2015-12-18 SELC Comments on Duke Corrective Action Plan for Buck

SOUTHERN ENVIRONMENTAL LAW CENTER
Telephone [phone] WEST ROSEMARY STREET, SUITE 220 Facsimile 919-929-9421
CHAPEL HILL, NC 27516-2356
December 30, 2015
Via Electronic and U.S. Mail
Donald R. van der Vaart, Secretary
North Carolina Department of Environmental Quality
1601 Mail Service Center
Raleigh, North Carolina 27699-1601
donald.vandervaartkncdenr. gov
Re: First Set of Comments on Errors in Duke Energy's Corrective Action Plan for the
Buck Coal Ash Ponds
Dear Mr. van der Vaart:
The Department of Environmental Quality ("DEQ") has stated that it values public
comment on the risk prioritization process for Duke Energy's ("Duke") coal ash basins. See
DEQ v. Duke Energy, 13 -CVS -14661, 13 -CVS -11032, Hearing Tr. at 90:14-16 (N.C. Sup. Ct.
Sept. 14, 2015) (Counsel for DEQ stated that "the public participation process is important, is
essential, is crucial"). On behalf of the Yadkin Riverkeeper and Waterkeeper Alliance
(collectively the "Riverkeepers"), the Southern Environmental Law Center provides these
comments on serious errors in Duke's Corrective Action Plan ("CAP") for the Buck Steam
Station ("Buck"). In particular:
1) Duke failed to fix errors that have been known for some time:
a. Duke has not placed any monitoring wells near the residential wells around
Buck;
b. Duke, once again, has made no attempt to evaluate the bedrock aquifer system
beyond its property boundary;
c. Duke still has not considered the effects of residential well pumping on
groundwater flow;
d. None of Duke's monitoring wells draw water from a similar depth as the
residential wells;
e. Duke still has not evaluated the water levels in residential wells;
f. Duke still has not evaluated the contaminants detected in residential wells;
g. Duke has conducted only a single round of sampling of its new monitoring
wells.
2) Duke's CAP is filled with internal contradictions:
a. Duke's entire CAP hinges on so-called "background" wells that Duke itself
admits may not represent background conditions;

b. Duke identified a background level of hexavalent chromium that cannot be
reconciled with the background level of total chromium identified;
c. Duke assumes that the ash basins are not sitting in the groundwater under a
cap -in-place scenario, yet all available evidence shows that the ash basins
would in fact extend down into the groundwater under this scenario.
3) Even though Duke's analysis obscures the full extent of contamination at Buck, it still
shows that the Buck coal ash lagoons pose a high risk:
a. Even accepting Duke's purported "background" levels of contamination,
many residential wells exceed these levels;
b. Duke's own analysis concludes that coal ash contaminants are flowing into
the Yadkin River;
c. Even accepting Duke's closure scenario analysis, contamination decreases
when the coal ash is removed, showing that the source of these contaminants
is the coal ash at Buck.
DEQ has further stated that high uncertainty will lead to a high risk designation. Duke
has had a year and a half since the passage of the Coal Ash Management Act to obtain the data
necessary to characterize the full extent of contamination at Buck; nearly three years since it
began its first attempt, which it later abandoned, to perform a groundwater assessment; and
nearly ten years since it first began monitoring groundwater at Buck. Yet Duke still has not
produced key information regarding contamination at Buck. Duke should not be permitted to
use its failure to supply potentially incriminating data to obtain a low risk designation for Buck.
Instead, DEQ should designate Buck as a high risk site, as DEQ regional staff did.
This is particularly true because even the partial information that Duke has supplied
shows contamination of the groundwater at Buck, contamination of the Yadkin River, and
contamination above background levels at dozens of residential wells.
The Riverkeepers note that there may be many other errors in addition to these glaring
and significant omissions in the Buck CAP. In addition, similar problems may pervade Duke's
CAPS for its other coal ash facilities. Although the CAP is available on the DEQ website,'
numerous CAP figures are unreadable, distorted, or lacking most or all of the information that
they are intended to present. These same problems pervaded Duke's Comprehensive Site
Assessments posted on DEQ's website, yet in the over four months that have elapsed since
conservation groups first alerted DEQ to this problem, DEQ has not taken any steps to resolve
the issues or make full and complete copies of the documents available to the public. As a result
of this lack of transparency, which has now been repeated in Duke's CAPS, the public has not
been able to fully evaluate the adequacy of Duke's CAPS. We expect to identify additional areas
of concern upon further evaluation of and access to complete and accurate figures and data.
I. Duke Failed to Fix Known Data Gaps.
A. Duke self-servingly placed bedrock monitoring wells away from the
residents.
1 http://edocs.deq.nc.gov/WaterResourcesBrowse.aspx?startid=221202&dbid=0&cr=1
N

