O 1 C- o -,-.
rv
^
(Y
REVIEW OF THE REMEDIAL INVESTIGATION/FEASIBILITY STUDY

                         FOR THE

                  NYANZA CHEMICAL SITE

                 ASHLAND, MASSACHUSETTS ,
                             PREPARED BY THE

                      ENGINEERING AND  SCIENCE UNIT

                       TECHNICAL ASSISTANCE STAFF

                    OFFICE  OF THE INSPECTOR GENERAL


                             SEPTEMBER 1990

                   Report No.  E1SGG8-14-0004-0400038
  CO
  CD
                                   ENVIRONMENTAL PROTECTION AGENCY
                                   WSHINGTON, D.C. 20460

-------
                          CONTENTS
  I.   PURPOSE	    1

 II.   SCOPE AND OBJECTIVES	    1

III.   SUMMARY OF FINDINGS	    2

 IV.   SITE BACKGROUND	    3

      FIGURE 1  (Location Map)	    4

      FIGURE 2  (Site Map)	    5

  V.   CONDUCT OF THE REMEDIAL INVESTIGATION/
                              FEASIBILITY STUDY	    8

 VI.   FINDINGS	-.	   12

VII.   CONCLUSIONS	I	   22

      OIG Response to Region' s Comments.	   25

      APPENDIX	   26

-------
               Review of the Nyanza Chemical Site
          Remedial  Investigation  and Feasibility  study
I.  PURPOSE


This report has been prepared by the Engineering and Science Unit
(ESU) of the Technical Assistance staff as a part of an ongoing
Office of Inspector General (OIG) effort to evaluate the
efficiency and effectiveness of EPA's Superfund Program.
Specifically, this report was prepared in compliance with Section
lll(k) of the Superfund Amendments and Reauthorization Act of
1986  (SARA), which requires the Inspector General to examine
"...remedial investigations and feasibility studies prepared for
remedial actions."  A summary of this report will be included in
the Inspector General's annual report to Congress regarding EPA's
conduct of the Superfund program.


II.  SCOPE AND OBJECTIVES


The scope of this report is limited to Phase I of EPA's Remedial
Investigation and Feasibility Study (RI/FS) conducted at the
Nyanza Chemical Waste Dump Site in Ashland, Massachusetts.  Our
specific review objectives are as follows:

1.   Evaluate the RI/FS conducted by NUS Corporation at Nyanza
     under contract with EPA to determine if the RI/FS formed an
     adequate basis for selection of a remedial action.

2.   Evaluate the contract management performance of EPA Region 1
     to determine if the Region took all reasonable and necessary
     measures to assure that the contractor carried out the
     Nyanza RI/FS as required.

-------
III.  SUMMARY OF FINDINGS


The Nyanza Chemical Site was ranked in the top ten sites on EPA's
first Interim Priority List of 115 hazardous waste sites which
were selected for cleanup by the Superfund Program in 1981.  The
site had been occupied by a series of manufacturers of textile
dyes and intermediates from 1917 until 1978.  In January 1983,
after the completion of a Remedial Action Master Plan for the
site, EPA .issued work assignments to NUS Corporation to conduct a
Remedial Investigation and Feasibility Study (RI/FS) to determine
the most appropriate remedial action for the Nyanza site.

Our review of the conduct of the RI/FS at the Nyanza produced the
following findings:

     1.   The Remedial Investigation Failed to Locate Significant
          Areas of Contamination at the Nyanza Site.

     2.   Inappropriate Sampling Methods Led to Erroneous
          Conclusions and Additional Expenditures to Characterize
          the Site.

     3.   EPA's contractor Exceeded Its RI/FS Budget Estimate and
          More Than Doubled Its Estimated Time of Completion, Yet
          Was Unable to Produce a Report Which Would Support the
          Agency's Record of Decision (ROD).

     4.   EPA Region 1, Although Aware of Numerous Deficiencies
          on the Part of the RI/FS Contractor, Was Unable to
          Effect Significant Improvements Despite Repeated
          Notices of Unsatisfactory Contractor Performance.

During the time the Nyanza RI/FS was in progress, EPA split the
task into three phases, with the initial work focusing on
hazardous sludges and other waste materials disposed of on the
Nyanza site.  Phases II and III deal with groundwater
contamination and contaminated sediments in the nearby Sudbury
River, respectively.  These tasks were covered by separate
contracts and are not included in this review.

After NUS had terminated its work on the Phase I RI/FS, EPA
engaged Camp, Dresser and McKee  (COM) to perform additional
investigations at Nyanza "...needed to fill certain data gaps in
the Remedial Investigation/Feasibility Study done by the NUS
Corporation."  CDM's field work produced significant
modifications of the estimated quantities and locations of sludge
deposits reported in the RI/FS.  In a July 1987 summary report on
the status of the Nyanza site, EPA estimated that $210,000 of the
CDM work was to compensate for RI/FS deficiencies.

-------
IV. SITE BACKGROUND


The Nyanza Chemical Site was the location of a series of
manufacturers of textile dyes and intermediate dye chemicals from
1917 until 1978, when the Nyanza Chemical Company ceased
operations.  Large quantities of solid and liquid wastes were
disposed of on site throughout the period of operation, resulting
in extensive contamination of the soil, groundwater and surface
water on and off the site.  Principal contaminants include the
heavy metals mercury, chromium, and lead, and many chlorinated
and non-chlorinated organics including nitrobenzene, aniline,
trichloroethylene, chlorobenzene, and a variety of phenolic
compounds.

The Nyanza site is a tract of approximately 35 acres located in
the southwest of Ashland, Massachusetts, which is approximately
30 miles west of Boston.  A location map and a drawing of the
site, both of which appeared in the Phase I remedial
investigation prepared by the NUS Corporation, are included in
this report as Figures 1 and 2.  Most of the original structures
of the Nyanza Chemical Company manufacturing complex were razed
after Nyanza discontinued operations in 1978.  The remaining
Nyanza structuresfand several new buildings on the site of the
old Nyanza complex now form an industrial-commercial complex
occupied by the Nyacol Corporation and several smaller companies.
This developed area occupies approximately one-fourth of the
total site area.

