent soil cleanup levels.
This is incorrect. As pointed out ~ove, numerical soil cleanup
levels were not est~lished and SPA is not proposing to cleanup
the vado.e zone. The range of degradation rates used in the
modeling factor in changes in degradation that may result from
different site conditions, including differences in degradation
rates resulting from different oxygen levels.
lHandbook of Environmental Dearadation Rates, P.H. Howard
et a1., 1991.
43
"'.'-"--'-,~ ---...----.--....--. ..-.- . --..-".
..-.. "'.---. -,-
-------
" '-"'.' -.--'--"'''-_0_._._.'",-, ~.._...'" -..-.-....- ...'
23. Degradation half-lives for DBCP are quoted to be between 0.5
years to 141 years. The long half-life is quoted for hydrolysis.
Recent studies conducted by Deely et. al., (1991) found for
typical groundwater conditions in groundwater from one site in
the Fresno. California areas demonstrated that the half-life for
DBCP is 6.1 years, much shorter than the 141 years quoted in the
RI. Therefore, the ass~ption of a long-half life for DBCP is
inappropriate and leads to an overestimation the concentration of
a DBCP predicted to occur in groundwater resulting in
. unnecessarily stringent soil cleanup criteria.
Response: Again, EPA did not consider only a single half-life
but considered a range of half-lives. The range considered for
DBCP was 1, 2 and 10 years (see Table 5.5 of the RI Report),
which is less conservative than considering just a , year half-
life.
24. The RI provides degradation rates for all chemicals found in
A-Zone groundwater on page 5-6. These degradation rates are not
taken into account in the fate and mobility modeling. The
degradation times are short relative to the hundreds of years
given as transport time from vadose zone soils to the A-zone
groundwater. Therefore, significant reductions in chemical
concentrations would be expected to occur due to degradation as
chemicals migrate through the vadose zone. Consequently, all of
the model simulations overestimate the concentrations of chemical
of concern predicted to occur in A-Zone groundwater resulting in
unnecessarily stringent soil cleanup criteria.
Response: The comment that "degradation rates are not taken into
account" is incorrect. EPA did take into account the degradation
rates for the volatile chemicals modeled by VLEACB. The results
are included. in Table 5.5 of the RI Report.
25. Water solubilities for chemicals of concern listed in RI
Tables 5-2 and 5-3 were measured for a pure solvent in contact
with water. The water solubility for a given chemical in a
mixture of solvents is lower than that for the pure chemical.
The use of pure chemical water solubility results in .
overestimation of the chemical concentration dissolved in water.
Therefore, the use of pure phase water solubilities for chemicals
of concern result in overestimation of groundwater concentrations
and unnecessarily stringent soil clean up levels.
Response: EPA agrees that solubilities would be lower in
mixtures as compared to pure chemicals and that this does make
for a small overestimate of the predicted impacts of the modeling
results. Bowever, this impact is even smaller since 9as phase
transport was typically found to be more critical than solute
transport, especially with the assumption that the site will be
capped. Bowever, as discussed previously, this impact does not
affect soi~ cleanup levels nor would it change the decision not
44
-------
to cleanup the vadose zone soil.
26. The fate and mobility modeling conducted by AScI for the EPA
also utilized incorrect basic assumptions. The overall effect of
these deficiencies is that the conclusions made regarding soil
cleanup levels and potential effects on groundwater quality are
incorrect. .. ....
Response: The basic premise for all of the comments regarding
the modeling conducted by AScI is that EPA made incorrect or
overly conservative assumptions that resulted in overly .
conservative conclusions and excessively stringent cleanup levels
for dinoseb. These comments ignore the basic conclusions made by
EPA from the modeling results. EPA concluded that treatment of
dinoseb in the vadose zone is unnecessary because the cap will
significantly retard the movement of dinoseb as a result of
cutting off the infiltration of water. Ho cleanup standard is
proposed for dinoseb in subsurface soil in FS Report.
27. As noted above, all model simulations ignore degradation
(AScI page 21, 22). . This results in the prediction of
unrealistically elevated concentrations of chemicals of concern
in A-zone groundwater, which in turn results in the establishment
of soil cleanup levels more stringent than necessary to protect
groundwater. quality.
Response: See comment l' regarding modeling assumptions and soil
clean-up levels.
28. Although descriptions of the lateral and vertical
heterogeneities that exist in vadose zone soils are described in
many sections. of the Draft RI, all model simulations conducted by
AScI to evaluate chemical transport utilize one dimensional
models. Therefore, lateral heterogeneities are not accounted for
in any ¥ay, and vertical heterogeneities are lumped into gross
classifications. The effect of these gross lateral and vertical
stratigraphy simplifications on model results is not discussed.
The oversimplification of the stratigraphic setting results in
the development of unnecessarily stringent soil cleanup criteria.
Response: See comment l' regarding modeling assumptions and soil
clean-up criteria. The assumptions used in modelling did not
result in the conclusion stated in the comment. See a180 the
response to comment 20.
29. Unrealistically elevated concentrations of Dinoseb were used
as MULTIMED model input. As described on page 17 of the AScI
report, the only field measurements used in the development of
the average concentrations for the source input for the different
depth in~ervals ~ere th~se greater than 100 ug/kg. In addition,
calculat10n of D1noseb 1n the top 0 to 25 feet ignored field data
45
... -._._---._-_._---,-_._~.--_._- -...--.-.-.-.-.
. . - -- . -"' - . - - -- - ." . .
-------
.-.._--_._----_.__._-._.~--_..- .-.. -----....
when the concentration measured was below 1000 ug/kg. This
averaging created unrealistically elevated concentrations of
Dinoseb for use in simulations and resulted in overestimation of
possible groundwater concentrations of Dinoseb and unnecessarily
stringent soil cleanup levels.
Response: The values used were for characterizing the worst
portions of the site. It was for these areas that EPA was trying
to make a remediation decision using the modeling results. 8ee
comment 19 regarding 'modeling assumptions and clean-up levels.
This approach did not result in the conclusion stated in the
comment. "
30. The comparison between model results and field data is
presented on pages 14 and 70. The report states that there are
significant differences between the observed and estimated
concentrations. Possibilities for this discrepancy which have
been ignored include facts such as the commingling of plumes from
multiple release sites is different than what is modeled, and the
, release of large volumes of water during the original release of
chemical (e.g. from the ponds and sumps on B & B property) means
that release conditions are not being adequately represented in
the model scenarios.
Response: The purpose of the modeling was not to model
historical conditions that led to the current problem. Instead,
the modeling was conducted to ev~luate the current and future
potential impacts from soil contamination currently found at the
site. The model results differ from the field data in that the
model predicted a smaller impact than what has been actually
observed in the field. This difference is due in large part to
the model's focus on current conditions. .
'31. The organic carbon content of the "perched zone aquifer" is
given as 0.005% (0.00005 mg/kg). The basis for this estimated
organic carbon content is not given. This unrealistically low
organic carbon content, given without backup, results in
overestimating groundwater concentrations of Dinoseb and the
-development of unnecessarily stringent soil clean up levels.
Response: The organic carbon ~alue was established by
extrapolating the decrease in organic carbon observed with depth
from existing data. Bowever, this conservative assumption did
not result in the conclusion stated in the comment (.ee above
discussion). .
32. The infiltration rate used in VLEACH model simulations
(possibly used in MULTIMED simulations, see below) was given as
25% total maximum rainfall. The basis of the estimation of
infiltration rates of 25% was not described. This infiltration
rate is likely to be 5 to 50 times greater than what would
actually be occurring in a climate as dry as that at the B&B site
46
-------
(stone, 1986; Nichols, 1987; Phillips et al., 1988). Use of
unrealistically high infiltration rates leads to the
overestimation of predicted concentrations of chemicals of
. concern in groundwater and, therefore, the development of
unnecessarily stringent soil cleanup criteria.
RespoDse: 8ee comment 19 regarding modeling assumptions and
soil clean-up levels. ' In should be noted that the VLEACH results
alsc included a modeling scenario vith an infiltration rate of
one.tenth the value quoted here. .
33. Infiltration rates used in model simulations are not clear.
The statement is made on page 24 of the AScI report that the
infiltration rate utilized is based on maximum rainfall. The
maximum rainfall given on page 25 of the AScI report is 16
inches/yr. Simulations utilize 25% of this as the infiltration
rate or 4 inches/year or 0.33 ft/year, or 0.1 m/yr. However, the
statement is made on page 25 that 0.032 m/year (25% of minimum
precipitation) is used for recharge rates. it is unclear if
there is a difference between infiltration rate and recharge 'rate
in model simulations, or if. this is the same parameter, which
value was used. A infiltration rate of 0.33 ft/year was used in
VLEACH model simulations presented in Section 5 of the RI. It is
unclear if VLEACH and MULTIMED simulations use the same
.infiltration values. '
Response: In the Hultimed model there are separate values for
recharge rates and infiltration rates, vhereas in the VLEACH
model only an infiltration rate is utilized. The ~ame
infiltration rate vas used for both models except that the VLEACH
modeling also used an infiltration rate of 0.033 for modeling
scenarios that assumed a cap.
34. In a layered soil stratigraphy, the infiltration rate would
be controlled by the soil layer with,the lowest saturated
intrinsic permeability. In the case of the soil system. described .
at B&B, layer 4, with a reported saturated"intrinsic permeability
of 1 x 10-8 em/sec (RI page 3-5), would control infiltration
rates. However, much higher saturated intrinsic permeabilities
were used in model simulations. Therefore, concentrations of
chemicals of concern have been overestimated in groundwater
resulting in the development of unnecessarily stringent soil
cleanup criteria.
Response: BPA disagrees. The thickness of the various layers is
also a factor, not just the permeability of the finest grain'
layer. 8ee also response to comment 32.
35. The air entry pressure head value and van Genuchten ALFA
coefficient used in model simulations as reported on page 25 of
the AScI report, represent gravels rather than the finer grained
.7
. .-~. p~--~-- ..-...,- ------- --- --- ..-.- ----- --- .
".-.--..' ..'
. "'...' .-.--. -. - ..
-------
. . - -_. ~ _._----- ._.~_._- "-._-' ----.- ---.".
- -.-- - . --._.~---_.~-----
soils common on the site. The use of parameters which represent
gravels creates a very conservative model simulation.
Therefore, the concentration of Dinoseb is overestimated
resulting in the development of unnecessarily stringent soil
cleanup criteria.
Response: See commen~ l' regarding modeling assump~ion8 an~ 80il
clean-up levels.
36. The area given for Dinoseb source soils is 50 by 200 meters
(10,000 square meters, page 25 of AScI report). This area is
much larger than any known source area on the B&B site. In
addition, this area is much larger than the source area sizes
reportedly used for VLEACH modeling as presented in the RI (2090
square meters for the pond area and 930 square meters for the
sump area, Appendix I of RI). The unrealistically large area
used as input for.the MULTlMED simulations causes an overestimate
of the concentration of Dinoseb which might occur in groundwater.
This overestimation of Dinoseb concentration causes the .
development of unnecessarily stringent soil cleanup criteria for
Dinoseb. .
Response: The commen~ has iden~ified an error in the modeling
report. The size of the source area should have been SO by 200
feet. However, since ~he model is a one-dimensional model, i~
does not take into account the size of the source area 80 the
error has no impac~. See also above discussion regarding cleanup
criteria for dinoseb.
37. The hydraulic conductivity for the 40 to 65 foot interval
reportedly used in MULTlMED model simulations as stated on page
25 of the AScI report was 1 x 10-4 em/sec. Hydraulic
conductivity.values reported on page 3-5 of the RI are lower,
.ranging from a maximum of 1 x 10-4 to 1 X 10-6 em/sec. This use
of the maximum possible hydraulic conductivity causes the model
to overestimate the concentrations of chemicals of concern which
might occur in groundwater. soil cleanup criteria developed
utilizing these estimates are unnecessarily stringent.
