Expert Report-Buck Steam Station-GMA-02292016

EXPERT REPORT
BUCK STEAM STATION
1555 Dukeville Road
Salisbury, NC 28146
PREPARED FOR:
Southern Environmental Law Center
601 West Rosemary Street, Suite 220
Chapel Hill, NC 27516
Telephone [phone]
PREPARED BY:
Groundwater Management Associates, Inc.
4300 Sapphire Court, Suite 100
Greenville, North Carolina 27834
Telephone: [phone]
GMA
February 29, 2016
Steven K. Campbell, Ph.D, PG
Senior Hydrogeologist
e�1.11ti1.Y{14i111 j!
CARS
SEAL
1414
.otoff`::'�
4� X
1
Richard K. Spruill, Ph.D, PG
Principal Hydrogeologist
I

I. Introduction and Scooe
Groundwater Management Associates (GMA) has been retained by the Southern
Environmental Law Center (SELC) to provide expert geologic and hydrogeologic
consulting regarding coal combustion by-product (CCB) storage and disposal by Duke
Energy, Inc. (Duke). GMA's services include producing this expert report concerning the
known or potential hydrogeologic impacts of CCBs located at Duke's Buck Steam Station
(BSS) in an open pile and/or within large, unlined pits (aka, basins, ponds, or lagoons).
Information reviewed by GMA indicates that natural stream drainages that historically
discharged surface water to the Yadkin river were dammed to create unlined pits that
were then filled with fly ash and bottom ash (two primary types of CCBs) produced
during operation of the BSS between 1957 and 2013. The CCBs within the pits are in
direct contact with groundwater occurring within a diverse group of consolidated and
unconsolidated geologic materials. Likewise, precipitation infiltrating the large,
uncapped and unlined pile of CCBs located next to one BSS ash pit will deliver leachate
to the groundwater system. Duke and/or their consultants and affiliates (collectively,
Duke) report approximately 4.43 million cubic yards of CCBs are presently located at the
BSS facility. The North Carolina Department of Environment and Natural Resources
(NCDENR) has reported that there are more than 5 Million tons of CCBs in the three BSS
ash pits. The character, quantity, and location of materials other than fly and bottom
ash within the pits at the BSS facility are not known, although liquid effluent and storm
water runoff are discharged to the CCB pits.
Duke states that "The purpose of this .. [Comprehensive Site Assessment]... is to
characterize the extent of contamination resulting from historical production and storage
of coal ash..." (CSA, page 1) in order to comply with the Coal Ash Management Act of
2014 (CAMA). Duke has not divulged volumes or disposal methods for the CCBs
produced for the three decades between initial operation of the BSS in 1926 and the
start of disposal in the pits in 1957. Duke has not indicated that they have performed
any inventory or investigation of other types of potential soil and groundwater
contamination that may be present at the Buck facility, such as polychlorinated
biphenyls (PCBs) or volatile organic compounds (VOCs).
page 1

To date, GMA has been provided, or has obtained from public sources, more than
31,000 pages of documents produced by Duke for the BSS. We obtained and reviewed
publicly -available maps and reports from a variety of sources, including state and federal
agencies such as the United States Geological Survey (USGS), to evaluate the geologic
and hydrogeologic setting of the BSS and the surrounding area. We have used these
data in combination with our professional training and experience to develop opinions
about the geology, hydrogeology, contamination sources and impacts of the BSS coal
ash, and how the associated groundwater contamination threatens off-site properties
and surface water near the BSS. It is obvious that the information available to GMA
does not include all of the data produced by Duke for the BSS site, and significant
fundamental and technical questions remain unanswered. A partial list of resources and
documents that we have reviewed is provided as Appendix A.
II. Qualifications
Dr. Richard K. Spruill and Dr. Steven K Campbell, both hydrogeologists at GMA, are the
authors of this expert report. These experts have visited the immediate vicinity of the
BSS property, and they have examined physical and hydrologic aspects of the area.
Both of these experts have provided litigation support and testified previously regarding
geology, hydrogeology, and environmental contamination. Their descriptions,
interpretations, conclusions, and professional opinions described within this expert
report are subject to revision, expansion, or retraction as additional information
becomes available. A supplement to this report will probably be necessary because
"Part 2" of the BSS Corrective Action Plan (CAP) was first made available five days prior
to the date previously agreed for submission of this expert report, that single document
consists of more than 20,000 pages, and Duke was actively installing 22 additional
groundwater monitoring wells at the time this report was submitted.
Richard K. Spruill, Ph.D, PG, is GMA's Principal Hydrogeologist, president, and co-owner
of the firm. Dr. Spruill's practice is focused on the hydrogeological exploration,
evaluation, sustainable management, and protection of groundwater resources. He has
Page 2

been a geologist for over 40 years, and he is licensed in North Carolina as a professional
geologist. Dr. Spruill has been a faculty member in the Department of Geological
Sciences at East Carolina University (ECU), Greenville, North Carolina, since 1979. He
teaches hydrogeology, mineralogy, petrology, and physical geology at ECU. His resume
is provided in Appendix B.
Steven K. Campbell, Ph.D, PG, is a Senior Hydrogeologist at GMA, he is the firm's
Director of Environmental Services, and he manages GMA's office in Greenville, North
Carolina. Dr. Campbell's practice at GMA includes the hydrogeological assessment and
remediation of contaminated sites, groundwater exploration, and resource development
and management. Dr. Campbell has been a geologist for 35 years, and he is licensed in
North Carolina as a professional geologist and as a certified well contractor. Since 1999,
Dr. Campbell has taught field geology and physical geology in the ECU Department of
Geological Sciences. His resume is provided in Appendix B.
II. Expert Opinions about Duke's Assessment and Remediation
A. Opinions about the BSS Comprehensive Site Assessment (CSA)
In March of 2014, NCDENR sampled and analyzed water discharging at the surface as
seeps and springs located outside the "compliance boundary" for the BSS ash pits.
Some contaminant concentrations reported by NCDENR exceeded the 15A NCAC 2L
groundwater -quality standard, the 15A NCAC 2B surface -water standard, or the
groundwater interim maximum allowable concentration (IMAC). Duke admits in their
CSA that contaminant concentrations measured in samples that they collected in June of
2015 from seeps and springs located inside and outside their ash -pit "compliance
boundary" exceeded those same water -quality standards. This fact is inconsistent with
their claim that "No information gathered as part of this CSA suggests that water supply
wells or springs within the 0.5 -mile radius of the compliance boundary are impacted by
the source, aside from the single permitted well owned by Duke Energy." (CSA, page
ES -3).
Page 3

Laboratory analyses have demonstrated that some water discharging at the surface
inside and outside the BSS compliance boundary is contaminated above water -quality
standards. In September and November of 2014, Duke sampled and analyzed 11 seeps
or discharges, and those data were summarized in December 2014 within a (revised)
"Discharge Assessment Plan" (DAP) that identified 14 seeps or discharges for ongoing
monitoring, sampling, and laboratory analysis. Duke reports flow measurements for 10
seeps in millions of gallons per day within DAP Table 1, which equates to flow rates from
individual seeps as large as 22,500 gallons per day (GPD, or 15.6 gallons per minute).
The cumulative flow for these 10 seeps was 72,300 GPD, or more than 50 gallons per
minute. Flow rates for individual seeps, and greater cumulative flow from all 10 seeps,
are undoubtedly higher during periods of increased recharge to the groundwater system
(e.g., winter). Many of the seeps and springs identified by Duke are located near bodies
of surface water.
Duke admits that ash -derived groundwater contamination extends outside the BSS
compliance boundary. Duke claims that groundwater contamination derived from their
ash sources "... is generally limited to an area within the ash basin compliance boundary
and the area north of the compliance boundary near the Yadkin River..." and "Some of
these [ 2 L or I MAC] exceedances were measured outside the compliance boundary..."
(emphasis added, CSA, page ES -15).
Although Duke continues to deliver bottled drinking water to area residents who rely on
groundwater pumped from the fractured bedrock aquifer, they claim that they have
... found no imminent hazards to public health and safety... [and] ... no actions to mitigate
imminent hazards are required." (CSA, page ES -15). Duke states that the CSA
investigation has "...identified the horizontal and vertical extent of groundwater
contamination within the compliance boundary.." that originated from their ash pits
(CSA, page ES -15). However, that claim of full delineation is inconsistent with (1)
contaminated groundwater discharges at seeps and springs, (2) Duke's identification of
numerous "data gaps" in the CSA, including a need to install additional monitoring wells,
(3) the fact that Duke began installing 22 new monitoring wells around the periphery of
their compliance boundary in January of 2016, many months after the CSA was
Page 4