In the southwest quadrant of the site Megunco Hill rises
approximately 80 feet above the elevation of the industrial park.
Nyanza Chemical Company disposed of sludge and other dye
manufacturing waste products on this hill.  An active line of the
Conrail railroad forms the northern boundary of the site.  An
abandoned trolley right of way lies on the eastern edge of the
site.  Two intermittent streams originate on the site.  Chemical
Brook rises from springs in the northwest corner of the site and
flows easterly along the railroad embankment.  Trolley Brook
originates at a spring on the eastern slope of the hill and flows
northeast through a wetland formed by the trolley right of way
embankment.  The two streams join at the northeast corner of the
site and flow through a culvert under the Conrail line,
eventually discharging to the Sudbury River north of Ashland
Center.  The Sudbury River is less than a quarter mile north of
the Nyanza site.  Prior to 1970, Nyanza Chemical Company
discharged untreated and partially treated liquid process wastes
to both Trolley Brook and Chemical Brook.  In addition to the
liquid wastes, the brooks and the Sudbury River also received
sludge and contaminated sediment which was carried off site by
runoff.  .

-------

-------
"**»
 Off
         i
         I
         o
          Q.
          <
CD
t^

i
5
IAJ
CO
                UJ
   FIGURE 2

-------
detailed description of the processes employed by Nyanza
Chemical, raw materials utilized in the manufacture of some of
the principal products, and waste products produced.  The report
included copies of detailed company production records for the
late 1960's and the quantities of raw materials purchased by
Nyanza in 1970.  DEQE also described the waste treatment
practices employed by Nyanza Chemical in the past, including
drawings showing the locations of major waste treatment
components such as sumps and lagoons, as well as points of
discharge.   The report described disposal practices for both
liquid and solid wastes, and described the general location of
major waste deposits.

The DEQE report was the first comprehensive evaluation of the
Nyanza site.  The author pointed out that mercury, while a
serious health and environmental threat, was not the only
contaminant of concern at Nyanza.  As part of their assessment
DEQE reviewed the remedial plan originally proposed by COM and
concluded that it was no longer acceptable because of the more
stringent standards which had been adopted since 1974 and because
the previous investigations had failed to take into account the
extensive use and disposal of organic chemicals at Nyanza.  The
report stated that additional investigations would be required to
characterize the site and a new remedial plan, meeting the
updated cleanup requirements, would be necessary.  A copy of the
conclusions and recommendations from the DEQE report is attached
as Appendix A.

The new owner of the site, MCL Development Corp., engaged
Connerstone, Inc. and Carr Research Laboratory to evaluate the
site contamination and prepare a cleanup plan.  Connerstone.and
Carr prepared a site Compliance Program and Schedule, which was
approved (after some modification) by DEQE in March 1981.  MCL's
consultants then conducted a field investigation which culminated
in a report, On-site Hydrogeology and Initial Pollutant Studies
on the Nyanza Chemical Waste Dump, dated August 14, 1981.  DEQE
and EPA had serious reservations about the sampling and
analytical procedures employed by Connerstone and Carr.  Also,
there was considerable controversy concerning some of the
conclusions made in the report.  By early 1982, the differences
between DEQE and the consultants had become so great that the
site investigation had virtually stopped.  In December 1982,  MCL
Development Corp. notified DEQE that it had severed its
relationship with Carr Research Laboratory.

Following the passage of the Comprehensive Environmental
Response, Compensation, and Liability Act of 1980  (Superfund),
EPA expanded its hazardous waste site activities significantly.
EPA Region 1, which had cooperated with Massachusetts DEQE on the
Nyanza site in the past, was able to devote more resources to
Nyanza once the Superfund was established.  Nyanza was reported
as an uncontrolled hazardous waste site in the Agency's November

-------
2, 1981 Hazardous Waste Report.  It was ranked in the top ten
sites on EPA's Interim Priority List of 115 hazardous waste sites
which were selected for cleanup by the Superfund program.

As a first step in the cleanup process, EPA engaged Camp, Dresser
& McKee to prepare a Remedial Action Master Plan (RAMP),  which
was completed on September 23, 1982.  The RAMP reviewed the
history of the site and summarized findings of previous
investigations which had been conducted at Nyanza.   It also
identified additional information required to characterize the
site sufficiently to serve as the basis of a selection of a
remedial action.  COM estimated that the direct costs for the
required Remedial Investigation and Feasibility Study would run
between $210,000 and $300,000 and that the tasks could be
completed within a year.


V.  CONDUCT OF THE REMEDIAL INVESTIGATION/FEASIBILITY STUDY


In January 1983 EPA issued two work assignments (one for the
Remedial Investigation and one for the Feasibility Study) to NUS
Corporation, the Agency's REM/FIT contractor for Regions 1-4.
On March 23, 1983 NUS submitted a draft work plan which indicated
that the RI/FS would be completed in 10 months at a total cost of
$480,862—$362,206 for the RI and $118,656 for the FS.  The
Agency's technical review of the work plan concluded that it did
not adequately address air quality concerns and the plan was
returned for revision.  After additional revisions, EPA
officially approved the work plan on June 30, 1983 (more than six
months after the Agency issued the work assignments)  with the
following conditions:

     Schedule:           10 months
     Level of Effort:    6050 hours
     Cost Budget:        $490,944.

EPA's REM/FIT contracts are cost-plus-award-fee contracts, in
which the fee paid to the contractor varies depending on a rating
of the contractor's performance by the Agency.  The rating is
determined by EPA's Procurement and Contracts Management Division
personnel, with input from the Project Officers who are managing
the various work assignments utilizing the contractor.  Each
Project Officer submits an Award Fee Performance Event Report
which evaluates the contractor's performance every three months.
These ratings are reviewed by the Contracting Officer and are
given consideration in determining the amount of award fee paid
to the contractor.  The first Award Fee Performance Event Report
for the Nyanza RI/FS covered performance from February 1 -. May
31, 1983.  The Region 1 Regional Project Officer (RPO) noted that
the "Contractor was very responsive and willing to make necessary
changes.  It did however take too long to receive an acceptable

                                8

-------
final product."  The RPO awarded NUS an overall rating of 3 out
of a possible 5 points for the first four-month period of
performance.