Response: Because not all
time of the modeling, some
comment 19 with regards to
levels.
of ~he RI data was available at the
assumptions are conserva~ive. S.e
modeling assump~ions and cl.an-up
38. The SUMMERS Model does not allow for any sorption,
degradation, dispersion, volatilization or decay in the aquifer.
This model was utilized by AScI to evaluate groundwater
concentrations which would result from introduction of
unsaturated zone leachate, as estimated utilizing MULTlMED, to
upgradient groundwater. In the SUMMERS Model, regarding chemical
behavior in the aquifer no partitioning, is allowed between the
48
-------
aquifer solid matrix and the water. Therefore, the SUMMERS Model
will result in the overestimation of groundwater concentrations.
These overestimates are utilized to develop unnecessarily
stringent soil cleanup levels. In fact, on page 6 of the AScI
report, it is stated that "the results (predicted by the SUMMERS
Model) may be too conservative at times and if sufficient data
are available a more sophisticated approach is suggested to
arrive at a cleanup standard which is more economical".
Response: See comment l' regarding modeling assumptions and soil
clean~up level~. .
39. This review of the recommended groundwater remedial actions
presented in the Draft FS was performed by following guidance
documents provided by the U.S. EPA for remedial projects being
conducted in conformance with the pertinent criteria in the 6
February 1990 National oil and Hazardous Substances Pollution
contingency Plan ("1990 NCP"), section 300.430(e). Specifically,
the following.'P.S. EPA guidance documents (the "Guidance
Documents") offer explanations of the technical analyses expected
in a Feasibility Study completed consistent with the 1990 NCP:
o
EPA Guidance for Conductina Remedial Investiaations and
Feasibility Studies Under CERCLA, OSWER Directive
No. 9355.3-01, October 1988 (Interim Final); .
o
Guidance on Remedial Actions for Ground Water at Superfund
Sites, OSWER Directive No. 9283.1-2, December 1988 (Interim
Final).
The Draft FS does not provide the level of technical information
and analysis~f remedial alternatives specified in the Guidance
Documents.
In particular, all action alternatives in the Draft FS include
identical A-zone groundwater remedial actions' and costs. The
RIfFS Guidance documents state that alternatives should be
developed which achieve ARARs or health-based levels within
varying time frames using different methodologies. . The Draft FS
does not present such alternatives for groundwater extraction.
This document focuses on several areas of potential problems
associated with the recommended groundwater remedial action.
Since no alternatives to the recommended action are presented in
the Draft FS, it is beyond the scope of this critique to propose
alternative actions in detail. However, a general alternative is
presented in the conclusions which should be considered before
proceeding with the EPA's recommended action.
In addition,
described in
ambiguity in
made in this
the EPA's recommended groundwater alternative is not
detail in the Draft FS. Given this extraordinary
the Draft FS, the following assumptions had to be
document to provide a basis for an evaluation of .
4'
" '""... ..--,--~ -"'---.-..---.------.-------.-
--. . - -. - ",_.
-,.. '''''''''---''''.-.
-------
. - _......M_'.__- -- - --_..- --- - -- - .
- . M--- --"'- -.----..---.
M_'__'_'-..-.----.. -.--....-...-
EPA's proposed system design:
o The assumption was made that the area of A-Zone groundwater to
be remediated was approximate 5.6 acres. The details of the
groundwater extraction and reinjection system are not described.
On page FS-3-9, the estimated area and volume of .
A-zone groundwater containing 1,2-DCP greater than 10 times its
MCL are stated to be 5.6 acres and 3,650,000 gallons,
respectively; this area appears to constitute the assumed area of
A-zone groundwater requiring remediation in all action
alternatives. . .
o The assumption was made that the extraction/injection system
would consist of 75 wells arranged in 8 rows. It is stated on
page FS-4-4 that the groundwater system would consist of
"alternating rows of injection wells and extraction wells". In
Appendix A, an internal u.s. EPA Memorandum from Michelle Simon
to Cynthia Wetmore, dated 12 February 1993, states that the
groundwater extraction system will consist of "75 wells arranged
in 8-400 ft~ rows, 9 wells each row; each well will [be] used
,for both extraction and injection, separate piping system to each
well for extraction & injection." To facilitate review and
visualization of the proposed A-zone groun~water remediation
scheme, one possible interpretation of the proposed well network
is shown on the attached fiqure. This fiqure shows 72 wells on a
grid of 67 feet between rows and 50 feet between wells in each
row. Most wells are located off-site of the Brown & Bryant site.
o The proposed groundwater treatment system is assumed to be an
innovative technology consisting of a UV/hydrogen peroxide
oxidation treatment unit, utilizing a proprietary additive, ENOX
510, supplied by Solarchem. All action alternatives include a
capital cost. of $650,000 for this treatment system, which appears
to correlate with the estimated cost for the 10 gallon per minute
("qpm") system designed to reduce ethylene dibromide ("EDB") to .
"non-detectable" levels in the treated groundwater .(see Appendix
A, U.S. EPA Memorandum from Vance Fong to Cynthia Wetmore, dated
20 November 1992). . .
o The assumption was made that the cleanup goal for water
treatment prior to reinjection was "non-detectable levels".
However, the June 1993 U.S. EPA public notice/fact sheet states
that "the groundwater will be treated until it meets maximum
contamination levels established by state and federal
regulations."
o The assumption was made that the extraction rate for the system
was 10 qpm, although this estimate appears high. The extraction
rate from each A-zone extraction well is. estimated to be 100
gallons per day ("qpd") on page RI-3-7 of the Draft RI. However,
at 100 gpd (0.069 gpm) and assuming that half of the 75 wells are
used as extraction wells, the aggregate average extracted.
50
-------
groundwater flow rate would be only approximateiy 2.6 qpm, not 10
qpm.
o The assumption was made that the reinjection flow rate per well
is 100 qpd. There is no discussion in the Draft FS of the design
of the reinjection wells, water conditioning prior to
reinjection, injection well maintenance, or the planned injection
rate in each well. Because the appended u.s. EPA memo~andum
imply that the number of injection wel~s will equal the number of
extraction wells and no surface discharge options are discussed,
it is assumed that the proposed reinjection flow rate per well is
identical to the extraction rate, i.e., 100 qpd.
. Response: Al though the above section is intended for
introductory purpose and not as specific comment, it contains
erroneous assumptions which gave rise to incorrect comments that
follow. The comment stated that EPA did not follow its quidance
document because only one technology was proposed for the
remediation of the A-zone groundwater. EPA quidance2 does not
suggest that a minimum number of alternatives must to be carried
from the screening level to the detailed analysis of
alternatives; it only suggests the number of alternatives should
not exceed ten. EPA believes evaluation of more than one
remedial alternative would provide flexibility in the remedial
selection process. EPA does not and should not carry non-viable
technologies past the screening stage solely for the purpose of
having a specific number of alternatives. At some sites, there
are circumstances where there are not several technologies that
are feasible. . At Brown' Bryant, the site-specific contamination
has rendered many treatment technologies/process options non-
applicable, see Figure 3.6 of the Feasibility study Report.
Additionally, the comment erroneously assumes that the estimates
used in the cost analysis are the proposed design of the selected
alternative. The Superfund process includes three phases: the
remedial investigation and feasibility study phase; the remedial
design phase; and the remedial action phase. "The objective of
the RI/FS process in not the unobtainable goal of removing all
uncertainty, but rather toqather information sufficient to
support an informed risk management decision regarding which
remedy appears to be most appropriate for the given site~"l It
is the purpose of the remedial design phase to provide the
specific technical detail such &S well design, vell spacing,
extraction and reinjection rates, etc. Apparently, when the
commenter could not find the technical specifics usually reserved
for the remedial design in the body of the feasibility study, he
used estimates in the cost analysis appendix or made assumptions
included in the comment's introduction and commented on those
1 .
/ EPA Guidance for Conductina Remedial Investiaations and Feasibilitv
Studies Under CERCLA, OSWER D:i.rective'9355.3-01, OCtober 1988
51
. ....----.-.--.--.-. ----.. p_.
....-"..-.---.......--..- .
. . - -... _.....
-------
. - ~- .-. __h_. ._U_- ---..-. . -'.
assumptions as iftbey were design specifications.
Tbe comment incorrectly states that tbe clean-up goal for water
'prior to re-injection ia non-detect. Tbe JuDe 1"3 factsbeet and
tbe ARAR analysis in the feasi~ility study ~tb state tbat tbe
clean-up level prior to re-injection is maximum contamination
levels est~lisbed ~y atate and federal regulations. EPA will
,address otber "assumptions" tbe comment made aa tbey pertain to
tbe specific comment.
40.' Preliminary analysis suggests that the proposed' , ,
extraction/injection system may: cause a spreading of existing
contamination. This potential is of particular concern given the
lack of characterization of geologic and hydraulic
characteristics of the A-zone and given the lack of analysis of
the possible realistic response of the system. In particular,
the following issues have not been addressed:
o effectiveness of "clays" at the base of the A-zone on limiting
migration of impacted groundwater to the B-zone under,
reinjection.
o effect of heterogeneity of A-zone sediments'on engineering
control of extracted and injected groundwater, and.
o effect of changes on groundwater flow paths under the effects
of extraction and injection pulling contaminants into previously
clean areas.
Response: Tbe comment ~ases tbe concern for lateral and vertical
spread of contaminants on erroneous assumptions of wbat tbe
remedial design will look like (see response to comment 3'). A
properly designed extraction and reinjection system can prevent
lateral spread ~y, for example, placing tbe extraction wells on
tbe perimeter of 'tbe system. Vertical spread can ~. controlled by
monitoring tbe water levels in the reinjection welle sucb ~bat .
tbey do not place excess pressure on tbe A~zone clay. Latera~'
and. vertical spread, as well as otber design considerations, will
~e considered in tbe design pbase. %n addition, EPA plana to
pbase in tbe treatment system ~y installing and operating a small
n~er of extraction and reinjection wells, tben monitoring and
studying tbe A-zone groundwater's response to tbe system.
Tberefore, tbe contaminant extraction can ~e maximized and
contaminant spread can ~e minimized.
Tbe comment also suggests tbat geologic and bydraulic
cbaracterization of tbe A-zone is insufficient. EPA contends
tbat sufficient information exists to proceed witb a pbased-in
system. Tbe geology under tbe site is very beterogeneous and
as a consequence not every minor geologic feature can be known.
Tbe amount of information known and presented in tbe remedial
investigation report is sufficient to. make a decision and. to move
52
-------
~
into the next phase of the clean-up.
41. The "sandy clay" layer which forms the base of the A-zone is
an imperfect seal. In general, the understanding of the
continuity of the clay is based upon very little data (less than
40 data points). At all locations, boring logs prepared by
Kennedy/Jenks consultants describe the Layer 4 clays as "moist",
indicating that. groundwater penetrates these silty to sandy
clays.
Mounding pr04uced near the injection wells would increa~e the
vertical leakage potential at these locations and thereby would
increase the potential for increased uncontrolled vertical
migration of chemicals into soils beneath the A-zone. For
example, if the head in a given area were doubled as a result of
injection, the rate of groundwater flow through the clay would
double.
Therefore, given the heterogeneity of the unit and the general
permeabilities of the silty/sandy clays within the unit, it is
possible that injection over a wide, regularly-spaced grid into
the A-zone may actually result in a net increase of migration of
contaminants to the B-zone.
Response:
See response to comment 39 and 40.
42. If injection wells are placed at locations where highly
heterogeneous conditions occur, then it is possible that the
proposed "regular spacing" of extraction wells may not
effectively capture the additional groundwater flow due to
injection, especially for injection wells placed at the perimeter
of the extrac~ion well network. Given the fact that no
conceptual geologic or hydraulic model has been presented which
provides a basis for evaluation of these potential effects, there
is no information by which to judge whether the system may cause
additional lateral spreading of existing contamination.