submitted to NCDENR, and (4) Duke's heavy reliance on very poorly -established "natural
background" contaminant concentrations that exceed applicable water -quality standards
to bolster their general assertion that ash -derived groundwater contamination is
confined to small portions of the Buck property.
Duke's CSA is a massive document (>9,400 pages) that is only available to the public
through the North Carolina Department of Environmental Quality (NCDEQ, formerly
NCDENR) website: http://edocs.deq.nc.gov/WaterResources/Browse.aspx?isTimeout=1&dbid=0. The
downloadable CSA is missing all but one monitoring well construction record and at least
three figures (11-1, 12-1, and 12-2), including the site conceptual model (SCM), and
those deficiencies have not been corrected as of February 25, 2016.
It is our opinion that Duke's CSA is seriously flawed, and it does not provide an
adequate basis for production and implementation of a Corrective Action Plan (CAP).
More specifically, we opine that the CSA (1) contains an incomplete and simplistic
evaluation of on-site and off-site geology and hydrogeology, (2) is not an adequate
assessment of the fractured, crystalline, bedrock aquifer system that is the primary
source of water for all potable water consumers located near Duke's facility, and (3) it
disregards well construction, pumping -induced influences, and groundwater analytical
data for off-site, potable water -supply wells.
A.I. Duke's evaluation of on- and off-site geology and hydrogeology
is simplistic, incomplete, and inadequate
Duke's assessment of on-site and off-site geology and hydrogeology is simplistic,
incomplete, and inadequate as a basis for remediation (i.e., preparation and
implementation of a CAP). Duke's investigation of the geology and hydrology appears
designed primarily to (1) defend the hydrogeologic site conceptual model (SCM) that
was described in Duke's pre -CSA "Proposed Groundwater Assessment Work Plan (Rev.
1)" dated December 30, 2014, and (2) to support and constrain Duke's intent to rely on
SCM -validating computer simulations of groundwater flow and contaminant fate and
transport (F&T) within their CAP. Section 14 of the CSA acknowledges "data gaps" (i.e.,
Page 5

deficiencies), and Duke states explicitly in Section 15 that additional assessment
information and analytical data would be supplied to NCDEQ in a "CSA Supplement"
prior to submission of the CAP. However, the content of the supplement' is not
specified, Duke produced no 'supplement' prior to the CAP, and several tasks that Duke
identified to address CSA deficiencies were not described within the CAP Part 1 (see
Section B below). The TSA Addendum" included with CAP Part 2 is 3,384 pages
dominated by raw data or tabulations that are generally devoid of explanation,
interpretation, and/or integration into the CSA or the CAP (see Section C below).
Duke's geologic characterization of the BSS property was focused on
unconsolidated to semi -consolidated materials (collectively, regolith) overlying the
fractured, crystalline bedrock that transmits the groundwater used by water consumers
at and near the BSS property. Significantly, Duke has failed to demonstrate that most of
their so-called background' monitoring wells are located hydraulically'upgradient' of
their ash pits, even though their all-important designation of background'
concentrations of self -identified "constituent of interest" (COI; i.e., contaminant)
depends on that'upgradient' assertion. Duke admits in the CAP that two of their four
background' well pairs (i.e., BG-3S/BG-3D and MW-6S/MW-6D) "...may not truly
represent background conditions.." (CAP Part 1, Section ES -2.1, page 3), and they
repeatedly state that their critically -important bedrock background' monitoring well BG-
1BR is "dry." Meanwhile, Duke claims that "...one isolated exceedance..." of a
groundwater -quality standard is adequate reason to stop evaluating some COIs (e.g.,
arsenic), while they simultaneously use a single detection in a 'background' well to
create "proposed provisional background concentrations" (PPBCs). Duke provides even
more startling admissions regarding their background' wells, background' contaminant
concentrations, and their PPBCs, in CAP Part 1, as addressed below in Section B.
Duke determined the depth to the top of the fractured bedrock and reported
compositional variations at their borings and core holes, but they failed to evaluate
adequately the type of rock (lithology) or structural features (e.g., fractures). Duke's
"Site Geologic Map" (CSA figure 6-1) is not a geologic map, nor does it provide any
Page 6

meaningful information or interpretation about the spatial distribution of, or variations
in, the bedrock composition or structural geology inside the BSS property.
Duke's superficial geologic evaluation outside the BSS property boundaries is
exemplified by their fundamentally meaningless "fracture -trace study," which identified
selected linear features' that they admit "...may or may not correspond to actual
locations of fractures.." and the "...results are not definitive..." (CSA page 31). GMA's
cursory review of Duke's fracture -trace' maps identified obvious linear features' that
were overlooked or ignored, further demonstrating the inadequacy of their basic
geologic assessment.
Evaluating spatial complexity within the subsurface is a standard objective of all
CSAs, yet Duke chose not to use their on-site boring and core hole data to complete this
fundamental task. To demonstrate what Duke could have done, GMA used the CSA
geologic descriptions for borings that encountered bedrock, and we contoured the depth
to the top of bedrock (overburden thickness) at the Buck property using the Surfer° 12
software program. The resulting map provided below reveals complexity not
incorporated in Duke's SCM or illustrated within the CAP's groundwater modeling (see
Section B). One implication of GMA's top -of -bedrock map is that ash -derived
contamination has a greater potential for entering the fractured bedrock aquifer in areas
where rock is located relatively near the surface beneath the BSS ash pits. Strong
contrasts in porosity and permeability of fractured rock versus a heterogeneous package
of overburden further complicate the potential distribution of ash -derived contaminants
in the subsurface.
Page 7

There are significant problems with production and interpretation of the
equipotential maps included in the CSA (and repeated verbatim in CAP Part 1). These
problems include failing to divulge the correct date(s?) of water -level measurements
used to create the maps, failing to identify data that were selectively used or omitted
during contouring of hydraulic head, not identifying the computer software or other
method(s) used to perform or 'adjust'the contouring, and constraining the contouring to
specific portions of the BSS property to create a favorable impression of local
groundwater flow in some areas.
For example, CSA Figure 6-6 shown below is Duke's interpretation of the
distribution of hydraulic head measured in their "deep" monitoring wells (i.e., shallowest
bedrock or "transition zone"). Contouring near areas where surface water is visible in
ash pits 2 and 3 indicate to us that Duke combined surface water elevations with the
"deep" monitoring well hydraulic head data, which is not a sound hydrogeologic
practice. The resulting map shows equipotential contours that indicate locally -steep
Page 8

hydraulic gradients near ash pit surface water, which in turn implies local groundwater
flow that is not realistic.
POT—ON EYRIC sUREACE_DEEP —L& (D)
�o
8166DUKE E ERG
Y CARDLIHAS, LLC
FN gIlGU5T201S
.unLe fryBUCK
sTTau S—ON ASH B -IR
&6
Page 9
Iti
LEGEND
' '��
`JCOALA6H
a.. r�' �� �.�
WASTE®d.11)�ARY
✓
{
uia�reemnu_smrwcelr-rw.^�I
O H�cKcauPLu cE eauNo Hv
12C
/ ex -H 1
ANONIiowNc wEu.
//�'� :.• M6-11° `
d.4:d
- TA NOT USE�eY H9%t
/ -' a<5 I �. ..15-0G
'y � 1 {p _ da8.'
R£PItFSENTATNE
�{.ROtFIL�iAIER FLOW
dil SL /f
1 r
r�
�t
h14ti'-'l
E 4 /// 23 �1 --
D
CIMPREHENSIVEE SITE
ASSESSMENT
RNBUCK
STEAM STATION
ASH HASRN
Al1GUSF 23, Zd15
yd
1
111J yC
d98.1 IA':i'3U
96E Aid
COUNTGUR INTERVAL= 5 FT
0 -5d 1.1m
t ,-, a --
jDO.
2.:-
hh4 J
:•;•! ,d �
p •'f`-"
FRIVILEG EU AND C—DENTAL,
ArORNEY-CLIENT COMMUNICATION,
ATTORNEY WORN PRC'°LIT
DRAFT WORK --T
EQ NTIALMAP
OF
Qf THE "DEEP"
P-'
MONITORING WELLS,
_ --
BUCKFACILITY,
-c
ROWAN COUNTY NC
r
l
N
Cate_ 212512 0 1 6
Page 9

GMA used Duke's data for the same wells and the standard contouring software
program Surfer® 12 to produce the alternate "deep" monitoring well equipotential map
shown above. The computer program contoured the data, and we did not adjust, add,
or remove any data point or equipotential contour. Our map does include head
measured in two monitoring wells that Duke elected to ignore during their contouring
(see the GMA map legend) because we did not find any justification by Duke for their
omission of those data. We have added representative groundwater -flow lines to
illustrate the horizontal component of groundwater flow in the general vicinity of
residential water -supply wells located near Leonard Road east and southeast of Duke's
property. The equipotential map indicates that those residential wells are at greater
potential risk from ash -contaminated groundwater migrating generally eastward. As
described elsewhere in this expert report, pumping of wells can create local influences
that cause deviations in patterns of hydraulic head and groundwater flow.
In several places within the CSA, Duke perpetuates the common misconception
that groundwater flows horizontally in the subsurface, typically directly toward a stream
valley or the Yadkin River. Duke did not create a flow net or employ site-specific
hydrogeologic cross sections that depict the subsurface distribution of hydraulic head
and flow paths that show the three-dimensional aspects of groundwater migration
pathways. Nevertheless, Duke admits that the "... vertical gradient of groundwater is
generally downward across the site." (CSA, page 76). The fact that downward -directed
hydraulic gradients facilitate the introduction of contaminants to deeper portions of the
aquifer system is glossed over or outright ignored throughout the CSA.
Duke produced copious data for aquifer parameters such as horizontal hydraulic
conductivity, yet the CSA characterizes entire "hydrostratigraphic layers" using averaged
data or range end -members. Duke's calculations for the rates of groundwater migration
using those data indicate that groundwater movement through fractured crystalline
bedrock (-400 to 229,000 feet/year) is orders of magnitude faster than it is through
regolith (N5 to 92 feet/year) (CSA, page 77). The relatively high hydraulic head at the
saturated coal ash basins will drive contaminated groundwater downward through the
ash and underlying regolith until it enters the complexly fractured bedrock aquifer,
Page 10