The contractor's inability to promptly negotiate subcontracts
with drillers, geophysical experts and other firms required to
initiate field work at Nyanza put the project behind schedule
almost immediately.  Contractor administrative errors also
hampered Regional approval of payment vouchers.  The RPO, in his
second report on contractor performance covering the period from
June 1-September 30, 1983, expressed concern about these problems
and observed that "These delays have led to the loss of 3 prime
months for conducting field work in New England (July, August &
Sept.) and will undoubtedly lead to further delays as adverse
weather conditions are encountered."  The RPO, dropped NUS's
overall performance rating to 2.

Although the performance ratings were high during the time the
contractor was collecting samples, the RPO's ratings again
dropped to 2 once the contractor began to submit draft products
for his review.  In his rating covering June .- September 1984,
the RPO questioned the contractor's technical competence.
Although the contractor had agreed to initiate an accelerated
project schedule early in 1984 in an attempt to make up lost
time, the project continued to miss milestones, slipping further
behind the original ten-month schedule.  Figure 3 shows the RPO's
overall ratings of NUS's performance at Nyanza throughout the
term of the work assignment.

By early 1985, more than a year and a half after approval of the
work plan, EPA Region 1 decided to split the RI/FS into two
phases.  Phase I addressed the sludge on site and the
contaminated sediments which had been carried offsite by runoff.
Phase II was to deal with groundwater contamination.  In March
1985 NUS completed draft Phase I Remedial Investigation and
Feasibility Study reports.

The Phase I Feasibility Study discussed 11 remedial action
alternatives for the Nyanza site:

1.  No action.

2.  No action with monitoring.

3.  Selective excavation, onsite disposal, capping, surface
    water and groundwater diversion.

4.  Selective excavation, onsite disposal, capping, surface water
    diversion and groundwater isolation.

5.  Selective excavation, onsite disposal, capping, surface water
    and groundwater diversion, leachate collection and

-------
    treatment.

6.  Selective excavation, onsite disposal, capping, surface water
    diversion, groundwater isolation, leachate collection and
    treatment.

7.  Selective excavation, waste fixation, onsite disposal,
    capping, surface water diversion, groundwater isolation,
    leachate collection and treatment.

8.  Total excavation, waste fixation, onsite disposal, capping,
    surface water diversion, groundwater isolation, leachate
    collection and treatment.

9.  Total excavation, backfill excavation, onsite disposal in
    secure landfill  (impermeable cap, double liner with
    underdrain, total isolation from groundwater) surface water
    diversion.

10. Total excavation, offsite disposal in permitted facility,
    backfill excavations, surface water control.

11. Total excavation, waste fixation, backfill excavation, onsite
    disposal in secure landfill, surface water control.


On April 25, 1985, NUS delivered an addendum to the draft RI/FS
reports.  The addendum added two more alternatives to the list of
possible remedial actions:

12. Total excavation, onsite disposal in secure landfill,
    surface water diversion.

13. Capping hill deposits, selective excavation and offsite
    disposal of outlying sludge deposits, backfill and restore
    areas.

After reviewing the draft RI/FS reports, EPA Region 1 determined
that the reports could not be utilized as the basis of a
selection of a remedial alternative without supplemental
information.  EPA engaged Camp, Dresser and McKee, of Boston, to
provide additional technical assistance to fill the data gaps in
the NUS draft RI/FS reports.  Although the Agency characterized
the work as "pre-design activities," much of the work performed
by COM was identical to the tasks normally associated with a
remedial investigation.

Upon completion of the additional field work by COM, the Regional
Administrator signed the Record of Decision, Remedial Alternative
Selection, on September 4, 1985.  The selected remedy addressed
the sludge deposits on the site, sediments which had been
deposited in the streambed leading to the river, and an area

                                10

-------
north of the railroad tracks which was reportedly contaminated
due to a leaking pipeline which had been used by Nyanza Chemical
Company to convey liquid wastes directly to the Sudbury River.

The selected remedy includes the following actions to contain and
isolate hazardous sludge and soils:

     -Excavation of outlying sludge deposits/sediments and
      consolidation with Hill area sludge landfill.

     -RCRA capping of the Hill area.

     -Upgradient surface water/groundwater diversion system.

     -Downgradient groundwater monitoring system.
                                11

-------
VI.  FINDINGS


1.   The Remedial Investigation Failed to Locate Significant
     Areas of Contamination at the Nyanza Site.


Abandoned Treatment Basin

In November 1982 EPA's Environmental Monitoring Systems
Laboratory (EMSL) completed an aerial photo interpretation report
on the Nyanza site.  The EMSL report includes a series of aerial
photographs of the site from 1938 to September 25, 1981.  The
early photos show a rectangular basin on the east side of one of
the major production buildings.  This basin appeared in a photo
dated April 1979, although it had been filled in when the 1981
photo was shot.  This same basin was shown in sketches included
in DEQE's 1980 report, which indicates that the basin was a
collection tank which held wastewater prior to neutralization and
subsequent sedimentation in ponds elsewhere on the site.

In spite of these two detailed sources of information indicating
the presence of an abandoned waste treatment unit, the RI makes
no mention of any sampling in the vicinity of the basin.  CDM did
collect samples in the basin area while completing its
"predesign" assignment for EPA.  These samples confirmed the
presence of organic contaminants.  Subsequently, the
Massachusetts DEQE conducted a more detailed survey.  As a result
of the DEQE investigation of the "vault" in the summer and fall
of 1986, EPA initiated an emergency removal action, which the
Agency justified on the -basis that the high concentrations of
nitrobenzene in the soil near the vault represented a hazard to
the health and safety of Nyacol employees who worked in the
former Nyanza facilities adjacent to the old treatment unit.


Buried Drums in "The Hill" Disposal Area

In early May 1989 the remedial contractor, who was excavating a
portion of the sludge disposal area on the hill in the southwest
quadrant of the site, uncovered buried drums.  This discovery
resulted in a second removal action, which ultimately located
approximately 250 buried drums.  All but 22 of the drums were
empty when uncovered, possibly having lost their contents through
leaks due to corrosion or physical damage.  Preliminary gas
chromatographic analysis of the contents detected organic
chemicals including toluene, benzene and trichlorethylene.  The
22 drums which contained material were sampled, overpacked and
staged on site.  Laboratory analyses of the samples were not
available when this report was prepared.