Response: The comment confuses the assumptions used for the cost
estimate, sucb as "regular spacinq", as. remedial design
specifics. .Se. response to comment 39 and 40.
I
I
!
I
43. Since no "hydraulic analysis is presented for
operation of the system might have on groundwater
is possible that operation of the system may draw
laterally into previously clean areas.
the effects the
flow paths, it
contaminants
Response:
See response to comment 40.
44. No critical analyses of the potential short-term or long-
term effectiveness of the proposed groundwater
extractionlreinjectio.n system are presented in the Draft FS. In
order for any evaluat10n of the potential effectiveness of the
53
-". .~...,.-
I'
- _. .__.___n. .~. - -- _._- --..---'-'--'.- .. . - .
-------
.. .._.__."....~---_. ...
.....---.. .--.___e.~- ..-.-.---..--.--p --.---'---'.'---.'
proposed remedial system to be considered complete, a detailed
conceptual model which addresses the geologic and hydraulic
controls on the migration of chemicals and demonstrates good
agreement with existing data is necessary.
The following iS$ues would need to be addressed in order to
demonstrate the potential effectiveness of the proposed remedial
measure:
o effect of limited volumes of groundwater on system performance,
o predicted capture zones of 'individual wells,
o effectiveness of reinjection as a method for removing chemicals
from A-zone sands.
Response: Again, ~he commen~ seeks specific remedial design
details not required in ~he RI/FS, such as how ~he
extraction/reinjection system would handle variations in wa~er
levels, temporary dewatering, low flow rates, etc. (See responses
to comment 39 '40). EPA expects ~hat there may be temporary
dewatering and variable flowrates in the A-zone groundwater and a
system will be designed that is flexible and able to respond to
these condition. .
45. The Draft FS should evaluate how the proposed system
deal with dewatering, low flow rates, large variations of
level in short periods (i.e., following rainfall events),
intermittent operation.
would
water
or
Given the lack of understanding regarding the sources of water to
the A-Zone and the fact that the ,saturated thickness is low
(ranging between approximately 4 to 10 feet during the period of
time monitored to date - 1991 to 1992), additional evaluation is
necessary to determine if groundwater extraction could be
maintained in the A-Zone. If the sources of water to the A-Zone
are intermittent (irrigation percolation or seasonal rainfall),
continuous extraction may not be feasible. Hydrographs for A-
zone monitoring wells on the adjacent to the site presented as
Figure 3.10 of the Draft RI illustrate a steadily declining
groundwater thickness over the past two years. It is possible,
if the source of the A-Zone water is irrigation infiltration and
if irrigation is curtailed in the local area, that the A-Zone
will naturally dewater within 1 to 2 years. If so, there is no
reason to install and operate a groundwater extraction system in
the A-Zone. In any case, there is no analysis to confirm that
the effects of pumping will not result in a dewatering of the A-
zone even with the proposed reinjection of all produced water.
No discussion is included in the Draft FS that would suggest how
the proposed system would deal with dewatering, low flow rates1
large variations of water levels in short periods (i.e.,
54
-------
following rainfall events), or intermittent operation. Given the
existing data, it appears likely that all three of these
circumstances could occur during operation of the system.
Variations in the flow rates and chemical concentrations would
also impact the performance and costs of any associated, treatment
facili ty . .
Response: See Response ~o comment 44 wi~h regard ~o dewatering
comment. The comment s~a~es that if irrigation is curtailed, the
systam will na~urally dewater in 1 ~o 2 years. without any
backup for such claim, i~ is difficult for EPA ~o respond
quantitatively. However, limited containment options such as
purchasing and capping adjacen~ farmland was considered in the-
development and screening phase of ~he feasibility s~udy (see
Feasibility S~udy R.epor~). It was rejected because it would
require indefini~e main~enance of a possible large area. EPA is
also concerned tha~ with ~he removal of ~he irriga~ion, ~he A-
zone groundwa~er under ~he si~e migh~ qe~ water from o~her
sources such as irrigation and infil~ration from ~he residential
area ~o ~he eas~. -
Also, EPA believes tha~ ~rea~en~ is the preferred means by which
principal threats, such as ~he A-zone groundwater, are ,addressed.
containment is reserved for situations where there is large
volumes of low concen~ra~ions or where trea~ent is not possible.
46. The capture zone calculations, in a memorandum from Ralph
Lambert of Ecology & Environment to Cynthia Wetmore and Tom
Huetteman of the u.s. EPA, dated J November 1992, utilized the
Modified Nonequilibrium Equation (Cooper & Jacob), and the
Equilibrium Well Equation. The Modified Nonequilibrium is a
simplification of the Theis equation which assumes a uniform,
infinitely extensive and "confined" aquifer. The A-zone is not a
confined aquifer and the use of'the MOdified Nonequilibrium
Equation is inappropriate. . The Equilibrium Well Equation does
assume unconfined aquifer conditions; however, this equation is
not appropriate because the simplifying assumptions are not met.
RespoD.e: BPA agree. that the A-zone groundwater is a unique
feature that classic equations will no~ apply. In fact in ~he
ci~ed mamo i~ s~a~es " please keep in mind ~hat ... ~he water
bearing zone at BiB does not meet many of ~he conditions ~ha~
~hese formulas were based." Bowever, for ~he purposes which the
calculations were used, namely the cost estimate, the modified
non equilibrium equation is suitable. The reason for calculating
the capture zone in ~he FS is ~o make an estima~e as ~o ~he
number of wells solely for cost purposes. In ~he remedial
design, a cap~ure zone will be calculated based on the well
design and other necessary considerations chosen in the remedial
design phase. -
55
. "-'-~ .._-.-.-
'"--..- ----'-.'---'-"-'-"-'
-------
. -'u_'-'~ '~-._.~_._.... -....------ ..__...-.. - . -".
-- .----"'-------'--'--'.'
47. No evaluation of the effectiveness of the proposed
extraction/injection process at redu~ing contaminant
concentrations to the desired levels was presented. In fact,
recent studies have shown that pump and treat methodologies are
not effective at reducing contaminant concentrations where high
concentrations of chemicals adhere to and desorb from fine
grained sediments. Given the properties of the contaminants in
the A-zone and the heterogeneities of the geohydrologic setting,
it will be difficult to demonstrate whether the operation of the
proposed system would be capable of producing a reduction of
chemical concentrations in the saturated A-zone which would be .
cost-effective at achieving the EPA goal of protecting. the B-Zone
groundwater. .
Response: The comment questions the effectiveness of pump and
treat where chemicals have adhered to the 80il within the
groundwater. Its reference to studies indicating ineffectiveness
is somewhat misleading. Studies done at long-term pump and treat
superfund sites have shown that initially large amounts of
contamination can be removed. Bowever, after the initial
success, pump and treat is not effective at reducing
concentrations to low levels when chemica18 are adsorbed to the
soil. Simply put, pump and treat systems reach a point of
equilibrium, not necessarily at health-based levels, and further
reduction in contamination concentrations is difficult.
EPA contends that extract and treat in the A-zone groundwater
would be effective. First because of the high concentrations in
the water, there is a significant mass of contamination not
adhered to the. soil. Second, the chemical dinoseb i8 vater
soluble and would not as readily adhere to the soil as other
chemicals. EPA believes that a significant mass of chemicals,
especially dinoseb, can be removed. If, however, there comes a
point in the system operation where it appears that extraction is
no longer effective, EPA will re-evaluate the entire site clean-
up operations. . .
48. The proposed extracted groundwater treatment system is. an
innovative technology. Although limited pilot testing was
conducted by EPA, the evaluation of the design of this system is
inadequate in several key areas:
o Chemical Compatibility,
o Well Design,
o Chemical and Biological Fouling,
o Hydraulic Capacity, and
o Treatment Capacity.
56.
-------
Each of these is discussed in more detail below.
Response: EPA studied the use of UV/OXidatioh for the site at two
different tests. The first vas. full scale on-site operation of
the system during an emergency response action in which dinoseb-
contaminated rinsewater was treated. The second was a
comprehensive remedy selection treatability study conducted to
evaluate treatment effectiveness of uv/oxidation with respect to
EDB, 1,2-DCP, DBCP and other key volatile compounds using A-zone
qroundwater. The tests shoved that OV/OXidation can ~reat the A-
80ne qroundvater to maximum contaminant levels and provided more
specific information to be used during the remedial design.
The purpose of treatability studies during the feasibility study
phase i& to reduce cost and performance Uncertainties to
acceptable levels so that a remedy can be selected and to support
the remedial design of the selected alternative. EPA contends
that the treatability studies were adequate for the above-stated
goal. The remedial design phase of the clean-up process is
intended to detail and address the technical requirements needed
for the remedial action. For example, the selected remedy calls
for a RCRA-type cap. The details for this cap such as number of
layers, cap thickness, specific geotextiles, etc. are to be
determined in the remedial design.
49. Dihaloalkanes are generally degradable to Hcl and CO2
(Ollis, et. ale 1989);- however, those with saturated bonds, e.g.
EDB, are more difficult to treat using UV oxidation. Longer
reaction times and/or more powerful UVlamps are necessary to
attain effluent quality standards. The Draft FS should evaluate
other treatment technologies which may be more efficient.
..
The enhanced oxidation system did not remove chemicals to non-
detectable concentrations. It is not clear if additional tests
will be.run to determine if non-detectable concentrations are, in
fact, achievable or if the power and residence time required to
achieve non-detectable concentrations are consistent with
extrapolations. ..
Response: Pumigant& such as EDB and 1,2-DCP were expected to be
limiting agents by the treatment engineer; therefore the
previously discussed remedy selection treatability study was
conducted to evaluate the effectiveness of ov/oxidation
treatment. The treatability study data shows that BDB can be
treated to maximum contaminant levels. As stated in the proposed
plan, the clean-up goal for the treated vater is maximum
contaminant levels, not non-detect concentrations (see comment
39).
50. The u.s. EPA extraction/injection well design assumes a
gravel pack radius of ten ~ and a length of 20 feet. Thesoil
57
"'----'-'--..._._-~-. "--- ...-.._~._....----_._.
-------
. . .--.,.- ." _h.. _. .'--. ..~
~.- p-- -. .~..- ."'-"----'--'. .-.--.-....----...-. ......
boring required to install the well would be 20 feet in diameter
to a depth of 75 to 80 feet, which technically is not feasible.
It is assumed that the 10 foot gravel pack is a mistake and that
there was no intention of installing a well of this design.
However, the fact that it was utilized in a basic cost.
calculation illustrates the level of inadequacy of this document.
Response:
See response ~o FS commen~ &2.
51. Articles by Camp, and Nyer and Bitter (Camp, 1991; Nyerand
Bitter, 1991) indicate laboratory. or short duration field tests
of Advanced Ox~dation Processes ("AOP") with clean tubes can
result in unrealistically favorable results compared to operation
in a deposit-prone field installation. Solarchem is recommending
that OAP on Brown & Bryant groundwater be performed at a pH of 3.
This low pH is likely to prevent biological fouling and scaling
in the OAP reactor; however, the Draft FS should evaluate the
potential for scaling to occur downstream (e.g., in the
reinjection wells) when the groundwater is neutralized.
Water quality issues regarding the reinjection of treated
groundwater have not been considered. These issues include: the
difference in general mineral chemistry of native groundwater and
treated groundwater; and the potential effects of fouling due to
reinjection of treated water. A precipitate could occur in the
water treatment train, in the reinjection pumps, at the well
screens, or in the formation. Thus, there exists a substantial
potential that the proposed reinjection scheme would not work.