As Riverkeepers have pointed out already, the most obvious and egregious deficiency in
Duke's bedrock aquifer assessment at Buck is its failure to assess any bedrock areas to the south,
southeast, and southwest of its coal ash basins—the very areas where the majority of the
residential wells are located. See Buck Comprehensive Site Assessment ("CSA") Figure 6-7;
Buck CAP Figure 3-5; Figure 1 below. Duke failed to correct this major problem in its CAP,
despite the fact that it had many months to do so between submission of its CSA and CAP.
Figure 1: Excerpt of Buck CSA Figure 6-7, Modified to Depict Nearby Drinking Water
Wells; Highlighting Shows Bedrock Areas Ignored By Duke
i .ar :i:w4s�.T'a. fart"
C Drinking Water
LEGEND
-� APPRDM9AATE GROLRJOriRTER ROW PIRECTIGN
ASN BASIN ASSESSMENT SRODNWATER
MGNIYGRIPR31TE4.L
-- GRDDNDWATER CDNTOUR LINE
- � � DUKE ENERGY YROP£RTY BG18f81ARY
ASR BASIN WASTE BOIMIDARY
-ASW BASIN Ca1PLWNCE BOUNDARY
ASN BASIN OANPLIANOE BOUNDARY COWOIDENT
- - - W RH DUKE ENERGY PRDPERW BOUNDARY
SnR u
NR-NOREADMG
The sole bedrock monitoring well to the south of the facility, BG -IBR, was reported as
`dry', and thus it can provide no information about contaminant concentrations, hydraulic head,
or local groundwater flow. Duke's own CSA recognizes that "[a] replacement bedrock
background well in this location may be warranted if BG -1 BR is not a viable well." CSA at ES -
14. Yet Duke reports no effort to date to replace this well, and instead states that it intends to
wait until 2016—after DEQ's deadline to make initial risk prioritization decisions, many months
after Duke itself recognized the need for additional monitoring wells, and well over a year after
Duke began its site characterization process—before Duke even begins installation of wells in
response to the data gaps identified. CAP at 24.
3