                                12

-------
The RI/FS contractor conducted a magnetometer survey of the
sludge disposal area on the hill in the early stages of the RI.
NUS attempted to subcontract the geophysical investigation, but,
according to their status report for August 1983,

     "The geophysics investigation was to be subcontracted, but
     the two bids that were received were not responsive to the
     needs of the investigation.  In order to remain on schedule,
     this work was then performed in-house."

The manufacturer of the magnetometer used in the survey claims
that the instrument is capable of detecting a group of 20 drums
at a depth of 50. feet.  A group of 250 drums should have been
easily detectible, since depth to bedrock (the maximum feasible
burial depth) on the hill was less than 20 feet in most
locations, according to the geologic cross sections included in
the Remedial Investigation Report.

The area in which the drums were discovered also contained other
metallic objects which had been discarded (old refrigerators,
building debris, etc.).  This may have made the presence of the
drums less obvious; however, an indication of buried metallic
objects in the area which had been identified as the principal
waste depository for Nyanza chemical should have been followed up
with a positive determination of the nature of the material.


2.   Inappropriate Sampling Methods Led to Erroneous Conclusions
     and Additional Expenditures to Characterize the Site.


The selection of sampling locations, sampling, methodology, and
the skill with which the samples are collected, preserved and
analyzed are critical factors in determining the quality and
utility of the RI/FS.  If the information obtained from the
sampling activities is not reasonably representative of the site
conditions, any decisions based on those samples reflect the
errors introduced by the sampling deficiencies.  In light of the
high cost, public health and environmental implications of
hazardous waste site remediation projects, it is essential, that
the sampling program be well designed and executed.  At Kyanza,
several significant errors in sampling strategy greatly reduced
the value of the RI/FS produced by NUS.  Three of these errors,
soil sampling techniques, monitoring well construction and test
pit locations, were important, and are discussed below:

Composited Soil Samples

Nyanza and its predecessors deposited solid waste products in the
form of sludge at many locations throughout the site.  This
sludge, which contained high concentrations of heavy metals,
resulted from the manufacturing processes at the site.  Over the

                                13

-------
years extensive grading and excavation, lagoon berm washouts,  and
sediment transport by surface runoff have further distributed
these hazardous wastes.  An earlier investigation by Connerstone
and Carr utilized approximately 80 test pits to locate sludge
deposits.  Although this work did not employ sampling and
analytical methods specified by EPA, NUS nevertheless
incorporated much of the test pit soil profile data from the
Connerstone and Carr report to estimate the quantities of sludge
on the site, since the test pit information was based primarily
on visual observations of sludge deposits.  NUS also dug 18
additional test pits, and collected soil samples "at 67 locations
in three separate areas where heavy metal sludges were suspected
of being deposited," according to the RI.

The Remedial Investigation report, at Sec. 7.2.2, further states
that "All samples were composites taken over a depth of 68 inches
using a hand auger."  The analyses of 66 of these samples are
summarized and discussed in Section 7 of the RI report.  The
utility of soil samples composited over a depth of 68 inches
where the depth and thickness of the sludge deposits varies
greatly from sample to sample is highly questionable.  Although
useful for detecting the presence of contaminants during the
preliminary stages of an investigation, such sampling techniques
give no indication of the amount of contaminated material which
must be removed or of the concentration of contaminants within
the sludge, since there is no way to tell how! much sludge or
contaminated soil is in a given sample or how deep it extends.
Consequently, the tabulation and discussion of the soil sample
analyses have little meaning.

The lack of definitive information about the sludge deposits at
Nyanza necessitated extensive additional sampling after
completion of the Phase I RI/FS to accurately characterize the
sludge deposits at the site.  This extra work is discussed in
greater detail elsewhere in this report.


Monitoring Well Construction

The Nyanza site is located in a glaciated region, with complex
geological features.  Consequently, groundwater movement is
difficult to predict.  In addition, the groundwater pollutants of
concern at Nyanza include a wide range of inorganic and organic
chemicals.  The organic contaminants include both light and dense
non-aqueous phase liquids.  Light non-aqueous phase liquids (L-
NAPLs) are less dense than water and thus tend to float on top of
the groundwater.  Dense non-aqueous phase liquids (D-NAPLs) are
heavier than water and sink through the groundwater and move
along any impervious layer or the bedrock underlying the aquifer.
Both D-NAPLs and L-NAPLs are only slightly soluble in water.  D-
NAPLs, once released into an aquifer, are generally difficult to


                                14

-------
trace and recover, often moving in quite different directions
from the general groundwater flow.

The DEQE preliminary site evaluation of 1980 documented the use
of large quantities of nitrobenzene, a D-NAPL, at Nyanza.  Where
non-aqueous phase liquids are known or suspected to be present in
the groundwater under or near a hazardous waste site, monitoring
wells should be constructed in a manner to maximize the
probability of detecting those contaminants.

In an unconfined aquifer, such as the one at Nyanza, monitoring
wells are generally constructed with relatively short screen
lengths at the top of the aquifer, at the bottom, and—depending
on the thickness of the water bearing formation(s)—at one or
more intermediate depths.  The wells are constructed so that
samples can be withdrawn from any of the various levels
independently, without mixing with groundwater from other levels.
This can be accomplished either by constructing a number of
separate wells of different depths in a cluster or by drilling a
single well but isolating the different screens through special
construction methods.  The upper screen captures any L-NAPLs
floating on top of the groundwater and the lower screen is more
likely to collect D-^NAPLs moving at the bottom of the aquifer.
Six of the monitoring wells at Nyanza were constructed with at
least two screen intervals.