Response: The comment again expects a feasibility study repor~
to contain engineering considerations which will be addressed in
the remedial design phase. Solarchem has a proprietary design .
for wiping ~he UV lamps in ~heir ~reatmen~ sys~ems. since bio-
fouling and deposi~ion of insoluble salt could lower UV light
intensity, this issue was discussed between Solarchem and BPA
treatment engineers.
Low pH around 3 favors UV/OXidation ~reatment. pH sligb~ly higher
tban 4.6 (pRa of dinitrophenol compounds) would shif~ the
equilibrium to the direction of favoring s~abilization of
dinoseb's conjugate base which increases the treatmen~
effectiveness of horizontal flushing due ~o increased 8olubili~y.
Thus, the injectate will be readjusted ~aking into account mizing
wi~h ~he A-zone wa~er wbich may be a~ a higber pH. ~he
readjustment does no~ necessarily resul~ in pH above neu~rali~y.
The pH adjustment will be determined in ~be remedial
design/remedial action phases. BPA believes ~hat these minor
adjustments and optimization will no~ alter remedy selection.
Although the pH in. ~he injec~ate will not be grea~er ~ban
neutrality to trigger precipi~ation should precipi~ationoccur, a
simple additional module with flocculation or ion-exchange
capacity can be introduced.
58
-------
52. Mineral satUration calculations using chemical analytical
data for major ions in native groundwater, contaminated
groundwater and treated groundwater have not been conducted for
the recommended groundwater reinjection system which is common to
all action alternatives in the Draft FS. The calculations are
necessary to evaluate the potential for precipitation of minerals
during' treatment or in. some portion of the reinjection system.
Response: See response to comment 51. Calculations necessary to
evaluate potential fo~ precipitation _ill ~e made in the remedial
desiqn.
53. Data presented in the FS are not of sufficient quality to
perform mineral saturation calculations for contaminated
groundwater; an analysis of projected treated groundwater
chemical parameters is not included in the Draft FS.
Specifically, anion and cation charge balances given in Table
4.14 of the FS for wells clearly in the contaminated zone, range
from 27% to 80% difference. Analyses with such significant
charge imbalances should not be used for calculations regarding
water-mineral equilibria. In addition, the identity and
concentration of the analytes not included in the chemical
analyses could strongly influence mineral saturation
calculations.
Response:
See response to comments 51 and 52.
54. The EPA did not consider the likelihood of mineral
precipitation in the proposed treatment system at even a
rudimentary level. Simple manual calculations performed
utilizing the results of analysis of water collected from
monitoring well AP-01 demonstrate that the native groundwater is
slightly over saturated with respect to carbonate minerals (both
aragonite and. calcite) , as well as magnesium-calcium carbonate
minerals. Given this oversaturation, .when native or contaminated
groundwater is pumped to the surface and treated, precipitation
of carbonate minerals will likely occur. . .
Response:
See response to comment 51.
55. The Draft FS does not discuss changes in the treated
groundwater due to the proposed treatment process. Addition of
strong acid (sulfuric) and strong base (sodium hydroxide) will
change the water chemistry by addition of 570 mg/L sulfate and 26
mg~L s~dium (Solarchem Envir~nmental Systems, November 1992).
Th1S w111 further the potent1al of sulfate, containing calcium or
magnes~um ~i~era~s, to ~recipitate. Potential regulatory issues
regard1ng 1nJect10n of 1ncreased sulfate concentrations to the A-
zone aquifer are also not discussed.
Response:
Based on the groundwater c~assification exemptions in
5'
..-.- -----------~--- ___'__0_.__._----- -_..,_.~ --- -".- ...
--.- _. -.
-------
. '.. ---. . - ....- -' ..- .
.' .. ......-.-." .. --~.,_. --._.__--n___...' ..- -'--.-'.'-'" -
Resolution 88-63, the A-zone groundwater is not considered
suitable or potentially suitable for municipal or domestic water
supplies. Futhermore, it is the policy of the california
legislators that activities which may affect the quality of water
shall be regulated to attain highest quality of water which is
reasonable considering all the demands made on that water. Given
that, it is reasonable to reinject sodium and sulfate at the.
levels anticipated considering the limited demands on the A-zone
groundwater.
From a risk, toxicity and health perspective, there is no
negative impact on the quality of water with respect to sodium
and sulfate. Sodium effects the hardness of water; sulfate may
effect clogging of pipes. Existing average concentrations in A-
zone wells of sodium range from tt to 387 mg/l and of sulfate
range from 62 to 1814 mg/l. The MCL for Total Dissolved Solids
(TDS) is 500 mg{l; current average levels of TDS in A-zone wells
range from 615 to 12000 mg/l.
56. The UV/Oxidation design report discusses that anions such as
carbonate, nitrate, and chloride can inhibit the UV/oxidation
reactions desired. The report further discusses that analyses
performed on water from the site were within the acceptable range
for treatment. However, the actual analytical data were not
presented in the treatability study nor were the specific
groundwater wells from which the groundwater was collected
identified. Considering the wide range in anion concentrations
reported for groundwater collected in A-zone wells from the site,
it is not clear if a sufficient assessment of the potential
inhibition of UV/oxidation was conducted.
Response: The comment incorrectly refers to the remedy selection
treatability. study report as a design report. Solarchem has
quantitatively analyzed water samples from Brown' Bryant for
anions and has determined that the anion concentrations will not
negatively influence UV/OXidation effectiveness. SPA treatment.
engineer discussed matters concerning treatment effectiveness as
a function of various anion concentrations with Solarchem and has
concurred with the vendor's ,technical determination. Since the
anion concentrations were not a determining factor leading to
selection of the remedy, the National Contingency Plan section
300.800 does not require such technical detail to be included in
the administrative record.
. .
57. There is no discussion regarding the proprietary catalyst
and reinjection of the treated groundwater. The form of the
catalyst, particulate or dissolved is not discussed. If the
patented catalyst is a solid, removal prior to reinjection of
treated groundwater is necessary to prevent clogging. If the
catalyst is a dissolved chemical(s) it can effect the water.
chemistry, mineral equilibria, and the. consequent potential
60
-------
impact on reinjection in the formation can be critical. Potential
regulatory issues regarding injection of this proprietary
additive are also not discussed.
Response: SPA treatment engineer has determined that the use of
solarchemproprietary additive will not endanger human health and
the environment nor result in unmanageable engineering
difficulty. See response to comment 55 and 61.
58. The selection of a 10 gpm treatment system for remediation
is inconsistent when the conclusion that the A-zon~ extraction
wells will extract groundwater at a rate of only 100 gpd. If
half of th~ seventy-two wells are operating as extraction wells,
the total system flow rate would be approximately 2.5 gpm. A 5
gpm treatment system would seem to be adequate.
Response: Final flowrate will be determined during the remedial
design. A 10 gpm system was used for the FS cost estimating
purposes. Ass~ptions used for cost estimates should not be
confused with remedial design specifications. See response to
comment 60 regarding purpose of FS cost estimate.
59. The selection of a UV oxidation treatment system which
treats contaminated groundwater to non-detectable levels is
listed in the capital cost summary rather than a treatment system
which treats contaminated groundwater to MCLs or ten times the
MCL level. The capital cost treatment capacity is not consistent
with the preliminary remediation goal for the A-zone discussed on
page 3-3 of the FS report.
Response: SPA agrees the capitol cost associated with the
treatment system should be $492,000, instead of $650,000. The
cost estimate" in the Record of Decision reflects this change.
60. Appendix A of the Draft FS - Assumption for Cost of
Alternatives - presents some details of the cost estimates which
were used by the u.s. EPA to develop Up-front, Annual and Present
Worth Costs listed in Table 5.8 of the Draft FS report. Many of
the capital costs listed in Appendix A cannot be confirmed
without additional information. Assumptions are identified;
however, no explanation is provided for many assumptions.
Appendix A also contains errors in the calculation of costs.
Given the previous discussions regarding the shortcomings of the
technical evaluation of the proposed alternative, it is
impossible to determine whether the costs in Table 5.8 are lower
or higher than what it may ultimately cost to install and operate
such a system as that recommended.
Several specific instances of inaccurate cost assumptions are
presented below.
Response:
Before each cost estimate comment is addressed,
61
the
.. -. ._.-...... - ---'------------- -"-.. _.-- - -----
. -. .,. ---..~ .-- - -- _.
-------
." - .._-- ~ -._.. -- -- ...- ...- .. ...
- ...-..---.. -...-.... . ... --. ...... .--....--.....-.---.--..-. _.-.... ....- -.-'---.-.-"'.--'" ...-
purpose of the cost estimate during the feasibility study should
be clarified. It appears that the commenter believes this cost
estimate should have the same level of detail as required in the
remedial design. SPA quidance states that "typically, these
'study estimate' costs made during the FS are expected to provide
an accuracy of +50 percent to - 30 percent 113. Clearly, the
o cost estimate is to provide a basis for comparison of the
alternatives. EPA was conservative when cost estimating and
typically chose the worst plausible scenario when estimating
costs. A more detailed cost estimate vill be provided during
the remedial design~ . . .
61. The requirement of the use of proprietary additives and
extended reaction times make the advanced oxidation process
("AOP") potentially costly for groundwater applications with a
low flow rate and high concentrations of the more refractory
compounds, such as EDB.
The reported cost of AOP appears to be high relative to other
treatment options. Based on the design criteria presented on
page 3 of the Solarchem Design Test Report, GAC usage would be
approximately 6.5 lbs GAC per 1000 gallons of water treated. The
GAC usage rate is controlled by 1,2-dichloropropane (1,2-DCP).
Assuming a replacement cost of $4.50 per lb which includes
contractor oversight during replacement, the cost of GAC
treatment at Brown & Bryant is estimated to" be on the order of
$30.00/1000 gallons of treated water. .
This GAC usage cost is certainly competitive with the AOP
operating costs estimated in the Draft FS. It should be noted
that the AOP treatment cost of $83.18/1000 gallons estimated by
Solarchem did not include contractor oversight.
The $30.00/1000 gallon GAC and the $83.18/1000 gallon AOP are
solely treatment costs and do not include labor costs for. routine
treatment system operation. .
Response: The comment did not reference the source for his
assumptions and cost estimates for GAC. Bowever, SPA contends
that the actual cost for GAC could be significantly more. I~
general, GAC is most cost-effective for high volume and low
concentrations. The A-zone groundwater at Brown' Bryant is lov-
flow and very high concentrations.
EPA used the Freundlich isotherm relationship to estimate the
carbon use rate. These adsorption isotherms are useful screening
tools for determining preliminary carbon useage rates but have
several drawbacks that might underestimate actual carbon use
EPA Guidance forConductina Remedial Investiaations and Feasibilitv Studies
Under CERCLA, OSWER Directive 9355.3-01, October 1988
62
-------
rate. The adsorption isotherms 40 not take into account
competitive a~sorption of mUlticontaminants, nor other organic
material that might interfere with adsorption~ EPA estimated
that car~on useage rate for 1,2-DCP, under ideal conditions and
assuming it is the only contaminant, would ~e approximately .3
1~s/1000 gallons. Car~on useage rates of 3 1~s/1000 gallons and
6.5 l~s/~OOO gallons is considered extremely high rate for GAC
and would result in short ~ed lifes.
. .