By attempting to characterize the bedrock aquifer for only half of the Buck site—the
portion furthest from the bedrock residential wells—Duke has omitted information that is critical
to determining the full extent of the groundwater contamination and risk at Buck.
B. Duke, once again, has made no attempt to evaluate the bedrock aquifer
system beyond its property boundary.
In addition to not characterizing the bedrock aquifer in the southern parts of the Buck
site, Duke has also refused to install any monitoring wells offsite. Even DEQ has instructed
Duke to install offsite monitoring wells: "It is expected that Duke conduct assessment work,
including borings, wells, and surface water/bed sediment sampling, offsite (outside of property
boundaries) as needed for adherence to 15A NCAC 02L.0106 (g)(4) and (or) for the evaluation
of background conditions. The plans as submitted do not adequately account for offsite
assessment work or provide a justification that offsite work is not needed." DEQ Response to
Duke's Initial Groundwater Assessment Workplan for Buck (Nov. 4, 2014). Yet Duke has
resisted installing offsite monitoring well at every phase of the analysis, and most recently has
failed to include offsite wells in its CAP.
C. Duke still has not considered the effects of pumping.
Although pumping of residential wells can change the flow of groundwater dramatically,
Duke's CAP again avoids analyzing the effects of pumping of the numerous residential wells
near Buck. Instead, the CAP inexplicably states that "it will not be possible to assess the affect
[sic] of pumping wells[.]" CAP at 68. It is well understood that pumping water out of
residential wells can create a cone of depression that can reverse groundwater flow, pulling in
groundwater from areas that might otherwise be down -gradient of residential wells. Duke's
failure to analyze this phenomenon undermines its self-serving groundwater flow conclusions.
D. None of Duke's bedrock monitoring wells draw water from the same level as
residential drinking water wells.
Not only did Duke neglect to install any bedrock monitoring wells near neighbors' wells,
the handful of distant bedrock wells it attempts to rely on are improperly constructed. Although
Riverkeepers pointed out this problem months ago, Duke has persisted in relying on these
incomparable monitoring wells in its CAP. See CAP Figure 3-5.
To serve as the basis for determining groundwater levels and directional flow as
compared to residential wells, monitoring wells must draw water from the same range of depths
as residential wells. The depth of the well and the length of the opening in the well are two
related but distinct concepts: a well could be 200 feet deep, but only open from a depth of 100-
105 feet, with the rest enclosed in casing. In this example, contaminated groundwater flowing
through a crevice in the bedrock at a depth of 50, 110, or 210 feet would not be detected in the
well. To serve as a useful comparison to residential wells, the openings through which Duke's
monitoring wells withdraw groundwater must be at a depth and length similar to the openings in
nearby residential drinking water wells.
a]

Contrary to this well-established principle, Duke's bedrock monitoring wells at Buck
extend no deeper than 200 feet below the ground surface, while some residential wells extend
600 or more feet below the ground surface. In an unpredictable, fractured -bedrock aquifer
system such as the one underlying the Buck site, it is not possible to evaluate the likelihood of
contamination to drinking water wells by studying groundwater exclusively at much shallower
depths. Duke's bedrock monitoring wells at Buck also draw water from vertical openings that
are only five feet long (CSA Table 6-9), while the vertical openings on residential wells near
Buck range around 20 to 500 feet in length (CSA Table 4-1). As a result, Duke's test methods
did not replicate the real-world conditions in which the residential wells operate. It is futile and
misleading to compare the static water levels in these dissimilar wells, or to translate those levels
into directional flow of the aquifer at the depths from which residents are drawing their drinking
water.
E. Duke still has not evaluated the water level in residential wells.
Not only is it difficult to compare residential and monitoring wells because Duke has
improperly designed its monitoring well system, it is further impossible to do so because Duke
has not even gathered water level data from residential wells. In the year and a half that Duke
has had to complete its groundwater assessment, it still has made no reported effort to measure
the water level in residential wells. Data on the water levels in residential wells could help
determine whether groundwater flows from the coal ash lagoons into the residential wells.
Duke's failure to gather and evaluate this residential data further compounds its failure to install
monitoring wells on the south side of the Buck site.
F. Duke still has not evaluated the contaminants detected in residential wells.
Despite the fact that Duke collected other data on contamination in residential wells, it
makes no attempt to fully analyze that data in its CSA or CAP. Instead, Duke's CSA merely lists
the residential well data in comparison to the relevant health based standards. See CSA
Appendix B. Although the CSA refers to a conceptual site model that is purportedly contained
in CSA Figures 12-1 and 12-2, this model is not included in the CSA figures and has not been
presented to the public. And Duke's CAP completely ignores the data gathered from residential
wells within a half mile of the Buck site. Duke cannot be permitted to benefit from intentionally
withholding any analysis of whether the contamination in residential wells is transmitted from
the Buck ponds.
G. Duke never re -sampled its wells, or at least has not disclosed any additional
sampling.
As even DEQ pointed out in its conditional approval of Duke's groundwater assessment
plan for Buck, Duke's analysis "should include at least two rounds of groundwater samples
collected and analyzed in 2015" Attachment 1 to Conditional Approval, at 1 (Feb. 24, 2015).
Yet Duke's CAP reports and analyzes only a single round of sampling of its newly installed
monitoring wells. Duke has had almost a half year to conduct a second round of sampling after
this first sampling event. Duke's failure to act during the critical window before DEQ's initial
prioritization should not serve as grounds for downgrading the Buck risk determination.
5