where it migrates rapidly in response to fracture orientations, fracture connectivity, head
distributions, and hydraulic influences from pumping by area water -supply wells.
Pumping of area water supply wells in relatively low -permeability bedrock will induce
steep gradients around the pumping wells, which in turn greatly accelerates water flow
toward the pumped wells. The net result is that pumping water -supply wells can cause
ash -derived contaminants to migrate in directions different than may occur naturally
within an un -pumped aquifer, including 'upgradient' of Duke's coal ash pits. Transient
real-world well pumping can easily reverse the local ambient' hydraulic gradient around
wells during and after the pumping event, thus drawing contaminants in directions
significantly different from the groundwater -flow pattern derived from water -level
measurements from monitoring wells.
Duke's relies entirely on generic concepts of groundwater hydrogeology in piedmont
settings to fabricate their site conceptual model', and they did nothing significant to
refine or challenge their SCM between the time that it was described in their pre -CSA
"Proposed Groundwater Assessment Work Plan (Rev. 1)" and submission of their CAP
(Part 1 and Part 2). Fatal flaws in Duke's SCM include the failure to consider and
incorporate natural complexities of groundwater flow in the heterogeneous fractured -
rock aquifer, and disregarding the effects of groundwater withdrawals by numerous
water -supply wells.
It is common knowledge that the Piedmont of North Carolina is characterized by
heterogeneous geology and hydrogeology, and this is true at and near the Buck
property, including heterogeneous distributions of groundwater contamination.
Remarkably, Duke claims that "Heterogeneities, with regard to groundwater flow, were
not identified during completion of the CSA." (CSA page 10 1) This statement is
obviously false; for example, Duke reports that hydraulic conductivity (K) measured for
the fractured bedrock aquifer at the BSS property varies by four orders of magnitude,
which clearly illustrates the existence of a heterogeneous aquifer that must produce
complex groundwater flow.
Page 11

Duke claims that their SCM will be "...refined through completion of groundwater
modeling...", but using a computer model to'refine'the conceptual model that forms the
fundamental framework of the computer model is an inherently circular, self-fulfilling
exercise that does nothing to'refine'the initial assumptions. The SCM should reflect
direct observations from the site and the surrounding area, and it should be refined
constantly as new data are produced. The groundwater model should be constructed as
an accurate reflection of site conditions that were determined by actual field assessment
and data interpretations.
When addressing site'data gaps'that impact their SCM, Duke states that
Heterogeneities, with regard to COI concentrations, were not identified during
completion of this CSA." (CSA page 101) This is another obviously false claim; the CSA
demonstrates that COI concentrations vary considerably in the horizontal and vertical
directions across the Buck property, ash -derived contamination is present in the
heterogeneous fractured -bedrock aquifer, and Duke explicitly acknowledges that
groundwater contamination extends outside the BSS compliance boundary at monitoring
wells and at natural groundwater discharge points (i.e., seeps and springs).
In March of 2014, NCDENR sampled and tested water discharging at several
seeps and springs located on Duke's property. NCDENR identified six substances in
sampled discharges at concentrations exceeding the applicable 15A NCAC 2L
groundwater standards, four of which exceeded the Class WS 15A NCAC 2B surface -
water standards. In September and November of 2014, Duke sampled and analyzed 11
discharges, as summarized in December 2014 within their (revised) "Discharge
Assessment Plan" (DAP). However, the CSA does not mention Duke's 2014 sampling
and analysis, and instead is focused on their sampling and analyses performed in June
and August of 2015. Although Duke claimed that most of the seeps or springs were
'dry' in mid -2015, they reported that concentrations of 10 targeted COIs exceeded the
15A NCAC 2L or IMAC groundwater -quality standards in water samples. The CSA
generally treats NCDENR's early -2014 spring analyses as unrelated to Duke's mid -2015
seep analyses (CSA Tables 7-9 and 7-10, respectively), instead of synthesizing all these
analytical data (including Duke's 2014 analyses). Likewise, CAP Part 1 (see Section B
Page 12

below) provides no new data or synthesis regarding water discharges to the surface,
and we have seen no evidence that Duke is implementing the DAP and reporting
analytical data (if any) to NCDEQ.
In spite of this flawed effort, it is clear that some contaminant concentrations
reported by Duke and NCDENR exceeded a 15A NCAC 2L groundwater -quality standard,
a 15A NCAC 2B surface -water standard, or a groundwater IMAC. Specifically, Duke
admits that contaminant concentrations measured in samples that they collected in mid -
2015 from seeps located outside their ash -pit "compliance boundary" and immediately
adjacent to the Yadkin River exceeded those same water -quality standards.
A,2. Duke's evaluation of the fractured bedrock aquifer system is
simplistic, inadequate, and intentionally limited in scope
Duke's investigation and characterization of the fractured bedrock aquifer system
is simplistic, inadequate, and is focused intentionally on very short zones within the
uppermost bedrock at a few locations on the BSS property. Duke has essentially
ignored the regolith and bedrock aquifer systems outside the BSS property boundaries.
The CSA devotes only seven sentences to the structural geology at the Buck facility, and
a single generic sentence describes the influence of structural geology (e.g., fractures)
on groundwater flow. Duke produced data that could be used to evaluate the depth to
the top of bedrock and the topography of that surface, but they chose not to perform
even that basic structural evaluation of the bedrock geology and aquifer. Duke chose
not to employ down -hole geophysical tools (e.g., acoustic and/or optical televiewer) that
are commonly used to assess fractured bedrock aquifers. Duke's core descriptions
report numerous fractures with a wide variety of orientations ('dip' angle), but those
data were not evaluated, and their descriptions have limited use because the cores were
not oriented during collection. Geophysical tools can no longer be used in the original
core holes to determine fracture orientations and density because Duke converted the
holes to wells using screen -and -gravel pack construction. Therefore, Duke has avoided
using standard investigative tools appropriate to understanding fracture orientations,
distribution, and associated influences on groundwater and contaminant migration in the
Page 13

bedrock aquifer system. Such evaluations are critical components of any hydrogeologic
evaluation of fractured bedrock aquifers, and incorporating such data is essential to any
meaningful SCM and groundwater computer simulations.
Duke's bedrock' monitoring wells employ screens that are only 5 feet long.
Comparisons of rock core descriptions with the wells installed at those locations show
that Duke placed very short screen across a specific fracture interval, typically within the
shallower portion of the cored interval. The CSA provides no generic or well -specific
explanation of criteria used to select well screen placement. For example, the portion of
the CSA boring log for bedrock monitoring well GWA-913R shown below reveals that
substantial fractures are present at least 42 feet below the bottom of the well's screen.
BORING NUMBER GWA•9BR
Wf f rn
FP ,some soo BORING NUMBER GWAABR
��.suaTm rncf a or ,o
- St fires
PI®6.l'MMB®1244R1 PR(�T4T TOCITMM Roman Crone we
ooPn Eatlon Ml.�evumcn. Fe
LQ -xi V'9�glp, FacwY,lYm W.
TM�m Pawry qn¢, hIP'0 Mn£osN
' ]U.a'. ]Ir V Re�ei6 ffi'. Fe Mie Mn
frog
a]' fl�wfi�, li n
P¢rp M.. AswJ
,��,yy 3bJ1'+hwfwl'F'�Yd.Eim
i $� N lei BuiFp,EYAue
�l.ffelevel6".falenxOON.n.
"Ternunbf of RtR-Ld
Duke appears to have ignored completely the locations and construction of area
residential water -supply wells when planning and installing their'bedrock' monitoring -
well network.' Area water -supply wells typically have steel pipe (well casing) installed
to the top of the bedrock and an open hole bored through tens to hundreds of feet of
fractured rock below the well casing. For example, the public water -supply well serving
Page 14

the Buck facility has 82 feet of well casing and is 500 feet deep, so that well is open to
418 feet of fractured bedrock, or more than 80 times longer than their'bedrock'
monitoring wells equipped with 5 feet of well screen. As an aside, Duke makes no
mention that their on-site water -supply well had water -quality violations for sulfate and
iron at least as recently as February of 2011.
We conclude that Duke chose to evaluate only tiny intervals of the shallowest
portion of the bedrock aquifer system at a handful of locations on the BSS property, and
they did not perform any investigation of off-site hydrogeologic conditions. Duke's
sparse on-site 'bedrock' monitoring -well 'network' leaves hundreds of acres of the
bedrock aquifer across the BSS property unevaluated, especially the southern half of the
property and in the northwestern portion of the property nearest the Yadkin River, as
shown in the following map.
Le9ena
0.5 M ILE OFFSET FROM ASH BASIN
RECEPTOR MAP- AERIAL BASE A—11—
J ENERGY CAROLINAS, LLC
Bl1CK STEAM STATION ASH BASIN ua=
-.axwm xaw.nmx.�.un xom:crm�n w�Aeuw.. wu.sxns.wEo x.��� 1 ROWAN COUNTY, NORTH CAROLINA
A more appropriate approach to assessing the bedrock aquifer would include
identifying the construction of area water -supply wells and assessing the yield, depths,
and fracture orientations of the water -producing intervals that supply those wells. Duke
Page 15