The major OiM costs associated with GAC is car~on replacement
which increases as ~ed life decreases. Assuming ~.6.5 1~s/1000
gallon useage rate, it is estimated that the annual car~on
replacement rate would ~e 85,000 l~S{year. A GAC system for the
Glendale South pump and treat system estimated its car~on
useage rate at 90,200 ~s/1000 gal. Adjusting the Glendale cost
estimate to reflect conditions at Brown i Bryant, the.GAC capital
cost for Brown i Bryant is estimated at $344,000. (uv/oxidation
capital costs are $492,000). GAC O&M costs, excluding operator
costs, are estimated at $192,000 annually. Car~on replacement
cost at Brown i Bryant would ~e higher than at Glendale ~ecause
the car~on could not ~e regenerated on-site like at Glendale, and
would require proper handling, treatment and disposal at a
hazardous waste facility. This is not accounted for in the
$192,000 OiM estimate. (uv/OXidation OiM costs, excluding
operator, are $230,000 annually) Operator costs for uv/oxidation
are assumed to ~e slightly higher than GAC.Bowever, GAC
operator costs would ~e more than typical due to the frequent ~ed
changes a GAC system at Brown & Bryant would require. In
summary, GAC is not half the cost of uv/OXidation as the comment
. alludes ~ut is pro~ably slightly less. .
In addition to cost, there are many other factors EPA is required
to consider when selecting the alternative. CERCLA S121(~)
states "Remed.ial actions in which treatment permanently and
significantly reduces the volume, toxicity or mo~ility of the
hazardous substances .. asa principal element, are to ~e
preferred over remedial actions not involving such treatment."
uv/OXidation would reduce the toxicity and volume through
treatment. GAC simply transfers the co~tamination to another.
medium. In addition, "The President shall select. a remedial
action that is protective of human health and the environment,
that is cost-effective and that utilizes permanent solutions and
alternative treatment technologies or resource recovery .
technologies to the maximum extent practica~le." uv/oxidation is
considered an alternative treatment; GAC is not. .
Bowever, SPA will evaluate further the possi~le use of GAC during
the remedial design. If, as expected, the cost ~etween the two
4 nFea~ibility Study for the Glendale Study Area South
Operable UnJ.t",. Auqust1992.. (San Fernando Valley Superfund Site)
63
. --.-.. ..-".-....--.-.-
. ..- .~ - . '.." ...-
- "- ...--. . .
-------
..-_..... ._-~_. .- -'.---.-'--'-..--.--. "h...._.--.....-.-
systems are comparable, UV/oxidation will be used as the primary
treatment.
In addition, EPA will evaluate the use of GAC as an addition to
UV/OXidation. Chemical destruction efficiency using UV/OXidation
decreases as the concentrations decreases. Therefore, it might
be cost-effective to treat the majority of contamination using
UV/OXidation, then reduce concentrations to HCLs using GAC. The
point at which treatment technology can be switched from
uv/oxidation to GAC will be determined in the remedial design.
EPA will weigh the cost-effectiveness and the amount of chem~cals
,destroyed when determining when or whether to add GAC to the
treatment train.
The Record of Decision states that the selected method of
treatment is UV/Oxidation; however, if after completeing a
detailed cost analysis of the two treatments and GAC is
significantly cheaper, EPA will re-evalute its decision of
UV /oxidation..
Finally, the comment states that the propriety additive is a
requirement. EPA would like to clarify that any UV/OXidation
system that meets the performance standards would be acceptable.
more
62. The gravel pack cost calculation appears to assume a radius
. of ten feet and a length of 20 feet, which results in a unit cost
of $5,000 for the gravel pack for each well in u.s. EPA's cost
estimates (U.S. EPA, 1993). A tw~nty foot diameter soil boring
is technically not feasible to a depth of 75 to 80 feet. If the
gravel pack diameter is actually one foot, when the unit cost for
the gravel pack for each well is approximately $50. The total
cost per well is listed as $20,000; however, by deleting these
gravel costs; the total cost per well would be reduced to
approximately $15,000.
Response: The comment identified a typo in the cost estimate.
The typo did result in a miscalculation. However, the cost
difference because of the typo is not significant. The cost is
corrected in the Record of Decision.
63. The capital cost for surface pumps, electrical controls and
tanks is assumed to be equal to 100 percent of the capital piping
costs; however, no explanation is provided for this assumption.
Additional costs which may be encountered for locating piping and
wells on off-site properties an across roadways and railroad'
tracks are not considered in the Draft FS.
Response: The cost of extraction field piping and surface
equipment costs were estimated by using the rule of thumb of 100
percent of capital costs. This rule of thumb is used by the EPA
cost estimator at the Engineering' Risk Reduction Laboratory .in
Cincinnati. During remedial design a more specific cost estimate
64
-------
will be done.
64. The. analytical cost for on-going monitoring is listed as
$1,000 per sample; however, the analytical method and the
compounds being analyzed for are not identified. The number of
sampling rounds are also not specified.
Response: The cost estimate for on-going monitoring is based on
current monitoring at the site.
65. The treatment system capital cost is listed as $650,000.
This corresponds to the capital cost associated with a ten gpm
treatment system which meets non-detect levels in the effluent,
described in the 20 November 1992 EPA Memorandum from Vance Fong
to cynthia Wetmore. Although this. memorandum cites the Solarchem
Design Test Report (Solarchem, November 1992) regarding the
source of this construction cost estimate for a 10 gpm system,
the Solarchem Report provided only a cost estimate of $119~000
for a 1 gpm system. No information is presented on the manner in
which capital costs were scaled to account for different design
flow rates.' '.
Response:
See response to comment 58 and 59.
66. The location of the precipitation front and hence scaling
will depend on factors that include: contact time and agitation
with respect to air where carbon dioxide gas exchange will occur,
and addition of ,acids and bases during treatment. A precipitate
could occur during the addition of sodium hydroxide after
UV/oxidation in the water treatment train, in the reinjection
system pumps, at well screens, or in the formation. Such fouling
could lead to increased operating costs for the system and well
maintenance as well as to reduced averaged reinjection flow rates
if well or formation fouling occur.
Response:
See response to comment 60.
67. U.S. EPA advises' in the Draft Remedial
Investigation/Feasibility Study Report that it will examine the
necessity for'the feasibility of B-Zone groundwater remedial
alternatives in a Second Operable Unit for the Site. SPTC and
Santa Fe recommend that the selection and implementation of an A-
Zone remediation program be deferred until the completion of the
Second Operable Unit. First, there is substantial question about
the feasibility of the proposed A-Zone extraction system due to
dewatering and other technical factors. Second, the B-Zone
groundwater evaluation will assess the interplay between the A-
and B-groundwater zones to evaluate the best remedial alternative
to protect the B-Zone groundwater (the zone the EPA seeks to
protect). Assessing the groundwater remedial alternatives in
such a comprehensive fashion in the Second Operable Unit will not
increase any risks of exposure, sinc~, as EPA acknowledges in the
65
._-,'.- ...-- ---..--.-.----.---..
'---. --. - --.. --'''-..--
...-- -.---..... 'UU_-'-"'-
-------
- .- .- .-..-. .._. ..--
'_'_'''_4_'_'_.--'-'---.---'.'--''''- . .
Report, the primary concern of the A-Zone groundwater is its
impact on the B-Zone, and the B-Zone groundwater is not currently
impacting drinking water supplies and is not expected to do so in
the near term. .
Response: EPA does not agree that the decision on theA-zone
groundwater should be deferred until after the B-zone is
characterized. Data collected to date show that the A-zone is
leaky and has caused contamination in the B-zone groundwater. The
mass of contamination in. the A-zone is significant. The rate. of
leakage, although not exactly known, is sufficient to cause
levels in the.B-zone to exceed maximum contamination levels. If
left unabated, the A-zone groundwater will continue to be a
source of further contamination to the B-zone groundwater. The
investigation of the B-zone groundwater will not alter the need
for action in the A-zone groundwater.
The B-zone groundwater has dramatically different characteristics
than the A-zone groundwater. It can sustain higher flowrates and
currently has' lower contamination levels. As a consequence, a
more conventional pump and treat system can be used. .
Alternatives for the B-zone will include hooking into the
existing system, as well as other types of treatment.
Comments from Canonie Environmental on Behalf of Holland &
Griffin:
68. Page 2 Second Paragraph: A-zone mounding is referred to in
the southwestern corner of the site. Figures 3-6 and 3-7 show a
ridge in this area, but no mounding.
Response: The use of the term "mounding" to describe the
groundwater flow pattern was not intended to imply only a
circular pattern. On page RI-3-S a more complete description of
the feature is given.. .
69. Page 2 Second Paragraph: States that "from a slug test the
groundwater velocity was estimated at 53 feet/year.. II The
statement implies that velocity can be determined directly from
the slug test, which is performed to measure hydraulic
conductivity. other parameters, such as gradient and porosity,
must also be measured. This should be stated or referenced in
the text. .
Response: The groundwater velocity is based on a hydraulic
conductivity of 4 x 10-4 em/see derived from the slug test, an
effective porosity of 26% estimated from literature values, and a
gradient of 0.034, which is a localized gradient measured from
the wells used in the pump test. A reference to the slug test
report is included in section 3. of the RI report. .
66
-------
70. Page 2 Fourth Paragraph: The dinoseb spill area is referred
to in this paragraph as the principal hot spot for surface soil
contamination and the only location where high concentrations of
dinoseb were found in the construction zone. A major omission of
this report is that this area has been at least partially
remediated. This is only briefly referenced in the RI. The FS
(pages FS-1-1 and FS-3-10) states that in 1991, approximately 80'
cubic yards of the most contaminated material was excavated from
this area, treated by soil washing, and returned the treated soil
to the area. This activity is not adequately documented in
. either section of the report. The limtts of the excavation
should be defined, as well as verification sampling at the limits
of excavation, to confirm the extent of contaminated material
removed, particularly if material is subsequently replaced.
Response: EPA exc~vated and treated approximately 80 cubic yards
of the highest contaminated portion of the dinoseb spill .area.
The exact boundary of this area has not been mapped; however, it
has been marked in the field. The contaminated soil was dug to a
depth of approximatelY seven feet, treated with soil washing, and
then returned to the excavated pit. concentrations below. the pit
are still as high as 4,230 mg/kg (at 8 feet below ground
surface). Based.on soil boring data, EPA estimated that prior to
treatment, up to 650 cubic yards of soil in the dinoseb spill
area may exceed 80 mg/kg (see page FS-3-10). Following the
removal, up to 570 cubic yards remain. Bad EPA selected the
remedy that treats this soil, additional characterization of this
area would have been necessary to accurately determine the final
volume for cleanup. Such characterization might have occurred
concurrent with any excavation. EPA 40es not believe that
additional characterization is necessary for the RI/FS,
especially since the selected remedy does not involve additional
treatment of the soil at the surface or at depth. See also
response to comment 97.
71. ..page 3 Fifth Paragraph: References and areal extent, 5.5
acres at 50 micrograms per liter (ug/l) in relation to "target
concentrations" should be changed in discussions so that this is
not construed to be a cleanup concentration.
Response: More discussion on this point is provided in section 4
of the RI report.
72.
Page 4 Third Paragraph:
"Absorbed" should be "adsorbed."
Response:
error.
The comment is correct.
The error was a typographical
73.
Page 4 Fourth Paragraph:
Define "key site contaminants."
Response: section.. and 5 include discussions on how "key site
contaminants" were identified.
67
. .. --"._---------'--~----.'.~'---"'"
. .'n. __n- ~ _. -.- ._-. --. --..~.
-------
.-' --.. - - -. - - --.---..- - --" .- -- .._- --.
- --------.----------.---- --..- ----.-----.-......
74. Page 6 First Paragraph:
"MCL" should be "MCLs."
"Were" should be "where," and
Response:
Comment notec!.
75. Pag~ RI-1-2 Third Paragraph: The second stated principal
risk is poorly worded and. confusing: "the potential future risk
if site contamination were to reach current drinking water
sources or from the future use of potential drinking water.
sources that are currently or may in the future be contaminated
from the site." It should simply read: lithe potential threat to
drinking water sources." It was already stated on the previous
page that the B-zone groundwater is considered a potential
drinking water source for the purpose of setting cleanup
standards.
Response:
Comment notec!.