Monitoring well 8A (MW-8A), according to Table 6-1 in the RI
report, was constructed with a single 17-foot screen (although
the boring log in the RI Appendix B indicates that a 30-foot
screen was installed, covering almost the full depth of the
well).  The use of a single long screen is remarkable,  because,
on page 6-6, the RI report explains that MW-8A was the only well
which was constructed with a stainless steel casing and screen,
specifically because of the suspected presence of nitrobenzene, a
D-NAPL which attacks PVC.  All of the other wells installed by
NUS were of PVC construction.  Since nitrobenzene is much denser
than water and only slightly soluble in water, it would most
likely be found in the fractured bedrock underlying the
unconsolidated glacial till and sediments.  Yet MW-8A was not
constructed with the capability of sampling groundwater from
discrete levels of the aquifer.

In July 1987, more than two years after NUS completed the Phase I
RI/FS, EPA's Emergency Response Team (ERT) from Edison, NJ,
constructed a well immediately adjacent to monitoring well 8A.
The new well was equipped with a short screen length at bedrock
level.  Samples collected from the new well contained more than
54 mg/1 nitrobenzene, indicating gross contamination.  Samples
which ERT collected from MW-8A, however, contained less than 0.4
mg/1 nitrobenzene.  Had this information been available while the
RI/FS was in progress, it is possible that NUS would have
discovered the area of highly contaminated soil which prompted

                                15

-------
EPA to initiate a removal action in the immediate vicinity of
MW-8A in 1987.

The cover memo transmitting the groundwater report prepared by
the Environmental Response Branch to the On Scene Coordinator of
the 1987 removal action included the following statement:

     "The poor construction of M-8A (installation of a 15'
     screen) prevented the EPA from understanding the character
     of the contaminant plume at that point until ERT-2 was
     completed."

Groundwater Sampling

The first two rounds of groundwater samples were collected from
the monitoring wells in November 1983 and March 1984.  NUS field
staff did not filter suspended material from these samples prior
to field-fixing and sending them to the contract laboratory for
analysis.  The acid used to fix and prepare water samples for
metals analysis dissolved the metals naturally present in the
suspended solids in the samples.  Consequently, the
concentrations of heavy metals reported by the lab included the
metals associated with the particulate matter as well as those in
solution in the groundwater.                  .-

In subsurface conditions such as those found at the Nyanza site,
the particulate matter does not move with the:groundwater.  The
metals present in the particles would have little, if any, impact
on groundwater quality.  The technical experts who reviewed the
laboratory data assumed that the samples collected for metals
analysis had been filtered when collected.  Consequently, the
failure to filter the samples led them to believe that metal
concentrations in the groundwater were much higher than they
actually were.

This misunderstanding caused considerable confusion until EPA
learned that the samples had not been filtered.  NUS collected
filtered samples in December 1984, but results from this third
round were not available in time to be included in the March 1985
draft RI report.  On pages 6-16 and 6-21, in the Geology and
Hydrogeology section of the draft report, NUS refers to tables
showing the average total concentration of four metals (arsenic,
chromium, mercury, and lead).  In the discussion of the results,
the report states as follows:

     "Since the groundwater samples were unfiltered it is
     impossible to determine if the metal concentrations reported
     were in well sediments or dissolved in the groundwater, but
     it appears that the groundwater in the unconsolidated
     deposits is more contaminated than the bedrock.
                                16

-------
     "An additional round of samples has been collected to
     determine only the dissolved metals in the groundwater.
     This sampling will address the potential for metal migration
     in the groundwater at the site."

Consequently, the discussion of heavy metal contamination in the
RI is seriously flawed.  Furthermore, the costs associated with
collecting and analyzing the first two rounds of metals samples
produced information of questionable value.

Sludae Volume Estimates

NUS dug 18 test pits at Nyanza.  Most of these were located on
the hill in the western end of the site, which was the primary
disposal area for sludge from the dye manufacturing processes.  A
few test pits were located on the northwest slope of the hill and
in the wetland area in the extreme northwest corner of the site.
The locations of 16 of these sites are shown on Drawing No.
0714-06-03, which accompanies the RI report.  All of the NUS test
pits shown on the drawing are in Area I—Megunco Hill—one of the
eight areas identified as sludge and fill deposits.  NUS's
estimates of the extent and volume of the remaining deposits are
shown on this same drawing.  These volume estimates were
apparently based on the test pit logs from the Connerstone and
Carr investigation of 1981.  As mentioned earlier in this report,
EPA and the DEQE were critical of the methods employed by
Connerstone and Carr, yet NUS apparently did nothing to confirm
the data from their report which they utilized to estimate sludge
volumes.

After NUS had terminated its work on the Phase I RI/FS, EPA
engaged CDM "to perform -pre-design work needed to fill certain
data gaps in the Remedial Investigation/Feasibility Study done by
the NUS Corporation" (as explained in a letter of June 7, 1985
from Regional Administrator Michael Deland to U.S. Rep. Lucile R.
Hicks, U.S. Rep. David P. Magnani, and U.S. Sen. Edward M.
Kennedy).

CDM and its subcontractors conducted extensive field work at
Nyanza from June 3 through June 27, 1985.  Field work included
the following activities:

   - surficial soil sampling
   - water sampling
   - test pit excavation
   - soil borings and standard penetration tests
   - boundary and topographic surveying
   - analytical laboratory services.

CDM's field work produced significant modifications of the
estimated quantities and locations of sludge deposits reported in
the RI/FS.  In some cases, where Drawing No. 0714-06-03 in the

                                17

-------
RI/FS showed sludge deposits, CDM's investigation found nothing.
Other areas which were not identified in the RI/FS as
contaminated were found to contain hazardous sludge.  Because
approximately three quarters of the total estimated volume of the
sludge was located in the hill area, the errors in the RI/FS did
not result in a great change in the total amount of material to
be excavated to implement the recommended alternative.  However,
the differences between the NUS and the COM reported locations of
the contaminated material were considerable.  Although CDM's
final estimated total volume of 105,610 cubic yards (cu. yd.) was
very close to that of NUS (102,300 cu.yd.), the sum of all the
discrepancies between the two estimates totaled approximately
45,000 cubic yards, 44% of the original NUS estimate (see Table
1).

The COM followup study found that approximately 21,000 of the
102,000 cubic yards of material which NUS determined to be
contaminated did not have to be excavated.  Furthermore, COM
determined that more than 24,300 cubic yards of material which
the RI/FS indicated could be left undisturbed was in fact
contaminated and had to be excavated to reduce the environmental
hazard to an acceptable level.