could then undertake an assessment on and near the BSS property to characterize the
full thickness of the bedrock system that is being used for potable water supply. As
described in Section A.3 below, Duke has avoided assessing the portions of the bedrock
system that are known sources of drinking water for the adjacent community. Duke's
incomplete assessment places the community at greater risk of exposure to
contaminants because the most likely pathways of contaminant migration that could
impact human health were ignored.
A,3. Duke has ignored the presence and influence of off-site, potable
water -supply wells, including groundwater analytical data
available for sampled wells
Duke identified almost 200 known or likely water -supply wells within 0.5 miles of
the BSS Compliance Boundary'. Most of these wells extract groundwater from the
fractured, crystalline bedrock aquifer to supply potable water to area residential and
commercial users. Duke's "Proposed Groundwater Assessment Work Plan (Rev. 1)"
stated that the influence of pumping water -supply wells on groundwater flow would not
be considered, and that important impact on groundwater and contaminant migration is
indeed absent from the CSA. Duke's failure to consider, evaluate, and incorporate the
effects of pumping by dozens of supply wells on the groundwater system is a
fundamental flaw of the CSA. No meaningful SCM, computer simulation, or remedial
plan (i.e., CAP) can be produced without considering all influences on the groundwater
system, and especially impacts on the heterogeneous, anisotropic fractured bedrock
aquifer suppling potable water to the surrounding community.
The CSA devotes very little text to describing the potential risk that groundwater
contamination derived from BSS coal ash poses to the numerous off-site water -supply
wells located near Duke's property. Duke made no effort to evaluate or integrate the
analytical data produced by the "NCDENR Well Water Testing Program" for
approximately 90 nearby residential water -supply wells. Duke states repeatedly that
BSS -derived groundwater contaminants (aka, COI) pose no real or potential concern for
any water -supply well located within at least 0.5 miles, specifically claiming that "A// of
Page 16

the off-site private water supply wells are upgradient of the ash basin..." including "...off-
site receptors immediately south and east of the Buck site..." (CSA pages 86 and 105,
respectively). Duke's blanket statement "No information gathered as part of this CSA
suggests that water supply wells or springs within the 0.5 -mile radius of the compliance
boundary are impacted by the source. "(CSA page 110) demonstrates that they will not
accept responsibility for, let alone investigate, any groundwater `COI' detected outside of
the Buck property boundaries. Nevertheless, Duke has supplied bottled drinking water
to homes located near the BSS site for many months, a fact that is conspicuously absent
throughout the CSA and CAP.
B. Opinions about the BSS Corrective Action Plan (CAP) Part 1
The CAP is difficult to evaluate because "Duke Energy and NCDEQ mutually agreed to a
two-part CAP submittal...", and a risk assessment will be submitted under a separate
cover with the CAP Part 2..". CAP Part 1 is the focus of our opinions described in
Section B because of the short period between Duke's submission of CAP Part 2 and
production of this expert report. Our preliminary impressions of CAP Part 2 are provided
in Section C of this expert report, but we anticipate that a supplement will be required.
Numerous figures in the CAP (Part 1, unless otherwise specified) are devoid of all
information or are illegible, and the public can only obtain the CAP through NCDEQ's
website http://edocs.deq.nc.gov/WaterResources/Browse.aspx?startid=221202&dbid=0. CAP Part 1 is
563 pages long, not including an unknown number of datasets, electronic files, and
iterations of computer simulations.
In this paragraph of our expert report, for sake of argument and without accepting the
claims as correct, we are taking at face value (1) Duke's "proposed provisional
background concentrations" (PPBCs) for their self -identified "constituent of interest"
(COI; i.e., contaminants), (2) Duke's groundwater modeling and contaminant fate and
transport (F&T) simulations at face value, and (3) Duke's assumptions and claims
regarding their preferred 'cap -in-place' remedial solution to the coal ash pits. Duke
concludes that the following COIs are"...attributable to the source area atBuck...":
Page 17

antimony, boron, chromium, hexavalent chromium, cobalt, iron, manganese, nickel,
selenium, sulfate, total dissolved solids (TDS), thallium, and vanadium. (CAP, pages 45-
50) Duke concludes that ash -derived groundwater contaminant concentrations will
exceed the 15A NCAC 2L or IMAC standards "...at the compliance boundary and at the
Yadkin River..." (CAP, page 64) for numerous COIs, for decades or centuries into the
future. For example, Duke states that chromium concentrations "...exceed the 2L
Standards in all three groundwater flow layers at the compliance boundary and at the
Yadkin River at the start of the [computer] simulation period..." and "... concentrations
are predicted to increase and peak within the 250 -year simulation period..." (CAP, page
64) Duke makes similar conclusions for other COI under their preferred cap -in-place
'remediation' solution for the BSS ash pits. Collectively, Duke is acknowledging that they
have violated groundwater -quality standards for multiple contaminants, both inside and
outside their coal ash compliance boundary, and those violations will persist for decades
to centuries into the future. Duke further acknowledges that they will discharge
contaminated groundwater to the Yadkin River over that same time span.
We have three fundamental criticisms of Duke's CAP (Part 1): (1) the CAP is built upon a
simplistic, incomplete, and inadequate CSA and SCM, as discussed previously in Section
A, (2) Duke's claims about 'background' water quality rely upon data obtained from
several background' monitoring wells that they admit are not proven to be located
hydraulically upgradient of coal ash pits, and (3) the groundwater flow and contaminant
fate and transport (F&T) computer model employs a seriously flawed, and greatly
simplified, hydrogeologic framework that does not consider the complexities of
groundwater migration in fractured bedrock and ignores the influence of pumping from
many dozens of water -supply wells located near the Buck property.
B,1, Duke's CAP cannot be considered reliable because it relies upon
a simplistic, incomplete, and seriously -flawed CSA
The primary objectives of a CAP should be to identify and implement methods to
eliminate, or significantly reduce, risks and impacts to human health and the
environment, identify and evaluate site-specific remedial options to rectify contamination
Page 18

impacts, and choose the justifiably best option(s) that meet or exceed regulatory and
site-specific conditions and objectives. These interrelated objectives can only be
achieved by first producing a comprehensive assessment of the geology and hydrology,
and fully delineating the three-dimensional distribution of contaminants emanating from
the site. Duke's CSA fails to accomplish these essential tasks for the Buck property, and
even those few CSA deficiencies that Duke acknowledged were not addressed in the
promised "CSA Supplement" or in the CAP. (The "CSA Addendum" provided as Appendix
A of CAP Part 2 is addressed below in Section C.)
The CAP never mentions that Duke is supplying bottled water to off-site homes,
yet it misrepresents a primary risk to the public by claiming that "the public drinking
water supply source for the surrounding area of Buck...[is] ... the City of Salisbury
(Salisbury -Rowan Utilities,' which withdraws water from "...the Yadkin River...
approximately 5 miles upstream of the Buck site." (CAP page 14). This "public" source
of water is not available to any home or business located within 1.5 miles of the Buck
property because the nearest Salisbury -Rowan Utilities potable water line ends
approximately 1.5 miles west of the intersection of Long Ferry Road and Leonard Road,
near Interstate 85 at the intersection of Allen Lane and Long Ferry Road. Homes and
businesses located near the BSS generally have no viable alternative to using their wells
to obtain potable water.
B,2. Duke's claims about contaminant 'background concentrations
are genera//y not supported by their own data and are often
presented in a confusing or disingenuous manner
The CAP lists PPBCs for Duke's self -identified COIs that were allegedly
determined from evaluating groundwater analytical data for so-called 'background'
monitoring wells. It appears that Duke's background' analytical data may have been
processed to varying degrees, but the actual datasets are not provided in CAP Appendix
B (Background Well Analysis and Statistical Evaluation), the CAP does not indicate where
those data are located or presented, it is impossible to know which data outliers' may
have been removed' during their evaluation, and the CAP generally does not explain
Page 19