76. Page RI-1-2 Fourth and Fifth Paragraphs: "Contamination of
surface soils at B&B has resulted largely from spills and
improper housekeeping. During the RI, source areas for this
contamination were characterized."
"During the RI the principal source areas were characterized to
determine which locations 9n-site were and are currently
significant sources of contamination."
It is apparent that, in general, the RI, while. not relying on
previously-collected chemical data, used this information to
recharacterize the areas already known, rather than investigating
areas that may not have been established.. .
Response: In selecting areas for investigation c!uring the RI,
EPA relied on available site history anc! previous investigations
to target specific areas on anc! off site for sampling. During
the RI, EPA samplec! both areas that were anc! were not previously
sampled.
77. Page RI-1-4: The site description should include a
discussion of the potential source areas immediately south of the
Brown & Bryant fenceline, including the unlined waste water pit
identified in the u.s. EPA document TS-PIC-89826 dated September
1989. This pit was used as part of the adjacent potato shed
washing operations. Do DBCP residues from the potato-washing
disposed in the waste water pit cause soil and groundwater.
contamination? Was this waste pond closed in accordance with the
California Toxic Pits Cleanup Act?
There are several other potato-washing sheds with unlined waste
pits between the site and the Arvin water supply well. Are these
operations threatening the ~in water supply?
68
-------
The Arvin water supply well is on assessors parcel 192-12-8,
which previously had tenants who handled chemicals similar to
Brown & Bryant. Those tenants were: San Joaquin Buildinq Supply
company (1946-1957), Kinq Chemical, Inc. (1957-1961), and Bear
Mountain Dusters (1961-1969). A photoqraph of the buildinq on
the Arvin water supply well property is enclosed. Note the siqn
on the side of the buildinq indicatinq the type of chemicals and
supplies handled at the property.
Response: SPA did collect soil samples in the area of the former
waste pond as did Kennedy Jenks. None of the data from the soil
~orinqs located south of the sitesuqqest a source of-
contamination. As for other sources in the area, it is not
within the scope of this investiqation to investiqate and develop
cleanup remedies for other possi~le contamination sources. These
sources are ~est addressed ~y either state or local aqencies and
are not part of the Brown and Bryant superfUnd site.
78. paqe RI-1-4 Second Paraqraph: This paraqraph states that
canonie's closure plan does not have a date. The report date is
printed on the front cover -- March 1988. '
Response: Comment noted. The copy of this report provided to
EPA has a cover paqe with no date.
79. Page RI-1-7 sixth Paraqraph: "The data collected by these
(previous consultant) investigations were used durinq the RI to
identify areas of concern for additional sampling. None of the
analytical data collected by canonie or Harqis is presented in
this report because it is of unknown or questionable quality."
By the same rationale, the areas sampled should have been
questioned, and greater effort should have been directed towards
investiqatinq all potential source areas, includinq off-site
locations.
Response:
See response to RI comments 76, and' 77.
80. Fiqure 1.1 The fiqure does not depict any low-permeability
layerunderlyinq the A-zone .water-bearinq unit.
Response: This fiqure is a simplified cross-section intended to
identify the qeneral media of concern. A more detailed cross-
section is included in Fiqures 3-3 and 3-4.
81. section 2: In this section, which presents Remedial
Investigation Field Activities, no mention is qiven to the
emergency response cleanup work in the dinoseb spill area
conducted in 1991.
Response: This section presents a summary of field activities
for the remedial investiqation. Since the removal referred to in
the comment was not a data collection activity, it was not
69
- ..__.__._--_....,..-.,-----_.-..--_._----~ ..". ".,.--.
. --.-.."- - -. -...
-------
... -- - ---_..--- -_.~. - ._- .-- . ,
'.""-' -..'. .---.... -.---.
- ...----.... --------.--..-.-
included in the summary.
82. Page RI-2-3 First Paragraph: The executive summary mentions
that five groundwater sampling rounds were performed. This
paragraph mentions six months in which groundwater sampling was
performed. This should be clarified.
Response: . The executive summary is in error.
not five, groundwater sampling rounds.
83. Page RI-2-3 Second Paragraph: "The results for the B-zone
are largely inclusive because of an insufficient number of wells
and because the older on-site wells were screened over more than
one water-bearing unit in the B~zone; this zone will be
investigated further in a second operable unit RIfFS."
There were six,
It is not believed appropriate, based on the available
information, to screen only one sub-"zone" of the B-zone and
treat these zones as hydraulically~district.
Response: EPA has so far investigated discrete sub-zones in the
B-zone based on differences in hydrology and chemistry observed
between the different sub-zones. This level of analysis, EPA
believes, will help to define the portions of the B-zone where
contamination is at greatest concentrations. As a result, EPA
may be able to more effectively remediate the contamination.
More detail regarding the B-zone hydrology and chemistry will be
provided in the second operable unit RIfFS.
84. Page RI-2~8Second Paragraph: This paragraph states that
"analysis of soil samples for seven volatile organic chemicals
(see Table 2.2}..." Table 2.2 states six volatile organic
. compounds.
Response: The text is in error. The analysis was for six, not
seVen, volatile organic compounds.
85. Section 2 Figures: In general, the figures either do not
have figures, labels, or numbers, or have inadequate legends.
For example, if a figure depicts wells with WA and WB
designations, it would be appropriate to indicate the
designat~ons where the wells are screened, not merely that they
are wells. Figure 2.4 is not shown in a readable scale. It
would be helpful to locate all investigative sampling points on a
single drawing to indicate actual coverage of the investigative
programs. .
Response: The purpose of these figures is to locate the position
of the various soil borings, surface soil samples and wells
located at the site. EPA believes the figures meet this purpose.
Details regarding the depth or screened intervals are provided
elsewhere in the report. . .
70
-------
86. Page RI 3-3 Fourth Paragraph: states "The major groundwater
features within the Arvin area consist of a deep confined aquifer
which is located below the Corcoran Clay, and a shallower
confined aquifer (B-zone) located above the Corcoran Clay." It
is unclear whether this is taken from the 1964 or 1991 reference
in this paragraph. A confined zone becomes unconfined if water
elevations drop to below the upper confining layer. The RI is
presenting the ~ayer between the "B-1" and "B-2" zones as a
significant cor.:~ning layer, yet by definition, subzones within
the .B-zone are connected at least in limited sense.
Response: The appropriate reference for this statement is the
1991 reference. Xn borings conducted into the B-2 zone. the water
level is consistently measured above the top o~ the B-2 unit.
The water level at well WB2-1 is about 15 ~eet above the top o~
the aquifer. During recent pump tests, EPA also observed a high
barometric efficiency for the B-2 wells, which is a .
characteristic typical of a confined aquifers. Regionally,
these discrete zones may be connected.
87. Table 3.2: The total well depth should include depth-to-
bottom-of-boring because seals were not placed for many wells in
the interval between the well trap and bottom of the boring. For
example, bottom-of-boring for EPAS-~ is 93.5 feet below ground
.surface (bgs). The listed depth is 84 feet bgs, which is the
bottom of the trap. EPAS-4 may be a vertical conduit installed
by the U.s. EPA, which interconnects the B-zone with the A-zone.
Response: Comment regarding the table noted.
comment 89, 117, and 119.
see.response to
88~ Page RI~3-5 Fourth Paragraph: This paragraph states "A
possible limb or mound in the water table extends from the
southwest site corner, southward approximately parallel and next
to the railroad tracks. This groundwater limb has a similar
shape to the groundwater contaminate' plume which is discus'sed in
section 4." This should not be confused to indicate that these
shapes should be similar and are in agreement. Isoconcentration
contours and. water level contours should not necessarily match.
with the determined groundwater gradient, the groundwater flow
direction and, therefore, the contaminant plume should go toward
the trough shown on Figures 3.6 and 3.7.
Response: EPA agrees with the comment. The similarity noted in
the text between the groundwater flow pattern and the contaminant
concentration patterns in the A-zone groundwater is in reference
to the general direction of contaminant flow, which is to the
5B-~ Aau~fer Test Report. Task 11. Brown & Brvant. Arvin.
Cal1forn1a, Prepared for USEPA by Ecology and Environment
March 31~ 1993. '
71
. - _.._~. '--"----r---"'----:-....__-__n____- .""-"0-'.'-." .--- -
. _. ..". - .."----.
-------
-- ." -'~--~- ~.__.. .--'.~'-------'- --.-- ----_._-~. ~ .
south and west. The hiqh concentration ot DBCP at well EPAS-3 is
consistent with the trouqhtound in. this area.
89. Page RI-3-6 Third Paragraph: It is possible that EPAS-1 may
penetrate the clay layer underlying the A-zone, which is reported
as thin and not well defined. No boring log is provided for this
location or EPAS-3 in the Ecology & Environment, Inc. (E&E),.
Site Assessment Report (November 1990), which documents the field
program during which this well was installed, nor is complete
information provided in the u.S. EPA RIfFS Workplan. Complete
information on the monitoring wells installed by the.U.S. EPA
contractors should be provided in the u.S. EPA RIfFS. The bottom
of the sand pack for EPAS-1 is reported to be at a depth of 90
bgs on the well detail. This corresponds to an elevation of
approximately 339.71 mean sea level (MSL). The elevations on the
cross-sections shown on Figures 3.3 and 3.4 show that the clay
layer underlying the A-zone extends only to an approximate
elevation of 350 MSL. Further, the above-referenced report has
boring logs for EPAS-2 and EPAS-4 which indicate that the clay
layer may have been penetrated and sand pack placed across this
interval (instead of-grouting back to the proper zone). The
report also contains two different boring logs for EPAS-4. The
boring log that indicates a total depth of 92 feet bgs may be for
EPAS-l or EPAS-3. This boring log also indicates that the clay
layer may have been penetrated. Figures 3.6 and 3.7 show a
trough in the vicinity of EPAS-2 and EPAS-3, which may indicate
conduits at these wells. This should be investigated due to the
inadequate boring logs presented in the E&E report and the u.S.
EPA RIfFS Workplan. .
Response: The comment presents various issues which are
addressed in~ividually ~elow:
Boring loqs tor EPAS-l and EPAS-3 are missinq and can not ~e
located. There are two ~oring loqs for EPAS-4 ~ecause an
initial exploratory ~oring was conducted prior to the well
- installation.. The qeoloqy in the area of EPAS-l and EPAS-3
can be inferred from nearby soil ~orings. Soil ~orinqs
CA10, 11 and 17 and RBi 'HM and NN are located near to EPAS-l
and soil borings CAOS, 06, and 07 and CB04 are located near
to EPAS-3.
Due to the lack of water production in ~oth EPAS-l and AP-5,
EPA intends to ~andon these wells within the next year.
EPA believes that EPAS-l is most likely screened over the
clay layer, which accounts for the lack of water in this
well.
Based on an analysis of well loqs and soil ~rinq loqs from
the site, EPA ~elieves that wells EPAS-2, EPAS-3 and EPAS-4
are screened ~ove the base of the A-zone. However, EPA
acknowledges that the avaiable data is inconclusive but may
72
-------
indicate the bottom of one or more of the boreholes drilled
for these wells may be below the base of the A-zone.
See response to comment 117 and 119 in regards to possible
conduits.
90. Page RI-3-6 First Paragraph: The so-called "flatteninglJ is
only indicated by Well EPAS-4, which is located upgradient
without any adjacent wells to confirm trend. Across the site,
the gradients have not changed significantly. Although, because
the well elevation data is not validated, no groundwater gradient
maps can be properly constructed.
Response: While the flattening is more pronounced at well EPAS-.
4, EPA believes that the data shows does some flattening observed
over the site. Bowever, EPA did not intend to place a lot of
significance to this observation. See also response to comment
94.
91. Page RI-3-6 sixth Paragraph:
was reported as 10.4 to 10-5.
Response: The comment is correct.