3.   EPA'a contractor Exceeded Its RI/FS Budget Estimate and More
     Than Doubled Its Estimated Time of Completion, Yet Was
     unable to Produce a Report Which Would support the Agency's
     Record of Decision.                     ;

NUS submitted a preliminary RI/FS work plan to EPA Region 1 on
March 25, 1983.  The total estimated cost of the effort was
$480,862.  EPA had authorized $500,000 for the RI/FS.  The work
plan included a schedule indicating completion of both the RI and
FS reports within ten months of the start of work.

EPA eventually paid NUS more than $612,000 for its efforts.  NUS
worked on the project for 23 months, from June 1, 1983 to May 1,
1985, producing reports which EPA was compelled to supplement
with extensive additional investigations by COM in order to
provide sufficient support for the selected remedial action.  In
a July 1987 summary report on the status of the Nyanza site, EPA
estimated that approximately $210,000 of the COM "pre-design"
effort was to compensate for RI/FS deficiencies.

Furthermore, the final ROD addressed only the!on-site sources of
contamination  (sludge deposits), leaving the groundwater
contamination to a Phase II RI/FS which was initiated at a later
date.  Neither the RAMP nor the RI/FS Work Plan anticipated
dividing the project into multiple phases.  This decision was not
made until early 1985.  None of the original schedules or cost
estimates prepared prior to initiation of the RI/FS at Nyanza


                                18

-------
               Table 1.
         NYANZA CHEMICAL SITE
Comparison of EPA Contractor Estimates
Of Volumes of Material to be Excavated
Area Description
Hill Area + Wet-
land in NW Corner
Area North of
Nyanza Office
Parking Lot
Strip Along
Northern Boundary
Of Site
Wetland @ Conflu-
ence of Trolley &
Chemical Brooks
Wetland @ Head of
Trolley Brook
Filled Lagoon
Behind Derby Chem.
Trolley Brook
Sediments (North
of Conrail Tracks)
Areas Not Listed
As Contaminated
In NUS Rl Report
TOTALS
m
Name
I
II
III
IV
V
VII
VIII
N/A
IS CDM
Est.Vol.
Cu . Yd .
75,000
1,600
12,000
7,500
4,900
500
800
0
102,300
Name
B&C
N/A
E
F
6
I
J
D,L,
M,N
Est.Vol.
Cu.Yd.
92,400
; o
1
110
, 0
6,480
1,640
1,120
3,860
105,610
Differ
ence
Cu.Yd.
17,400
1,600
11,890
7,500
1,580
1,140
320
3,860
45,290
%
Diff.
+23%
-100%
-99%
-100%
+32%
+228%
+40%
+100%
                  19

-------
 indicated that groundwater contamination would be addressed in a
second phase.  On the contrary, NUS's work plan specifically
included a groundwater investigation as one of the tasks to be
undertaken as part of the RI/FS.

At the time NUS prepared the RI/FS work plan,, there was a
considerable body of available information on the Nyanza site.
Studies funded by EPA, the DEQE and three of the site owners had
been completed and had been reviewed by both agencies.  Nyanza
was recognized as a complex site, with multiple deposits of
hazardous solid wastes and documented inorganic and organic
groundwater contamination which had already migrated off site.
The NUS field investigation did not reveal any findings which
differed markedly from information previously reported.  Yet the
RI/FS exceeded its budget and schedule by wide margins, and the
objectives of the effort were not achieved, even after
supplemental work by an additional EPA contractor.


4.   EPA Region 1, Although Aware of Numerous Deficiencies on the
     Part of the RI/FS contractor, was Unable to Effect
     Significant Improvements Despite Repeated Notices of
     Unsatisfactory Contractor Performance.


EPA's Regional Project Manager (RPM) for the Nyanza RI/FS was
responsible for managing the contractor's efforts.  The RPM
typically monitors contractor activities through regular progress
reports, review of work products, meetings and site visits, when
appropriate.  In addition to managing the contractor, the RPM is
the primary responsible party for overall conduct of the RI/FS,
including liaison with State and local government, public
information, and coordination with other EPA personnel who are
involved in the project (e.g., specialists required to review
draft contractor reports).  The RPM originally assigned to Nyanza
was also responsible for three other RI/FS efforts.  This was a
typical work load for RPMs in Region 1 at that time.

The administrative demands of project management leave little
time for careful RPM review of the technical aspects of an RI/FS.
Consequently, at Nyanza, as with most projects, NUS had
relatively little technical input from EPA prior to undertaking
field work.  Most of EPA's technical expertise was devoted to the
review of draft reports prior to their release by the Agency.

EPA's contracts with the REM contractors are cost-plus-award-fee
contracts, in which a portion of the contractor's fee is
influenced by RPM performance ratings.  This mechanism was
introduced as an incentive to improve contractor performance,
since higher performance ratings will produce larger fees.
                                20

-------
As discussed in Section IV of this report, the contractor rating
forms periodically submitted by the RPM indicate that
deficiencies in NUS's performance surfaced early in the course of
the RI/FS.  By the second four-month period, the overall score
had dropped to 2, and the RPM had assigned the lowest possible
score, 1, in two of the categories (schedule and cost control,
and resource utilization) which make up the overall score.  By
the last rating period, February 1, 1985 to May 31,1985, the
contractor's rating had dropped to 1 in all categories and
overall.  In the Work Assignment Completion Report, which is a
comprehensive rating of the contractor's performance for the work
assignment, the RPM recommended that NUS be denied any of the
available award fee, characterizing the contractor's performance
as "abysmal".

These increasingly low ratings did not produce the desired effect
on the contractor's performance, and EPA had no other feasible
mechanism to remedy the situation.  Consequently, NUS continued
to work until the REM I contract expired, delivering RI and FS
reports which were not sufficiently rigorous to support the
selection of a remedial alternative.  EPA Region 1 then engaged
COM to fill the remaining "data gaps" in the RI/FS reports.
Although EPA characterized the COM assignment as "pre-design",
much of the work was to correct deficiencies in the RI/FS
performed by NUS.
                                21

-------
VIZ.  CONCLUSIONS
1.   Did tba RI/FS Conducted by NOS corporation Provide an
     Adequate Basis for Selection of a Remedial Alternative at
     Nyanza?