how and why data were, or were not, considered. Tables provided in CAP Appendix B
indicate that statistical treatment was limited to analytical data for the MW -6S and MW -
6D well pair, probably because those are the only 'background' wells that Duke has
disclosed to have been sampled and tested repeatedly. However, Duke admits that
wells MW -6S and MW -6D are not suitable for use as'background' monitoring wells
because "The source of groundwater in we/l5 MW -65 and MW -6D is uncertain... [and]
... localized groundwater flow may be from the direction of the ash basin to the north",
which places them "... downgradient of Cell I" (CAP Part 1, Appendix B, page 4).
Likewise, 'background' wells BG -3S and BG -3D are not suitable because the wells are
located "...close to cell 1 (590 feet) and groundwater elevations in this area (June 2015)
are 14 feet lower than Cell 1 water level. This well [pair] may be downgradient of Cell
1." and the BG-3S/D "...well group will continue to be evaluated for its suitability as a
background well as more water level and analytical data are collected." (CAP, Appendix
B, page 6) Collectively, Duke admits that two of their four 'background' monitoring well
pairs (i.e., BG-3S/BG-3D and MW-6S/MW-6D) are not definitively located hydraulically
upgradient of the ash pits, and thus they cannot be used to establish naturally -occurring
'background' concentrations of detected contaminants.
Duke's single'upgradient' bedrock monitoring well (BG-1BR) is consistently
identified as a'dry' well throughout the CSA and CAP (Part 1), which one would infer
means that groundwater samples cannot be obtained for analysis. Well BG-1BR is not
mentioned in CAP Appendix B, yet the analysis of carcinogenic hexavalent chromium
(Cr(VI)) reportedly produced for a water sample obtained from BG-1BR is the sole
source of Duke's claim that the 'natural' background concentration of Cr(VI) is 78
microgram per liter (fag/L). Duke's claim that high concentrations of hexavalent
chromium occur naturally at and near the Buck facility is discredited by Duke's PPBC for
total chromium (1.9 fag/L), which by definition must include all CR(VI). Duke also
ignores the fact that 14 sampled water -supply wells contained Cr(VI) at concentrations
greater than their PPBC for total chromium. (CAP Part 2 describes sampling and
analysis of'dry' well BG-1BR in October and November, 2015, and we address those
results in Section C of this expert report.)
Page 20

All of these geologic, hydrologic, and analytical background well' problems were
known by Duke prior to submission of the CSA, yet the CAP produced several months
later shows no evidence that these serious issues were addressed. Nevertheless, Duke
uses analytical data from non -background monitoring wells to (1) argue for natural'
contaminant concentrations that they use to downplay risks to human health and the
environment, and (2) propose alternative cleanup' goals for the BSS facility. Duke
specifically states that if NCDEQ approves the "...PPBC concentrations.." for cobalt,
hexavalent chromium, iron, manganese, and vanadium, then the PPBCs for those five
contaminants "... will be used for identifying groundwater quality exceedances of COIs
instead of the 2L standards, IMACs, or DHHS HSLs.." (CAP, page 27) Because Duke has
not established naturally -occurring background' concentrations of their COIs in the
groundwater system at or near the Buck site, they cannot identify appropriate remedial
goals or propose alternatives to established standards for groundwater contamination.
Duke chose the "highest laboratory reporting limit for non -detects" for some
'background' concentrations to fabricate a PPBC, while it appears that they assigned a
random number for other PPBC (e.g., antimony = 1 fag/L). Duke makes no distinction
between analytical data for their "shallow," "deep," and "bedrock" monitoring wells,
which indicates that they are assuming that a single, homogenous background'
concentration exists throughout the area in all saturated geologic media and at all
depths for each COI. Although Duke repeatedly claims or implies that elevated
background concentrations' of COIs are commonplace at and near the BSS property,
they have not completed a credible evaluation of natural 'background' concentrations for
gny tested analyte. CAP Table 2-2 lists ranges of "Regional Background Groundwater
Concentrations" for 10 COIs, but the number (n) of analyses used to create the
concentration ranges are not disclosed. Likewise, Duke sampled and tested an
undisclosed number of their employee's residential supply wells reportedly located
between "2 and 10 miles" from the Buck property in mid -2015, a fact omitted from their
CSA. The "2-10 Private Well Data" reported in CAP Table 2-2 indicate that vanadium
ranged from <1 fag/L to 16.4 fag/L. Because Duke does not divulge or evaluate the
underlying '2-10' data, this effort provides no insight into background' concentrations in
the region, and certainly not near the Buck facility. However, if one assumes that n=10,
Page 21

then it is possible that vanadium in 9 of the '2-10' analyses was <1 pg/L and a single
anomalous analysis contained 16.4 pg/L, or, perhaps, the opposite is true. If Duke had
acknowledged that 13 NCDENR-sampled water -supply wells near the Buck property
contained vanadium concentrationsra eater than their 8.8 pg/L PPBC, then perhaps they
would feel compelled to apply more rigor to their contaminant 'background' evaluations.
The actual distribution of contaminant concentrations can produce a significantly
different impression than what one may get from the concentration ranges in CAP Table
2-2. For example, Duke cites the USGS' National Uranium Resource Evaluation (NURE)
database (http://mrdata.usgs.gov,) as the source for vanadium's 'regional background'
concentration range (<0.1 pg/L to 42.8 pg/L) detected in wells located within 20 miles
of the BSS site. GMA's inquiry of the NURE database (shown below) on February 1,
2016, shows n=86, the average (mean) vanadium concentration was 3.35 pg/L, and
roughly two-thirds of these 86 analyses were at or below 2.7 pg/L. We are not
proposing that the naturally -occurring 'background' concentration for vanadium is 2.7 or
3.35 pg/L, but we are saying that Duke has failed to perform the evaluation required to
establish gny natural 'background' concentration for gny'COI' at or near the Buck
property.
Statistics cif KURE Mnile resuhs
FiNt1
FFrequency Distribution
slat3ticv
80
count8f
Minimum: -01
Maximum: 42.9
ao
Sum: 238.5
Maan. 3.3541x51
40
StaedardDeviatlan: ED98049
20
0
LZ-1
shot of 2/1 12416 inquiry
0 17 26 35
The CAP lists PPBCs at concentrationsrQ eater than the applicable groundwater -
quality standard for the five metals identified in the table provided below. Four of those
metals (cobalt, iron, manganese, and vanadium) are the most common COI to exceed
water -quality standards in sampled seeps and springs located in areas topographically
Page 22

downgradient of the BSS coal ash pits. Duke proposes to use those five PPBCs to
establish "...groundwater quality exceedances of COls instead of the 2L standards,
_TMA C5, or DHHS HRR..." (CAP, page 27)
Hexavalent 0.07 78
Chromium
Cobalt
Iron
4
300 370"
Manganese 50 160
Single highest value (78 dig/L in well BG-1BR)
Single highest value from MW -6 well pair; n=2
indicates one sample date for the two wells
Average parametric 95th Upper Prediction Limit
(UPL) from the MW -6 pair
Average parametric 95th Upper Prediction Limit
(UPL) from the MW -6 pair
Vanadium 0.3 8.8 Single highest value from MW -6 pair; n=2
indicates one sample date for the two wells
* The groundwater standards listed are per 15A NCAC 2L, IMAC, or the NCDHHS HSL.
Concentrations listed are in micrograms per liter (pg/Q. Duke proposes a PPBC of 1.9 pg/L
for total chromium, or 41X their PPBC for the hexavalent chromium valence state (Cr(VI)).
^ Value rounded to the nearest whole number from Duke's PPBC of 369.81 pg/L.
Although Duke claims that "...one isolated exceedance..." of a groundwater -quality
standard is an adequate reason to stop evaluating some COIs (e.g., arsenic), they are
willing to establish PPBCs based on a single detection in a so-called 'background'
monitoring well for others (e.g., vanadium). Duke proposes a PPBC for hexavalent
chromium that is more than 40 times greater than their PPBC for total chromium, which
by definition includes all Cr(VI). The methods that Duke uses to discount some
concentrations of their self -identified COIs and to fabricate PPBCs eliminates the
credibility of their proposals to employ less stringent, BSS -specific remediation criteria.
Until Duke conducts a much more rigorous and credible evaluation of'background'
concentrations, they should be required to remediate contamination to the applicable
groundwater standards, including the NCDHHS HSLs for hexavalent chromium and
vanadium.
Page 23

B.3. Duke's groundwater flow and contaminant fate and transport
computer model is based on an incomplete and inadequate assessment of
the geologic and hydrogeologic conditions at the site, and it does not consider
the complex, heterogeneous nature of the fractured bedrock aquifer system
The CAP is inherently flawed because it is built on an inadequate and incomplete
CSA that was biased toward validating the most favorable situations that Duke could
imagine for their unlined coal ash pits. The groundwater -flow model and contaminant
fate and transport (F&T) simulations suffer from many of these same issues, including
(A) reliance on an inadequate and overly -simplistic CSA and SCM to establish the
hydrogeologic framework, (B) failure to account for aquifer heterogeneities and pumping
effects of area water -supply wells, and (C) arbitrary assumptions and conjectures for
groundwater flow and contaminant F&T scenarios.
Duke claims that their groundwater flow model and F&T simulations (CAP
Appendix C) were produced by the "University of North Carolina at Charlotte" (UNCC),
although neither author of Appendix C is listed as an employee of UNCC. Appendix C is
a stand-alone modeling report that requires signing and sealing by a NC -licensed
professional engineer (PE) or professional geologist (PG), and labeling it draft' is
inconsistent with the way that Duke relies upon it throughout the CAP. The same
requirement is true of the "UNCC" report in Appendix D, which is not identified as a
draft.' Duke's "third -party peer review" of their 'calibrated [groundwater] flow and
[contaminant] transport model" (CAP page 6) was conducted through the Electric Power
Research Institute (EPRI), an entity that enjoys major financial and other support from
Duke.
B.3.A Duke's computer modeling cannot simulate reality
because it is based on a CSA and SCM that are flawed and
overly simplistic
Our concerns about the flawed CSA forming the basis of the CAP have
been addressed in Section A of this expert report. Duke's modeling report in CAP
Page 24