4 to 10-6, not 10-4 to 10-5.
The permeability of the A-zone
The statement should read 10-
92. Page IR-3-7 First Paragraph: This paragraph states that
groundwater velocity was calculated using assumed porosity;
however, porosity measurements were said to be taken by E&E.
Response: At the time that the slug test was conducted, the
porosity data was not available, so a literature estimate of 26%
effective porosity was used. Bowever, this estimate is for
effective porosity whereas the laboratory measured total
porosity. EPA believes that the effective porosity value used is
consistent with the laboratory values.
93. Page RI-3-a Fourth Paragraph: This paragraph states that
the water levels in the new wells cannot be correlated to the
water .levels in the old wells because the old welis are screened
across several of the SUbzones of. the B water-bearing unit. As
referenced earlier in the comment to Page RI-2-3, second
paragraph, the. report has also stated that the results for the B-
zone are largely inconclusive. The O.6-foot difference cited in
water-level elevation between the "old" and "new" wells does not
indicate these zones are not hydraulically interconnected. A
pump test could be performed to make such a determination.
The 0.6-foot difference could also be due to surveying the
mon~toring wells to different elevation controls.
73
". . ...... _...~.- -. --'--..-'---'----.'--."..--.--..-. ..... .-.
.' "'-'. -...
-------
'0 .-- _._.__._._.._....__._-----~..~._-_.
. - - .. ._- . ... -. _n-- . ..
0_-'__-- - -.-...--- ---"-.--. _.-.
The potential fault that cuts the site could influence water
levels. The data may confirm the presence of the fault. Why is
the potential fault not shown on any map in the RIfFS?
The potential fault could isolate the Brown & Bryant site from
the Arvin water supply well.
Response: See response to comment 83, 86 and 9C. As to the
potential fault at the site, early in the investiqation there vas
speculation of a fault ~ased on the dry wells located on the vest
side of the site. Bowever, ~asedprimarily on litholoqical loqs
and E-loqs from soil ~rinqs conducted ~y EPA and Kennedy Jenks,
no evidence of a fault could ~e detected. EPA has not discussed
in detail the possi~ility of a fault in the RI report ~ecause of
the lack of evidence reqardinq its existence. The dry vells
located on the west side of the site appear to ~e dry ~ecause of
the position of the well screens relative to the water t~le (see
paqe RI-3-6).
94. Table 3.2: The table should make it clear that "survey
elevation" is measured to the top of well casing or measuring
point elevation. Ground surface elevation would also be good
information to list. This table does not correspond to Table 2.1
in the U.S. EPA Workplan. Which elevation data are correct? If
the wells were resurveyed, it should be discussed in the RIfFS.
What is the true groundwater flow direction in the A-zone and the
B-zone aquifers? If the elevation controls are incorrect, the
groundwater flow directions calculated from them will be
incorrect. Are the more recent monitoring wells installed by
Kennedy Jenks Consultants surveyed to the same vertical control
as the other monitoring wells?
Response: Durinq the RI, EPA had all of the EPAS and older vells
resurveyed and est~lished for all wells the north side of the
casinq as the elevation control. Kennedy/Jenks conducted ,a
separate survey of there wells. It can not~e verified at this
time that the same vertical control. was used for each survey.
However, when the newB-zone wells are installed for the second
RI/FS, all the B-zone wells .will ~e resurveyed and the A-zone
well surveys will ~e verified. '
95. Figures 3-2 through 3-7, 3-9: The source of the figures is
Ecology and Environment, Inc., 1993. There is no reference
citation for this contribution.
Response: The reference to Ecoloqy and Environment is intended
to record the oriqinator of the fiqures: it is not a reference to
a separate document. '
96. Page RI-4-3 Second Paragraph: This paragraph states "To
further identify areas for potential cleanup, Figure 4.1 also
identifies eight other samples with concentrations of dinoseb
7C
-------
between 8 mg/kg and 80 mg/kg." only four samples are shown on
Figure 4.1. Once a clean-up concentration, level of 80 mg/kg has
been established, it is not appropriate to cite lower
concentrations for cleanup purposes. '
Response:, The text is in error. There are only four samples
between 8 and 80 mq/kq, not eiqht. The reason for showinq these
samples in the figure is to characterize areas on-site,that may
approach the cleanup level. This data may be used to 1dentify
portions of the site for further characterization,during cleanup.
97. Page RI-4-3 Second, Third, and Fourth Paragraphs: The u.S.
EPA emergency response cleanup action in the dinoseb spill area
is not adequately addressed in the RI (action is mentioned only
in the FS). Approximately 80 cubic yards were removed from this
area, treated by soil washing, and replaced. The third paraqraph
states that five locations in the dinoseb spill area were above
the 80 mg/kg level. These areas may have already been removed,
which would only leave isolated surface soil "concentrations" as
a potential health risk.
Paqe RI-4-3 Fourth Paragraph: Again, addressing soils in the
"construction zone," the dinoseb spill area is the only area of
concern, and it may have already been removed. Further, it is
not likely that the construction zone, based on sewer line data,
would be used along this site boundary.
Response: The RI does characterize the dinoseb spill area prior
to the removal. Part of the reason for doinq this is to more
thorouqhly document the contamination problem that led to the
removal. ultimately, the final remedy selected does take into
account the fact that the removal reduced the risk from dinoseb
in the worst area of surface soil contamination (down to 7 feet)
at the site. ,EPA selected to only cap the southern portion of
the site in larqe part because the dinoseb problem had been
substantially remediated by the removal action. Furthermore,
treatment of the remaininq isolated hot spots was not selected
because of the small volume of contaminated soil exceedinq 8~
mq/_.,'
98. Page RI-4-8 Table: Target concentrations are misleading and
can be confused as cleanup levels. It is not appropriate to
estimate quantities in relation to target concentrations that may
not ultimately be used as cleanup goals. Further, the figures
generated in this section are all related to these target goals,
whereas standard concentration contours would be helpful to allow
simpler interpretation to any concentration limit.
Response: page RI-4-6 e~lains the rationale for using the
target concentrations in the report. on this paqe it is stated
that the concentrations "are not intended as clean-up levels. II
EPA did not draw contamination contour lines for these figures
75
-~--_.- .._-~_....__..--_._--_._....._.-. ,-.-
. ---... --"-.-----'-'--..-,.- or
-------
, - ~-------_..__._-----_.__. -'.
because EPA believes that such contours would be too subjective
due to the occurrence of multiple sources and the heteroqeneity
of contaminant concentrations.
99. paqe RI-4-18 Second Paraqraph: While it is important to
note the different screen intervals, data may be gathered from
these wells that may be compared. An inflatable packer could be
used to sample discrete intervals in the "old" wells. The
elevation data collected is likely correct. Well AR-Ol should be
video-logged to determine the screen interval.
Response:
EPA agrees and is considerinq the actions proposed.
100. Table 4.8: Units are not specified. In the first row of
the table, the average number given "is "38,8511.""
Response:
The unit, uq/kq, was mistakenly left off of the table.
101. Figures:" In general, several fiquresare missing a north
arrow. For figures depicting concentration contours, past the
date of the sampling event and concentrations detected for each"
well location. Contours should be in relation to standard
intervals (e.g., 10, 100, 1,000) instead of target concentration
intervals (e.g., 7, 70, 700). Figure 4.9 shows a pattern that
could be indicative of an off-site source. This is not discussed
in the text of the report.
Response: Concentrations used for Fiqures 4.7 - 4.' are averaqes
of 1992 data. The results are included in Table 4.12. EPA"
believes that presentinq the concentrations relative to MCLs is
more informative than usinq an arbitrary interval, especially
since the MCLs for the different chemicals are in some cases more
than an order" of maqnitude different. 80il data collected south
of the site does not indicate an off-site source (see response to
comment 77).
102. Figure 4.9: The point source DBCP anomaly surrounding
Monitoring Well EPAS-3 appears to be an off-site source located
near the off-s"ite liquid waste disposal pit identified in Fiqure
4 of U.S.EPA document TS-PIC-89826, September 1989, titled
"Aerial Photographic Analysis of Brown & Bryant, Inc., Arvin,
California, EPA Region 9" prepared by Environmental Monitoring
Systems Laboratory, P.O. Box 93478, Las Vegas, NV 89193-3478.
This liquid waste disposal pit is south of the Brown & Bryant
property, and it is associated with the potato shed facility due
west of the pit. DBCP was used as a pesticide on potato crops
during the 1960s. The potato shed operations included washing
the soil and its adsorbed DBCP off the potatoes and discharging
the waste water to the disposal pit.
This potential off-site source for groundwater contamination of
DBCP should be addressed in the RI/FS.
76
-------
I
i
Response: DBCP was no~ de~ec~ed in any soil samples collec~ed by
ei~her EPA or Kennedy Jenks in ~he area of ~he pi~ and po~a~o
shed, excep~ for soil samples a~ ~he wa~er ~able (see Fiqure 2.4
of the RI Report for ~he number of soil borings located in this
area). See also response ~o commen~s 77 and 101.
103. Page RI-5-13 First'Para9raph: It is not clear whether the
dinoseb modeling effort for A-zone soils included the results for
soil already removed. If so, the modeling is invalid.
Response: The modeling conducted for dinoseb was a screening
type model. It utilized conserva~ive assumptions to represent
the highest contaminated areas of the site. Concentrations
utilized for the model were biased to the highest contaminated
areas at the site, including the area ~hat was excavated.
Despite these conservative assumptions, EPA concluded from the
modeling results that addi~ional treatment of the dinoseb hot
spot was not necessary. A cap over this area was selected to
adequately address the contamination remaining after the removal;
this was the least costly alternative for soil treatment outside
of the no ac~ion alternative.
104. Page RI-6-2 Fourth Paragraph: Preliminary Remediation Goals
for residential soil are all less stringent than referenced
target concentrations.
Response: The preliminary Remediation Goals referenced here are
based on ingestion or inhalation of contamination from soil.
These goals do not take into account the potential for
contamination to leak into groundwater. Elsewhere in the RI, EPA
investigated the potential for the contaminants in soil to impact
groundwater and derived the target concentrations for that
purpose (see response to comment 98). For the risk assessment
conducted on surface soil and soil in the construction zone, the
pathway of exposure that was analyzed was incidental ingestion of
contaminated soil.
105. Table 6.1: Analytical results on the first page of the
table do not.correlate with the values used in the soil sample
calculation section on the second page of the table. The highest
value listed in Table 6.1 is 520,000 ug/kg not 5,200,000, as
noted on the second page of the table. The calculations at the
end of Table 6.1 are incorrect if the values in Table 6.1 are
recorded correctly.
The higher values in the table are listed with the data qualifier
(J), indicating approximate values.
Response: Table 6.1 includes two ~ypographical errors. The
concentrations for samples from location 10 are each missing an
additional. zero at the end and should be 5,200,000 and 3,500,000;
the correct concentrations were used elsewhere in the risk
77
.--" '~m .~. -- ---------..-.-.-- .-. .. --. .
"<-". .------...
_.. -.. "'_n.
. . '--."-'. . --.
-------
! -
, - .'-"__'n"_.,_-,,--- .---- -- ...- .._-- ".'."
. _."on_On-'.'
'-'. .,.."-.-
. .._----- .------....-- -
assessment section.
106. Table 6.5: Numbers in columns CDI (ave) are miscalculated
for the Child and Young Adult, according to the 0.1 to 5,000
mg/kg values in the range column.
Is it appropriate to use analytical numbers that have the data
qualifier (J), indicating approximate values as numbers used to
calculate heal~h risk? .
Are the high values on the table from samples collected in the
dinoseb'spill area that has already been remediated? If so, they
should not be used because they do not reflect current site
conditions and falsely over-state the risk.