II. Duke's CAP Is Fraught with Unexplained, Internal Contradictions.
A. Duke's entire CAP hinges on so-called "background" wells that Duke itself
admits may not represent background conditions.
The foundation of Duke's CAP is its background analysis. Yet Duke itself admits that of
the four background well sets that it attempts to rely on—BG 1S/D, BG 2S/D, BG 3S/D, and
MW-6S/D—both DEQ and Duke have recognized that three out of four of these wells are likely
located within the influence of the coal ash, and therefore cannot be considered background
wells, and the only remaining deep background well was dry. As Duke and DEQ have
explained, "background wells BG-3S/D and MW-6S/D may not truly represent background
conditions; therefore, installation of additional background wells may be warranted," CAP at 3,
and "[b]ackground monitoring well locations BG-2S/D and BG-3S/D are located in areas which
likely fall within the influence of the ash storage areas," DEQ Response to Duke's Initial
Groundwater Assessment Workplan for Buck at 9 (Nov. 4, 2014). BG-3S/D and BG 6S/D are
located in close proximity to properties with residential wells. Duke has not drilled replacement
background wells in the numerous months that have passed since this recognition in its CSA, nor
in the more than a year that has passed since DEQ first raised this concern. The only action that
Duke has taken in response to these concerns is to shift BG-2S/D slightly south. The only
remaining deep "background" well, BG- IDBG-IBR, was dry, and therefore could not be
sampled to determine background concentrations for all contaminants. CAP at 43.
Moreover, for all of the new wells, Duke has sampled only once—three samples total
between all of the supposed background wells—and therefore its results are not statistically
significant. DEQ has stated that "at least four rounds of data" are necessary in these background
wells. DEQ Conditional Approval, Attachment 1 at 1 (Feb. 24, 2015). And Duke recognizes
that its analysis is not based on "sufficient" or "statistically valid" data set. CAP at 3, 22.
Duke's failure to obtain adequate or representative background samples in advance of the
initial risk designation is inexcusable, and should not serve as the basis for a low risk
designation.
B. Duke identified a background level of hexavalent chromium that cannot be
reconciled with the background level of total chromium identified.
Hexavalent chromium, a potent carcinogen, is a subset of total chromium. It is physically
and logically impossible for there to be more hexavalent chromium in a system than total
chromium. Yet that is exactly what Duke's CAP reports: a background level of total chromium
of 1.9 parts per billion ("ppb"), and a background level of hexavalent chromium at 78 ppb. CAP
at 23. In other words, Duke pretends that there is more than 40 times more hexavalent chromium
than chromium under background conditions at Buck. This absurdity cannot be true.
The only "background" well that Duke sampled that reported 78 ppb of hexavalent
chromium was BG-1BR, which as noted above, Duke also reported as a dry well. Duke has not
explained how it was able to obtain a sample of hexavalent chromium from a dry well. And to
1.1