Appendix C (hereinafter, "model" or "report") explicitly states that the CSA's site
conceptual model is the basis of the model. The report claims that "In general,
groundwater within the shallow, deep transition zone (TZ), and bedrock layers flows
radially from the ash basins and northward toward the Yadkin River." (CAP Appendix C
page 1). Radial flow is not demonstrated anywhere by Duke's model, and northward
flow is required by their model because a "no -flow" boundary encloses three-fourths of
the model domain on the west, south, and east, and "drains" (seeps/springs/discharge
areas) and specified head boundaries comprise the remainder of the model's "domain"
(Appendix C figure 4 provided below).
Figure 4. Numerical Model Boundary Conditions
Page 25

The model's appearance of complexity, and the believability of the
report's numerous and colorful images for F&T simulations, are compromised by relying
on a groundwater -flow model that was constructed from data in a CSA that was used to
corroborate a simplistic, preconceived SCM. Many assumptions inherent in the SCM and
computer flow model are presented as fact, such as Duke's unrealistic claim that
topographically -defined surface water drainage basin boundaries coincide exactly with
multiple groundwater `basins' at depth.
Likewise, the model's description and images do not report the thickness
of primary and secondary model layers that were apparently assigned unique aquifer
properties, nor are the thickness of layers shown graphically with a scale or shown to
have consistent or justified model cell heights (e.g., Appendix C figure 2 shown below).
As one example, Duke admits that "...in the area to be modeled for which borehole data
is not available, dummy boreholes were used to extend the model [layers] to the model
boundaries" and those invented '...boreholes were based on the hydrostratigraphic
thickness of the existing boreholes and the elevation of the existing boreholes based on
the assumption that the hydrostratigraphic layers are a subdued replica of the original
topography of the site and geologic judgment." (CAP Appendix C page 9). There would
be no need to create fictitious boreholes if Duke had installed enough boreholes and
wells to define fundamental aspects of the geology and hydrogeology at and near the
Buck facility.
Page 26

A' A
Figure 2. Model Domain North-South Cross Section (A -A') Through Secondary Ash Basin
13.3.13 Duke's computer model greatly simplifies aquifer
heterogeneities and ignores pumping effects of area
potable water -supply wells
Selection criteria for aquifer parameters used in Duke's computer model
are poorly described, justification of assigned model layer thicknesses and aquifer
parameters are generally absent, and the model ignores any stresses induced on the
natural system by pumping of area water -supply wells. Duke's stated intent of the
model is to "...predict the groundwater flow and constituent transport that will occur as
a result of different possible corrective actions at the site." (CAP Appendix C, page 2).
The limited objective of predicting contaminant F&T under competing scenarios of coal
ash remediation may appear to be a reasonable application of Duke's computer model,
but their model cannot be used to simulate contaminant migration anywhere outside of
the model domain (i.e., most water -supply wells near the Buck property), nor can their
Page 27

model challenge or refine' Duke's many assumptions about area groundwater
hydrogeology and background' concentrations for self -identified COIs.
B.3.0 Duke's F&T scenarios employ poorly -documented
assumptions and conjectures
Simply put, there are numerous assumptions and arbitrary selections of
facts' employed within Duke's F&T simulations. For example, Duke states that
"Sensitivity of the COI transport model was evaluated by varying key model
assumptions forporosity, dispersivity, and/C,...". (emphasis added, CAP, page 63)
These modeling "assumptions" cannot be evaluated independently without access to
copious data and the underlying groundwater -flow model, none of which has been
provided to us for review. Duke's F&T simulations were run for 250 years through the
year 2265, but they neglected to identify the significance of that period or if other
periods were simulated. Nevertheless, Duke concludes that ash -derived groundwater
contaminant concentrations will exceed the 15A NCAC 2L or IMAC standards "...at the
compliance boundary and at the Yadkin River..." (CAP page 64) for numerous COIs for
decades or centuries into the future. It is important to recognize that Duke's modeling
cannot evaluate the potential delivery of ash -derived groundwater contaminants to the
majority of known potable water -supply wells located within 0.5 miles of the Buck
facility, nor was that attempted for those potable wells that are located inside Duke's
model domain.
C. Preliminary Opinions about the BSS Corrective Action Plan (CAP) Part 2
There was inadequate time for GMA to review the 20,000 pages comprising Duke's CAP
Part 2 that first became available to the public on February 22, 2016 through NCDEQ's
website http://edocs.deq.nc.gov/WaterResources/Browse.aspx?startid=221202&dbid=0. We
anticipate that a supplement to our expert report will be necessary to address Part 2 of
the CAP. Nevertheless, there are a few notable aspects of the CAP Part 2 (CAP',
hereinafter) that deserve comment in the context of our expert report to this point.
W -MO -

The CAP includes a 3,384 -page "CSA Addendum" as Appendix A that is predominantly a
mass of raw data or tabulations devoid of explanation, interpretation, and/or integration.
Appendix A is referenced only once in the CAP, and no description of the contents or
relevance can be gleaned without significant effort. For example, Duke "revises" CSA
tabulations of critical aquifer properties and reports them in the CSA Addendum, but
they make no effort to explain why revisions were necessary, or how those changes
might impact any aspect of the SCM and/or groundwater -flow model.
The CAP reports new sampling and analysis of `background' monitoring wells, including
analyses for groundwater obtained from 'dry'well BG-1BR in October and November,
2015 (CAP table 2-5, page 1 of 8). Dissolved and total chromium was detected at 3.7
pg/L and 9.2 pg/L, respectively, in the BG-iBR groundwater sample collected on
October 8, 2015. These analytical results were undoubtedly available to Duke PELor to
submission of CAP Part 1 five week later, and those results should have made Duke
accept that hexavalent chromium does not occur naturally at 78 pg/L at or near the
Buck property. Dissolved chromium and (total) hexavalent chromium was detected at
6.1 pg/L and 6.5 pg/L, respectively, in samples collected from well BG-iBR on November
19, 2015 (CAP table 2-5, page 1 of 8). It is obvious that Duke's persistent claim that the
naturally -occurring 'background' concentration of Cr(VI) is 78 pg/L is false.
Nevertheless, Duke continues to misrepresent and misapply their Cr(VI)'background'
PPBC in CAP Part 2, stating that the concentration of hexavalent chromium '...across the
site outside of the ash basin is 781ug/L." (CAP, Appendix B, page 21). Duke conducted
new fate and transport simulations that apply a 78 pg/L'background' concentration for
Cr(VI) throughout the horizontal and vertical extents of their model domain, stating that
"51nce the background concentration is higher than the NCDHHS HSL, the model
simulations depict concentrations above the standard everywhere within the model
domain." (CAP, Appendix B, page 21). This claim is obviously not consistent with the
analytical data available for Duke's monitoring wells or the off-site water -supply wells
located near the Buck property, none of which are presently known to have reported
Cr(VI) concentrations near 78 pg/L.
Page 29

The two figures provided below show Duke's F&T model output for the "initial (2015)"
distribution of Cr(VI) as they reported in CAP Part 1 (Appendix C figure 124) and CAP
Part 2 (Appendix B figure 71), respectively. The difference in appearance of these two
F&T simulations is striking, and the second image appears to imply that Duke's ash pits
are helping to reduce hexavalent chromium concentrations in the bedrock aquifer
beneath the Buck property.
Figure 124, initial (2015) Hexavalent Chromium Concentrations (Ngrls) in the Bedrock Groundwater Zone
Page 30

LEGEND
P P7 A�+bl+gi � 2
NNcA .'
1 pg�L = nHvcLgrams par
�p•1G
-
LIOr
2. F1a1 aralerd Lh .p
�64-1tA
_ MU 1 L"UL7
F411� i14RH�pY
aa{1aFarr F,'ydylk�nr
9.07 ML
hSM Y0.91M wars
`— a4Jxwar
�� r
'SH ROLM COMVUAW-_G
■EUMCIAMY
-
i9H R424H COMPLAW-E
r
BUJMIR.RY CMW$IMNT
e rA+M buKEE "COY
M93011WY &M J4D MY
PIWN I,ftAL S#O t5kNr4
" MMPuAW-'E BCJMQ4`Rr
LdMFL- VOUMN
MrVe
4414 $A
Ilumoffr
i
7
AC "iri�SlM
A -SH OV
OL6 aP1PS?,R Cie:
CELL CJAID
Figure 14 Initial {2415) Hexavalent Chromfum Concentrations in Bedrock Groundwater Zone
The equivalent bedrock F&T map for total chromium provided in CAP Part 2 (Appendix B
figure 47) is provided below. Duke apparently fails to recognize that Cr(VI) is, by
definition, included in total chromium, as demonstrated by the striking difference
between the figure for Cr(VI) shown above and total chromium shown below. For
example, areas shown in figure 47 (below) that Duke claims contains concentrations of
total chromium between 39 and 68 pg/L (below) display Cr(VI) concentrations between
4 and 8 pg/L (figure 71 above). We do not believe that either of these maps reflects
reality inside or outside the Buck property limits.
Page 31
NNcA .'
1 pg�L = nHvcLgrams par
LIOr
2. F1a1 aralerd Lh .p
_ MU 1 L"UL7
allZ MSL valve -
�
9.07 ML
Feel
3. Hl rAvukrr-LiumaHIRI
� �
s r �•la aLF �
ep 791a,
Figure 14 Initial {2415) Hexavalent Chromfum Concentrations in Bedrock Groundwater Zone
The equivalent bedrock F&T map for total chromium provided in CAP Part 2 (Appendix B
figure 47) is provided below. Duke apparently fails to recognize that Cr(VI) is, by
definition, included in total chromium, as demonstrated by the striking difference
between the figure for Cr(VI) shown above and total chromium shown below. For
example, areas shown in figure 47 (below) that Duke claims contains concentrations of
total chromium between 39 and 68 pg/L (below) display Cr(VI) concentrations between
4 and 8 pg/L (figure 71 above). We do not believe that either of these maps reflects
reality inside or outside the Buck property limits.
Page 31