The Nyanza Site, like most hazardous waste sites, presented a
complex puzzle to EPA and its RI/FS contractor, NUS Corporation.
A series of chemical manufacturers had disposed of a wide variety
of organic and inorganic wastes at numerous undocumented
locations throughout the manufacturing complex for approximately
60 years.  Because Nyanza was one of the earlier Superfund sites,
EPA had not developed much of the guidance now in place to assist
Agency remedial project managers.

Nevertheless, the deficiencies manifested in the Nyanza RI/FS are
remarkable even in a program which has been documented as having
severe problems.  The contractor exceeded the original budget by
25% and required 23 months to complete an assignment which it had
committed to finish in 10 months.  More importantly, the final
product of this effort failed to fulfill its fundamental
purpose—to characterize the nature and extent of risks at the
Nyanza site and to provide a rational basis for the selection of
a remedial option.

The residents of Ashland had formed an active citizens group
which was closely monitoring the progress of the Nyanza cleanup.
The significant discrepancies between NUS's assessment of sludge
deposits and CDM's findings were a major blow to the credibility
of the RI/FS and the Agency.  The two removal actions undertaken
at Nyanza in response to hazards undetected during the Remedial
Investigation further diminished the community's confidence in
EPA's cleanup effort at Nyanza.  EPA's expenditure of more than
$200,000 to obtain additional data to support the Record of
Decision is conclusive evidence that the Nyanza Remedial
Investigation/Feasibility Study was deficient.
                                22

-------
 2.    Did EPA Region 1  Take  All  Reasonable and Necessary Measures
      to  Assure  that the Contractor  Carried Out  the Nyanza RI/FS
      as  Required?

 Agency personnel responsible  for  oversight of the contractor's
 performance,  although  acutely aware of deficiencies  early in the
 process,  were unable to rectify the situation using  the remedies
 available to them  through the contract management process.  NUS
 was not  sufficiently concerned  about the problems at Nyanza to
 take  decisive corrective action until April  1985, when the firm
 replaced its Project Officer—after the draft RI and FS reports
 had been completed and far  too  late to have  a major  beneficial
 impact on the project.   EPA Region  1 engaged a  second contractor
 to correct the  deficiencies in  the  RI/FS.  Although  the second
 contractor completed the additional investigative work prior to
 the issuance of a  Record of Decision, the Region characterized
 the contract as a  "pre-design".contract.  According  to Region 1
 personnel, at least $200,000  of the cost incurred under the
 supplemental contract  was clearly for work which should have been
 completed as part  of the RI/FS.   Much of this additional work
 duplicated work accomplished  by the original RI/FS contractor.

 The superfund program  has been  structured to utilize contractors
 to perform Remedial Investigations  and Feasibility Studies under
 the direction of EPA program  officials.  Because of  the obvious
 uncertainties associated with most  hazardous waste sites, the
 Agency employs  cost-plus-award-fee  type contracts for RI/FS
 assignments.

^Although this contracting method  provides the necessary
 flexibility  to  allow the contractor to react to changing demands
 of the project, it also relieves  the contractor from strict
 accountability  for the quality  and  cost of the  final product.
 Whether  or not  EPA is  satisfied with the contractor's
 performance,  a  cost-plus-award-fee  contract  assures  the
 contractor that all costs,  including overhead,  will  be
 reimbursed.

 The award fee is intended to  provide an incentive for superior
 contractor performance. At Nyanza, however, where the contractor
 was apparently  unable  to provide  a  satisfactory level of
 performance  in  spite of the incentive, the award fee was
 ineffective.  This problem  was  aggravated by the lack of
 sufficient numbers of  experienced,  skilled technical personnel at
 EPA Region 1.  When they are  responsible for several projects as
 large and complex  as a typical  RI/FS, Remedial  Project Managers
 cannot effectively keep abreast of  the rapidly  changing
 developments which are commonplace  at remedial  sites.  This
 situation tends to place RPMs in  a  reactive  mode, unable to
 detect and address developing issues before  they become major
 problems.


                                23

-------
EPA Region 1 was unable to obtain acceptable performance from the
Nyanza RI/FS contractor, although EPA staff appear to have taken
all reasonable measures within their control toward that end.
The contractor's failure to take effective corrective action
after receiving repeated poor award fee performance ratings from
the Regional Project Officer is strong evidence that the remedies
available to the Agency under the REM/FIT contract were
insufficient to address the poor performance encountered at
Nyanza.
                                24

-------
                PIG Response to


The DIG provided Region 1 with a draft version of this report for
review and comment in August 1989.  The Acting Regional
Administrator responded with the Region's comments in a letter
dated September, 1989 (See Appendix).  In general, the Region
agreed with the substance and thrust of the report.  The Region
also noted that it anticipates that the Agency's recently
established Alternative Remedial Contracting strategy (ARCS) will
give the regions somewhat greater control over the assignment of
contractors to remedial projects.  However, the ARCS program  ,
employed a contractor selection process which 'prevented the
consideration of prior experience in ranking the competitors.

Consequently, NUS has been selected to perform the RI/FS for the
Sudbury River sediment contamination attributed to Nyanza
Chemical discharges.  Thus the Region is faced with the prospect
of conducting another RI/FS in Ashland under the scrutiny of an
active, well-organized group of Ashland citizens, many of whom
have little confidence in the Environmental Protection Agency or
its contractor, based on previous experience.

Furthermore, the ARCS contracts employ the same cost-plus-award-
fee payment system which the Region unsuccessfully attempted to
utilize to improve contractor performance in the original Nyanza
RI/FS.