add to the implausibility, even the total chromium level for this very sample was less than half of
the hexavalent chromium level—which, again, is not physically possible.
Duke cannot be permitted to rely on a background level of hexavalent chromium that is
completely inconsistent with the rest of its CAP.
C. Duke assumes that the ash basins are not sitting in the groundwater under a
cap -in-place scenario, yet the available evidence shows that the ash basins do
in fact extend down into the groundwater.
Duke's CAP "assumes ash left in the ash basin system and ash storage area is above the
water table" under a cap -in-place scenario. CAP at 7. Yet Duke's CAP also shows that the ash
basins extend deep into the groundwater table, see, e.g., CAP Figure 3-2.1, 3-2.2, and that even
under the 33 foot reductions in the groundwater table that Duke assumes, "the groundwater level
will remain in the ash" in Cells 2 and 3. CAP at 62 (emphasis added). Duke offers no support
for its assumption that groundwater levels would be reduced 33 feet. And it makes no attempt to
reconcile the blatant contradiction between its assumption that the ash is above the water table
under cap -in-place, and its conclusion that the ash is below the water table under that same
scenario. Duke's conclusions about the diminished presence of contaminants in the groundwater
under a cap -in-place scenario therefore cannot be trusted or relied upon.
III. Even with Its Numerous Attempts to Conceal the Full Extent of Contamination,
Duke's Analysis Shows That Buck Is High Risk.
A. Even accepting Duke's purported "background" levels, many residential
wells are above these levels.
Even if you accept at face value the background levels that Duke has put forward, these
show that many of the neighboring residential wells at Buck have contamination levels above
background. For example, Duke identifies a background level of total chromium of 1.9 ppb.
CAP at 23, tbl. 2-2. As explained above, hexavalent chromium cannot exceed total chromium,
and so the background level of hexavalent chromium cannot be higher than 1.9 ppb. Yet
numerous residential wells near Buck contain levels of hexavalent chromium that are higher than
this level, and reach levels as high as 22.3 ppb. See DEQ, Well Water Testing,
http://portal.ncdenr.or /g web/guest/wellwatertgging. Similarly, Duke identifies a background
level of vanadium of 8.8 ppb, CAP at 23, tbl. 2-2, yet many residential wells contain much
higher levels of vanadium.
Given that Duke's own reports to date show that carcinogens and other toxins are present
above background levels in residential wells near Buck, the coal ash at Buck must be
characterized as high risk.
B. Duke's own analysis concludes that coal ash contaminants are flowing into
the Yadkin River.
7

Duke's CAP admits that many contaminants in the groundwater beneath its unlined coal
ash lagoons at Buck flow into the Yadkin River, even under a cap -in-place scenario. CAP at 8,
64. For example, the CAP concludes that chromium remains above the health -based standard of
10 ppb—and therefore above Duke's background level of 1.9 ppb—through the next 250 years
in the shallow and transition zone layers, and these layers "discharge to the Yadkin River to the
north and east of Cells 2 and 3." CAP at 64. Similarly, "Vanadium in groundwater discharges to
the Yadkin River north of Cell 1 in the all groundwater flow layers and to the north and east of
Cell 3 in all groundwater flow layers." CAP at 65.
Continued contamination of the Yadkin River for the next 250 years, even according to
Duke's own reports, is another reason that Buck cannot be designated as low risk.
C. Even accepting Duke's closure scenario analysis, contamination decreases
when the ash is removed, showing that the source of these contaminants is
the coal ash at Buck.
A comparison of the "existing conditions" scenario, in which the ash is left in place as -is,
to the excavation scenario, in which the ash is removed to safe, lined storage out of the
groundwater, further shows that contamination at Buck is coming from Duke's coal ash. Under
the existing condition scenario, "all modeled COIs increase or reach steady-state conditions at
concentrations above the 2L Standards, IMACs, or DHHS HSL" during the next 250 years. CAP
at 63. In the excavation scenario, after the ash is removed, many of these constituents decrease,
meaning that the source of the contamination is the coal ash at Buck. For example, hexavalent
chromium remains above the health screening level at the compliance boundary for the next 100
years, rather than 250 years or more, under the excavation scenario. CAP at 66. This further
confirms that hexavalent chromium is sourced from the Buck coal ash, while simultaneously
showing that groundwater remediation must accompany removal of the coal ash source.
For all of these reasons, Duke's own studies show that contamination at Buck is flowing
from the coal ash there, despite the fact that Duke's studies are designed to obscure the full
extent of its contamination. Buck therefore poses a high risk to the people who use and enjoy the
Yadkin River and groundwater in the area, and must be designated as high risk under the Coal
Ash Management Act.
MDB/lap
Sincerely,
Myra Blake
Staff Attorney
n

cc: Stanley (Jay) Zimmerman, Director, Division of Water Resources, Central Office
Bruce Parris, DEQ Mooresville Regional Office
Harry Sideris, Senior Vice President of Environmental, Health & Safety, Duke Energy
Anita LeVeaux, North Carolina Department of Justice
Jim Cooney, Womble Carlyle
0]