LEGEND N£
-14 i5Wd1,di-t1
18.15
® 15-18 Ir
®85.99
1-39.68 Zti r __
� 1
DUKE ENERGY
PRop"'o"No4RY
ASH BASIN %WSTE ---
BCLINDARY ti` r r :ASN-Bnsaw ��
$�CONpAR}
ASH BASIN COMPLIANCE -- CELL-
BOUNCARV
ASHFIASINCOMPLIANCEBOU3�
1mrH ouRY ENeRad(:F1T
WITH ARTY ENERGY
PROP'ER4Y BOUNDARY -
PROViSICNALASH BASIN ` ACTIVE !
COMPLIANCE BOUNDARY ASN UASIN
MODEL DOMAIN d f PRIMARY CE,L
i
1
1 ASH �
t STORAGE
iAb
{
f T
NEV­.,'V 4" c N
Notes,
1. pg1L =micrograms i III "'lll
per liter 0 500 1,000
2. Chromium 2L _
Standard= 10 p91L - S �" Feet
3. Chromium PPRC = 1 -
1.9pg1L �� •••3rt'napping.
'iJ►��'�1yfCi .i .a._ rr,rrian,., _
Figure 47 Initial (2015) Chromium Concentrations in Bedrock Groundwater Zone
It is apparent to GMA that removal of ash from the unlined pits and placement of the
ash in an engineered, lined, and water -free repository is the most certain and protective
remedial option. Duke did not evaluate excavation of the ash pits and disposal of that
ash in a fully lined repository within CAP Part 2.
Our preliminary review of Duke's CAP Part 2 has revealed other significant problems that
we anticipate will be addressed within an addendum to this expert report.
Page 32

Appendix A: List of References
We anticipate that additional documents and data will be reviewed as they become
available, so the list below is not considered comprehensive.
Butler, J.R. and Secor, O.T., Jr. 1991. The Central Piedmont, in, Horton, J.W., Jr. and
Zullo, V.A., eds., The Geology of the Carolinas: The University of Tennessee Press,
Knoxville, p. 59-78.
Cunningham, W.L. and Daniel III, C.C. 2001. Investigation of ground -water
availability and quality in Orange County, North Carolina: U. S. Geological
Survey, Water -Resources Investigations Report 00-4286, 59p.
Daniel III, C.C. and Sharpless, N. B. 1983. Ground -water supply potential and
procedures for well -site selection upper Cape Fear basin, Cape Fear basin study,
1981-1983: North Carolina Department of Natural Resources and Community
Development and U.S. Water Resources Council in cooperation with the U.S.
Geological Survey, 73 p.
Driscoll, F.G., 1986, "Groundwater and Wells", Second Edition, Johnson Screens, St.
Paul, Minnesota, 1089 p.
Einarson, M., 2006, Multilevel ground -water monitoring, In: Practical Handbook of
Environmental Site Characterization and Ground -Water Monitoring, ch. 11, ed. D.
Neilson, CRC Press, Boca Raton, Florida, p. 808-848.
Fan, Y.,L. Toran, and R.W. Schlische, 2007, Groundwater flow and groundwater -stream
interaction in fractured and dipping sedimentary rocks: Insights from numerical
models. Water Resources Research, 43, W01409.
Freeze, R.A. and Cherry, J.A. 1979. Ground Water, Englewood Cliffs, NJ, Prentice -Hall,
1979.
Harned, D.A. and Daniel III, C.C. 1992. The transition zone between bedrock and
regolith: Conduit for contamination?, p. 336-348, in Daniel, C.C., III, White, R.K.,
and Stone, P.A., eds., Groundwater in the Piedmont: Proceedings of a Conference
on Ground Water in the Piedmont of the Eastern United States, October 16-18,
1989, Clemson University, 693p.
Hatcher, R.D. Jr., Bream, B.R., and Merschat, A.J. 2007. Tectonic map of the
southern and central Appalachians: A tale of three orogens and a complete
Wilson cycle, in Hatcher,
R.D., Jr, Carlson, M. P., McBride, J. H., and Martinez Catalan, J. R., eds., 4-0 Framework
of Continental Crust: Geological Society of America Memoir 200, p. 595-632.
HDR, 2014x. Buck Steam Station Ash Basin Closure — Conceptual Design Data Report,
Draft — February 2014.
HDR, 2014b. Buck Combined Cycle Station Ash Basin. Proposed Groundwater
Assessment Work Plan, NPDES Permit NC0004774. September 25, 2014
HDR, 2014c. Buck Combined Cycle Ash Basin. Drinking Water Supply Well and Receptor
Survey, NPDES Permit NC0004774. September 30, 2014.
Page 33

HDR, 2014d. Buck Combined Cycle Station Ash Basin. Plan for Identification of New
Discharges, NPDES Permit NC0004774. September 30, 2014.
HDR, 2014e. Buck Combined Cycle Ash Basin. Supplement to Drinking Water Supply
Well and Receptor Survey, NPDES Permit NC0004774. November 6, 2014.
HDR, 2014f. Buck Combined Cycle Station Ash Basin. Proposed Groundwater
Assessment Work Plan (Rev. 1), NPDES Permit NC0004774. December 30, 2014.
HDR, 2014g. Buck Combined Cycle Station Ash Basin. Topographic Map and Discharge
Assessment Plan, NPDES Permit NC0004774. December 30, 2014.
HDR, 2015a. Comprehensive Site Assessment Report. Buck Steam Station Ash Basin.
August 23,2015.
HDR, 2015b. Corrective Action Plan Part 1. Buck Steam Station Ash Basin.
November 20, 2015.
HDR, 2016. Corrective Action Plan Part 2. Buck Steam Station Ash Basin. February
19, 2016.
Heath, R.C., 1983, Basic Ground -Water Hydrology, United States Geological Survey,
Water -Supply Paper 2220, Washington, United States Government Printing Office.
Heath, R.C., 1984, Ground -Water Regions of the United States, United States Geological
Survey, Water -Supply Paper 2242, Washington, United States Government Printing
Office.
Heath, R.C., 1988, Hydrogeologic Setting of Regions in Black, W., J.S. Rosenshein, and
P.R. Seaber, eds., The Geology of North America, Volume 0-2, Hydrogeology:
Geological Society of America, Boulder, Colorado, pages 15-23.
Horton, J. and Zullo, V. 1991. The Geoiogyof the Carolinas. Knoxville: University of
Tennessee Press.
Horton, J. W., Jr., Drake, A A, Jr., and Rankin, D. W. 1989. Tectonostratigraphic
terranes and their Paleozoic boundaries in the central and southern Appalachians,
in, Dallmeyer, R; D., ed., Terranes in the circum -Atlantic Paleozoic orogens:
Geological Society of America Special Paper 230, p. 213-245.
LeGrand, H.E., 1988, Region 21, Piedmont and Blue Ridge in Black, W., J.S. Rosenshein,
and P.R. Seaber, eds., The Geology of North America, Volume 0-2, Hydrogeology:
Geological Society of America, Boulder, Colorado, pages 201-208.
LeGrand, Harry E., 2004. "A Master Conceptual Model for Hydrogeological Site
Characterization in the Piedmont and Mountain Region of North Carolina, A
Guidance Manual," North Carolina Department of Environment and Natural
Resources Division of Water Quality, Groundwater Section.
McDonald, J.P. and R.M. Smith, 2009, Concentration profiles in screened wells under
static and pumped conditions. Ground Water Monitoring and Remediation,
29(2):78-86.
National Research Council of the National Academies, 2006, Managing Coal Combustion
Residues in Mines. Washington, DC: The National Academies Press.
Page 34