Five of the Region's six specific comments were relatively minor
corrections or expansions of statements in the draft report.
These have been incorporated in the final version of this report.
The sixth comment disputes the report's critical discussion of
the groundwater investigation—specifically the assertion that
monitoring well MW-8A should have been constructed to monitor
groundwater at different elevations.  The Region noted that some
of the terminology in the discussion of the monitoring wells
(e.g. D-NAPL and L-NAPL) was unknown at the time the RI/FS was
performed.  Although it is true that the terminology is recent,
the concept of sampling aquifers at various levels is not.  In
fact, six of the groundwater monitoring locations which NUS
utilized at Nyanza were capable of sampling groundwater at two or
more depths.  Furthermore, as the final version of this report
indicates, OSWER's Environmental Response Branch was also
critical of the single long screen installed in MW-8A.
Consequently the OIG's position on the groundwater investigation
remains unchanged.
                                25

-------
APPENDIX
   26

-------
**:>„
*-.         UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

 8                      .     flEGION I
 f
*/          J.F. KENNEDY FEDERAL BUILDING, BOSTON, MASSACHUSETTS 02203-2211
   September 18, 1989
                 /

   Stephen E. Burbank
   Acting Divisional Inspector General
   Eastern Audit Division
   J.F.K. Federal Building
   Room 1911
   Boston, MA  02203-2211

   Re:  Nyanza Chemicals RI/FS

   Dear Mr. Burbank:

   Enclosed  please  find  comments  prepared  by the  Region's Waste
   Management Division on the draft report prepared by your Engineer-
   ing and Science Unit on the RIFS  for the Nyanza site  conducted by
   HUS and of EPA's oversight of NUS  on that project.  The Region  gen-
   erally agrees with the substance and thrust of the report although
   we question the value of  its inclusion in  the,1989 Report to  Con-
   gress since the work  in question was completed  in  1985.   I would
   also  like to provide  you with  the Region's  perspective on  the
   transition  from  the REM  contracts  to  the  newly established  ARCS
   program.

   In interviewing  prospective A/E  firms for the  newly  established
   Alternative Remedial Contracting Strategy (ARCS), the Agency selec-
   tion panel members were.specifically directed by headquarters not
   to use prior  experience with an A/E firm (under REM for instance)
   as a criterion for evaluation.   Consequently, NUS has been awarded
   one of the seven ARCS contracts  for Region I and,  ironically,  has
   been  tasked to perform the  RIFS for  the  third operable  unit at
   Nyanza (the Sudbury River sediments).  I'm  sure you can appreciate
   the considerable  discomfort experienced by the RPM in  trying to
   explain this  development  to the community.

   Notwithstanding the aforementioned, Region  I believes the ARCS  pro-
   gram, although utilizing the same cost  plus  award fee system as the
   REM program,  will  provide the  Agency will  greater leverage to
   secure better contractor  performance for two reasons.   First,  the
   ARCS  contractors  will  be  guaranteed only a  minimum floor of  work
   under the contract; additional work will be assigned to those  con-
   tractors whose past performance merits  it.   Secondly, the ARCS  con-
   tracts  run for 10  years.  This  will  allow the Region,  in  many
   cases,  to retain  one contractor throughout  the RIFS,  Remedial
   Design, and Remedial Action phases  of  a project and should result
   in greater contractor accountability.   By contrast,  the 2-4 yr. REM
   contracts resulted in continual contractor turnover from one phase
   of a project  to the next.
                                27

-------
We appreciate the opportunity to comment on the draft report.  If
you would  like to  discuss any of  our comments,  please contact
Richard Cavagnero in the Waste  Management  Division at FTS:   833-
1730.
Sincerely,
Paul G. Kebugh
Acting Regional Administrator

cc:  Merrill S. Hohman
                             28

-------
r
                                       Attachment:

           p. 2/3    According to State records, Nyanza never declared bank-
                     ruptcy.  The Region is currently  investigating the issue
                     of corporate succession for possible cost re-covery.

           p. 8      It was during CDM's "predesign" sampling activities that
                     the  vault  was determined  to contain organics  and not
                     metal sludges.   COM conveyed this information to EPA's
                     RPM, who initiated discussions with DEQE about the pos-
                     sibility of DEQE performing a separate study on the vault
                     area.  EPA was  out of funds at  the time due to pending
                     CERCLA funding  authority  termination which lasted over
                     a year due to lack of reauthorization. These discussions
                     concluded with the signing of an  "Advance Match Coopera-
                     tive Agreement"  by EPA and DEQE, under which the State
                     would fund the  vault investi- gation and later receive
                     credit towards its ultimate 10%  cost share.

           p. 9      Without defending  the work of NUS,  it  should  be noted
                     that the area where the drums were unearthed during re-
                     mediation was  a debris dump  for trash  including metal
                     objects (empty barrels, refrigerators, remnants of Nyanza
                     buildings).  The presence  of so many metal objects un-
                     doubtedly hindered the magnetometer survey.

           p. 10     EPA's concern  with the Connerstone and Carr  work was
                     primarily  with  the  borings,   monitoring  wells,  and
                     sampling techniques rather than the test pits, which were
                     excavated by conventional means.  NUS1 greatest failure
                     was  in not performing confirmatory sampling to confirm
                     or deny the earlier work.

                     The  use  of  composited  soil samples,  while  clearly
                     problematic  in  accurately  defining sludge  depths  and
                     volume, was in part necessitated by the realities of the
                     CLP program which, at this period of time, was seriously
                     lacking in capacity and ability to provide timely turn
                     around.

           p. 11     The  discussion  of L-NAPLs  and  D-NAPLs constitutes  a
                     revisionist perspective.  Dr. John Cherry, possibly the
                     leading  researcher  into  the  behavior  of  D-NAPLs  in
                     groundwater systems, had not yet even coined the term D-
                     NAPL during the time period when the RIFS was performed.
                     The  report  author's  conclusions regarding  the  proper
                     placement of  well screens also  neglects the fact that
                     this issue was the subject of considerable debate at the
                     time the  RIFs was performed.    Indeed,  certain states
                     actually required that the entire saturated  thickness be
                     screened as  a precaution against  the possibility that
                     discretely located screens might miss a permeable lens
                     conducive to contaminant transport.
                                             29

-------
The critique  oversimplifies the problem  of accurately
defining groundwater  contaminant transport  in complex
hydrogeologic  settings  like New  England and  fails to
acknowledge that  the  body of knowledge  in  this field,
while evolving rapidly,  is still  in its infancy and that
the number  of real experts  in  the field is also very
limited.
                   30

-------