Response: The CDI(ave) values in Table 6.5 are correct using the
150 average concentration. It is consistent with BPA policy to
use concentrations that are quali~ied with a "J" in risk
assessments. Quali~ied results are used because it is usually
preferable to not using any value. The highest concentration
used in the surface soil risk scenario is not ~rom the area that
has already been remediated. The highest concentration used ~or
the worker scenario was from the area cleaned up by BPA. This is
noted in the conclusions in section 6.5 of the RI. report. It
should also be noted that BPA is not proposing additional soil
cleanup in the construction zone.
107. Page RI-7-1 First Paragraph: Addressing the nature and
extent of contamination, again, removal action in the dinoseb
spill area should be detailed.
Response:
See response to comment 70 and '7.
108. Page RI-7-2 Third Paragraph: Estimating extent of chemicals
above target concentrations (i.e., 1,2-DCP over 5.5 acres at
concentrations 10 times the MCL) is meaningless. Cleanup
concentrations need to be established. The FS states that
cleanup concentrations for the A-zone groundwater between 10'and
100 times the MCL would be protective of the B-zone, based on
modeling performed. '. '.
Response: As noted in the comment, the PS study established a
cleanup standard of 10 to 100 times KCL. There~ore, it is Dot
"meaningless" to discuss the area of groundwater contaminatioD
equal to or greater than 10 times the KCL. Such a fiqure
provides a volumes estimate for cost purposes. BPA Region 9's
policy at Superfund sites is to use KCLs as a basis for cleaDup
standards in groundwater. .
109. Page RI-7-4 Third Paragraph: Data gaps include limited off-
site testing to investigate areas such as the potential off-site
source area for DBCP (Figure 4.9). .
78
-------
Response:
See response to comment 77, 101 and 102.
110. Page RI-7-6 First Paragraph:
future" is vague.
Response: Tbe potential impact from contamination in tbe B-zone
will be addressed in more detail in the second operable unit. '
No expected impact in "near
111. Page FS-1-1 First Paragraph: "Applicable or Relevant and
Appropriate Regulations" should be "Applicable or Relevant and
Appropriate Requirement.s."
Response:
The comment is correct.
112. FS-1-2 First Paragraph: The statement is misleading. A
RCRA cap is not automatically required for all waste ponds and
sumps in operation' after 1982., However, at closure, surface
impoundments in which waste residues remain must be closed as a
landfill.
Response: The comment correctly clarifies that a RCRA cap is not
required until after an impoundment is closed. Because RCRA is
applicable at Brown' Bryant, either a cap (landfill closure) or
clean closure of tbe Brown' Bryant pond and sump are required.
A clean closure would require cleanup of all the contamination.
since the contamination from the Brown' Bryant pond and sump has
spread to a considerable depth and extent, EPA does not believe
that a clean closure of tbese units is feasible; therefore, EPA
chose to use a RCRA cap.
113. FS-1-4 Third Paragraph: "Applicable or Relevant and
Appropriate Regulations (ARARs)," again, should be "Applicable or
"Relevant and ,Appropriate Requirements."
Response:
The comment is correct.
114~' FS-3-1 Third ~aragraph: "Surface soils at the sit~ are
contaminated with dinoseb at levels up to 7,400 ppm" should read
"were." As stated on page FS-1-1, "In 1991, EPA excavated and
treated the most contaminated soil containing the pesticide,
dinoseb," which would include this hot spot."
Response: The comment is correct that tbe area where the dinoseb
concentration was found at 7,400 ppm was cleaned up; however,
dinoseb concentrations at up to 5,200 ppm still remain on-site.
115. FS-3-1 Fifth Paragraph: States that "under current
conditions (no cap), the model indicated a cleanup level of 2
mg/kg would be protective" of the A-zone groundwater. This model
apparently included the "most contaminated" concentrations that
have been removed from surface soils. A separate model should be
run to reflect existing conditions. ,The model should also be run
79
----------' ----"...
"' --,-- -'--'-'-'''---- ----._-- _.--..' -.' ..'
-------
.. -. ..'. --- _.~ .... --....--- .-.-
.~. ._____0._-...
using only concentrations less than 80 mg/kg for the 0 to 7 feet
below ground surface range because several of the evaluated
remedial alternatives involve removal of contaminated surface
soils to a cleanup level of 80 mg/kg. This would model the no-
action alternative for the A-zone soils and its potential impact
on the groundwater.
Response: EPA does not intend to rerun the model for dinoseb.
Even though tbe model provided a conservative prediction of. the
potential impact of dinoseb contamination in soil to groundwater,
EPA'concluded that there is not enough of a threat to warrant
treatment of dinoseb contaminated soil in the subsurface,' .
especially considering the effect of capping the site on reducing
contaminant transport. See also the response to comment 118.
116. FS-3-10 Paragraph 3: The volume of A-zone subsoils that may
pose a threat appear overestimated. Soil samples are stated to
not exceed 2'mg/kg below 20 feet for dinoseb; however, the
estimate includes soil to a depth of 40 feet. Furthermore, the
upper 7 feet of soil is considered separately as surface soil.
Response: The volume of 48,000 cubic yards is based on the
distribution of volatile organic contaminants in the sump and
pond area and not on the dis~ribution of dinoseb contamination.
This an estimate of the volume of soil that would have been
treated by soil vapor extraction.
117. FS-4-3 Paragraph 4: For groundwater alternatives, an action
item should be instituted to investigate and destroy potential
conduits to the lower zone. In particular, EPAS-1 through EPAS-4
appear to be improperly constructed.
Response: Since well EPAS-1 serves no useful purpose, EPA agrees
it should be destroyed. As for wells EPAS-2, 3 and 4, EPA does
not believe that these wells should be destroyed at this time;
see response to comment 11t for the rationale.
118. FS-4-5 Paragraph 5: The no-action alternative may be .
acceptable for A-zone soils once proper dinoseb concentrations
are input to the MULTIMEDmodel. Also, no-action for this zone
should be considered with cleanup of surface soils to below 80
mg/kg concentration levels.
Response: The no action alternative for soil would not be
acceptable because RCRA regulations require a cap over the sump
and pond. Furthermore, EPA believes that a cap over the
remainder of the most contaminated portions of the site is an
appropriate remedy to control the spread on subsurface
contamination in the soil. Rerunning the KULTIKED model would
not change the s.elected remedy.
119. The RIfFS should present data that corrects and documents
80
-------
the shortcomings in the u.s. EPA's Brown & Bryant field program.
complete boring log descriptions for Monitoring Wells EPAS-l,
EPAS-2, EPAS-3, and EPAS-4 should be provided in the RIfFS.
There should be a discussion in the RIfFS concerning the
acceptability of using silica sand at the bottom of the
monitoring wells where the bo~ings appear to penetrate the bottom
of the A-zone aquifer. Using the limited boring information
available in the U.S. EPA RIfFS Workplan dated December 1990.
Canonie has prepared well design diagrams with elevation data and
the possible position of the AfB Aquitard (enclosed Figures 1
through 4). The potential that the EPA monitoring wells are
cross-contaminating the A-zone and B-zone aquifers should be
discussed in the RIfFS. .
Response: Two issues are presented in this comment. The use of
silica sand at the base of the wells is consistent with standard
practices. Bowever, EPA agrees that in this case, the use of
silica sand was not the best choice. In future well designs for
the site, bentonite or grout will be used, where necessary, to
fill in the space between the bottom of the borehole and the.
bottom of the well. The second issue concerns the possibility of
cross-contamination from these wells. EPA acknowledges that
there is the potential for leakage through the A-zone at these
wells and will investigate this concern further during the second
operable unit. However, based on current information, EPA
believes that this leakage is not significant enough to warrant
abandonment of wells EPAS-2, 3 and 4. . The leakage is believed to
be insignificant based on the fact that water levels within these
wells are maintained over time at levels consistent with other
site wells, and that the area over which leakage could occur is
small (especially compared to area over which the A-zone is
contaminated). However, EPA will.investigate this concern
further during the second operable unit and consider again at
that time whether to abandon these wells.
120. The discrepancy between the monitoring well elevation data
reported in the RIfFS and the RIfFS Workplan should be discussed
in the RIfFS. Groundwater flow dir.ections cannot be evaluated
until.groundwater elevation' data can be validated. .
Response:
See response to comment 94.
121. There are rusting waste drums on the site, apparently from
the U.S. EPA field investigation program. Who is responsible for
them? Do they contain hazardous waste? Are they a potential
source for surface contamination? .
Response: Drums on site are either empty or contain
derived wastes. The drums are in good condition and
leaking. EPA will dispose of this waste at the time
cleanup.
sampling
are not
of final
81
. --- '-.-------..---
'-.'-'-"- --_.
. .. -._-- " -_.._.. . ._..~- ..
-------
-_.~. .-.--.-.-..--..- .
. ....-...._~-.._.. .----- --'---
------.------------..'. ----- ------._--
.---- '_.~-'_._"
122. The u.s. EPA conducted excavation and treatment operations
within the dinoseb spill area. These operations should be fully
described in the RIfFS. Maps and cross-sections should be
provided to show the present conditions in the dinoseb spill
area. If adequate information is not availab~e, it should be
collected before the RIfFS is rewritten.
Response:
See response to comment 70.
123. The RIfFS does not contain certified analytical reports and,
therefore, it is impossible to verify the correctness ~f the .
individual tables that summarize the analytical values. As noted
in our comments, several tables contain reporting errors.
Response: All RI data has been validated. A discussion of data
quality is provided in section 2.7 of the RI report and a list of
data validation reports is provided in Appendix D, and the
reports are included in the administrative record for the RIfFS.
124. The dinoseb threat appears overstated, thus, overstating the
threat to groundwater from the A-zone soils. .
Response: EPA disagrees. The least costly remedy for addressing
the dinoseb contamination has been selected. The no action
alternative for soil is unacceptable fora variety of reasons.
In particular, it does not meet ARARs and does not address areas
of surface soil contamination onsite that exceed the health based
cleanup standard. See also response to comment 70 and 97.
125. The risk assessment appears to be based on contamination
values that have already been remediated or are within the
construction zone, which will be remediated.
. Response:
See response to comment 106.
126. The A-zone aquifer is not an appropriate operable unit, it
should be combined with the B-zone aquifer for a complete
groundwater interpretation.. . .
Response: EPA disagrees. . There is sUbstantially more
contamination in the A-zone groundwater as compared with the B-
zone groundwater. In order to treat the worse portion of the
site first and to control the current source of contamination
threatening the B-zone, EPA has proposed to address the A-zone as
a first priority. See also response to comment 67.
127. The RIfFS does not appear to evaluate all sources, including
a point source of DBCP south of the Brown & Bryant fence, former
potential contaminant sources near the Arvin water supply well,
and the point sources potentially caused by the u.S. EPA wells.
All of these contaminant sources will influence the choice and
scale of. remedy.
82
-------
Response: EPA does not aqree that these other sources will have
any affect on the remedy selected for this ROD or tbe scale of
tbe remedy. EPA ~elieves that tbere is overwbelminq evidence
tbat Brown and Bryant is tbe primary source of tbe contamination
proposed for remediation in tbis ROD. See also response to
comment 77, 101 and 102.
128. Databases used to reach a remedy need to be verified.
certified analytical results are not provided, so there is no
opportunity to assess the accuracy of the database used to reach
a remedy. Similarly, the qroundwater database is suspect ~ec~use
the U.S. EPA well logs, well details, and elevation data are not
complete nor are they consistent. It appears that nonverified
data influenced the groundwater remedy.
Response: Data used in tbe RIfFS bas qonethrouqh a thorouqh
validation and wells were resurveyed durinq tbe RI for consistent
and accurate measurements of tbe qroundwater qradient. See also
response to comment 94 and 123.
83
. '-'-".~--_._,.... ".."--.---.- . ""'--
------- |