North Carolina Department of Environment and Natural Resources. 2007.
"Performance and Analysis of Aquifer Slug Tests and Pumping Tests Policy,"
May 31, 2007.
North Carolina Department of Environment and Natural Resources. 2007.
"Hydrogeologic Investigation and Reporting Policy Memorandum," dated May
31, 2007.
North Carolina Department of Environment and Natural Resources. 2013a. 15A NCAC
2B .0200s. Classifications and Water Quality Standards Applicable to the Surface
Waters and Wetlands of North Carolina. NC and EPA Combined Surface Water
Quality Standards and Criteria Table. May 15. URL:
http://portal.ncdenr.org/web/wg/ps/csu/swstandards
North Carolina Department of Environment and Natural Resources. 2013b. 15A
NCAC 02L. Groundwater Rules. Groundwater Standards Table. April 1. URL:
http://portal.ncdenr.org/web/wg/ps/csu/gwstandards#4
North Carolina Department of Environment and Natural Resources. 2013c. 15A NCAC
02L. Groundwater Rules. Interim Maximum Allowable Concentrations (IMACs)
Table. April 1. URL: http://portal.ncdenr.org/web/wg/ps/csu/gwstandards#4
Secor, D. T, Barker, C., Balinsky, M., and Colquhoun, D. 1998. The Carolina terrane in
northeastern South Carolina: history of an exotic volcanic arc: South Carolina
Geology, v. 40, p. 1-17.
Shapiro, A.M., 2002, Cautions and suggestions for geochemical sampling in fractured
rock. Ground Water Monitoring and Remediation, 22(3):151-164.
Shapiro, A. M., P.A. Hsieh, W.C. Burton, and G. J. Walsh, 2007, Integrated multi-scale
characterization of ground-water flow and chemical transport in fractured crystalline
rock at the Mirror Lake site, New Hampshire. In: Subsurface Hydrology: Data
Integration for Properties and Processes, D.W. Hyndman, F.D. Day-Lewis, and K.
Singha eds., American Geophysical Union Geophysical Monograph Series, v. 171, p.
201-225.
Thornton, S.F. and G. P. Wealthall, 2008, Site characterization for improved assessment
of contaminant fate in fractured aquifers. Water Management 161, WM6:343-356.
Tiedeman, C.R., D.J. Goode, and P.A. Hsieh, 1998, Characterizing a ground water basin
in a New England mountain and valley terrain. Ground Water, 36(4):611-620.
Trainer, F.W., 1988, Plutonic and Metamorphic Rocks in Black, W., J.S. Rosenshein, and
P.R. Seaber, eds., The Geology of North America, Volume 0-2, Hydrogeology:
Geological Society of America, Boulder, Colorado, pages 367-380.
US Geological Survey. 1997. Ground Water Atlas of the United States: Delaware,
Maryland, New Jersey, North Carolina, Pennsylvanis, Virginia, West Virginia.
Piedmont and Blue Ridge Aquifers. URL: http://pubs.usgs.gov/ha/ha730/chl/L-
text4.html
Winter, T.C., J.W. Harvey, O.L. Franke, W.M. Alley, 1999, Ground Water and Surface
Water - A Single Resource, United States Geological Survey, United States
Geological Survey Circular 1139, Denver, Colorado.
Page 35

Appendix B: Resumesfor the Expert Report Authors
Page 36

EXPERIENCE SUMMARY
Dr. Spruill formed Groundwater Management
Associates (GMA) in 1986, and he is DMA's principal
hydrogeologist and president. He is a licensed
professional geologist in North Carolina, he is past
president and a current member of the National Board
for the Association of State Boards of Geology
(ASBOG°), and he is past chairman of the North
Carolina Board for Licensing of Geologists.
Dr. Spruill has been a faculty member in the Department
of Geological Sciences at East Carolina University for
over 35 years, and he is the Founding Director of ECU's
Coastal Water Resources Center. Dr. Spruill has
conducted and supervised numerous hydrogeological
projects through East Carolina University and GMA, and
his accomplishments in the field of hydrogeology are
extensive and well known in the state and region. Many
of his graduate students have taken positions with
environmental and engineering consulting firms, state
and federal government, and oil companies after
completing Master's theses concerning hydrogeological
problems.
Dr. Spruill has served as an editor for Ground Water, a
peer-reviewed journal published by the National Ground
Water Association. Dozens of times each year, he is a
guest hydrogeological speaker for various state and
national organizations. He has evaluated groundwater
resources, designed well fields, and assisted with
development of environmental regulations for State and
local government. Dr. Spruill was a leading advocate
and technical advisor to the stakeholder committee for
new regulations for the protection of groundwater within
the Central Coastal Plain of North Carolina. In 2000, he
received the Distinguished Partnership Award for his
"invaluable service to and special partnership with the
Division of Water Resources and the North Carolina
Department of Environment and Natural Resources".
GROUNDWATER MANAGEMENT ASSOCIATES INC.
THE GROUNDWATER EXPERTS
4300 Sapphire Court, Suite 100
Greenville, North Carolina 27834
Telephone [phone]
www.gma-nc.com
Dr. Spruill has been intimately involved in the permitting and
development of North Carolina's first Aquifer Storage and
Recovery (ASR) project being implemented by Greenville
Utilities Commission in Pitt County. He is currently working on
ASR projects in Hilton Head (SC) and Wilmington (NC). He
has provided hydrogeology litigation support and expert
witness testimony for environmental litigation in North Carolina
and Maryland. In 2011, his testimony was instrumental in
winning a $1.513 verdict for groundwater contamination in
Maryland.
PROFESSIONAL HISTORY
Principal Hydrogeologist & President -1986 to present
Groundwater Management Associates, Inc.
Associate Professor —1979 to present
East Carolina University, Department of Geological Sciences
Director, Coastal Water Resources Center — 2010 to 2012
East Carolina University, Department of Geological Sciences
EDUCATION
Ph.D. Geology, 1980, University of North Carolina -Chapel Hill
M.S. Geology, 1978, East Carolina University
B.S. Geology, 1974, East Carolina University
LICENSES
North Carolina Licensed Professional Geologist # 942
Page 37

GMA ®►
EXPERIENCE SUMMARY
Dr. Campbell joined Groundwater Management Associates,
Inc. (GMA) in 1994 after working at the Florida Geological
Survey and teaching field geology, physical geology,
environmental geology, oceanography, and earth science at
two universities and a community college. Dr. Campbell is
GMA's Director of Environmental Services and he manages
the firm's Greenville office. He is responsible for ensuring the
timeliness and quality of GMA's environmental projects,
providing hydrogeology support and project management for
environmental and water -supply projects, and supervising
hydrogeology staff members, staff and project geologists. His
clients include government agencies, industry, municipalities,
developers, and commercial interests.
Dr. Campbell is GMA's project manager for the first Aquifer
Storage and Recovery (ASR) system permitted and
constructed in North Carolina. He provides expert witness
services in support of environmental litigation involving
petroleum and other contaminants in Maryland and North
Carolina. Between 2001 and 2008, Dr. Campbell was the
lead hydrogeologist for coastal plain and piedmont sites
administered by the Federal Trust Fund (FTF) program of
NCDENR. He managed the assessment and remediation of
over 100 FTF sites with petroleum -contaminated soil and
groundwater. He has assessed soil and groundwater
contaminated with chlorinated solvents and metals.
Dr. Campbell is an Adjunct Teaching Assistant Professor in
the Department of Geological Sciences at East Carolina
University, where he has taught field geology, physical
geology, and advanced igneous petrology. Since 1999, he has
taught field geology at the North Carolina Geology Field
Course held each summer in New Mexico and Colorado. Dr.
Campbell has taught at 22 geology field courses since 1982.
Dr. Campbell is a member of several professional
organizations, including the Geological Society of America,
the National Ground Water Association, the American
Geophysical Union, the Colorado Scientific Society, and the
New Mexico Geological Society (lifetime member).
GROUNDWATER MANAGEMENT ASSOCIATES INC.
THE GROUNDWATER EXPERTS
4300 Sapphire Court, Suite 100
Greenville, North Carolina 27834
Telephone [phone]
www.gmc-nc.com
PROFESSIONAL EMPLOYMENT
Senior Hydrogeologist-1994 to present
Groundwater Management Associates, Inc.
Adjunct Teaching Assistant Professor— 2013 to present
East Carolina University, Department of Geological Sciences
Visiting Assistant Professor— 1999 to 2013
East Carolina University, Department of Geological Sciences
Geologist -1991 to 1994
Florida Geological Survey, Florida DEP
Visiting Assistant Professor— 1990
East Carolina University, Department of Geological Sciences
Instructor -1986 to 1993
Florida State University, Department of Geology
Adjunct Instructor —1989 to 1994
Tallahassee Community College, Division of Mathematics
and Science
EDUCATION
Ph.D. Geology, 1994, Florida State University
M.S. Geology, 1985, East Carolina University
B.S. Geology, 1981, East Carolina University
LICENSES
North Carolina Licensed Professional Geologist # 1414
Virginia Certified Professional Geologist # 2801 001800
North Carolina Certified Well Contractor # 2815-A
TRAINING AND CERTIFICATIONS
OSHA HAZWOPER Certified (40 hours)
OSHA HAZWOPER Annual Refreshers (8 hours)
MSHA Part 48 Surface Miner Training (24 hours)
Maryland Drinking Water Sampler Certification # 8023SC
WO i