NC0004961_Assessment Report/2012 Supplemental GW Monitoring Report_20140915

-a
Pat McCrory
Governor
NCDENR
North Carolina Department of Environment and Natural Resources
Division of Water Resources
Water Quality Programs
Thomas A. Reeder John E. Skvarla, III
Director Secretary
Mr Sean DeNeale
Duke Energy Carolinas, LLC
P O Box 1006
Mall Code EC13K
Charlotte, NC 28201-1006
September 8, 2014
RECEIVEDIDENRIDWR
SEP 15 2014
Water Quality
Pemtitting Section
RE Duke Energy Carolinas, LLC
Riverbend Steam Station Ash Basin Groundwater Assessment Report and
2012 Supplemental Groundwater Monitoring Report
NPDES Permit NC0004961
Gaston County
Dear Mr. DeNeale,
Thank you for your submittal of the Groundwater Assessment report for the Riverbend Steam
Station Ash Basin. We received the electronic copy on May 31, 2013, and the follow up hard
copy on June 3, 2013. We received the 2012 Supplemental Groundwater Monitoring Report in
electronic form on June 21, 2013, and the follow up hard copy on June 24, 2013
The groundwater reports were submitted with the goal of addressing the report outlined in
paragraph 52 of the Proposed Consent Order filed on October 8, 2013, requiring determination
of the naturally occurring concentration of substances in the site's groundwater The submitted
groundwater reports indicated progress toward that goal but did not fulfill that obligation. As
you know, Secretary John Skvarla of the North Carolina Department of Environment and
Natural Resources (NCDENR) sent a "Notice of Regulatory Requirements" (NORR) to Duke
Energy on August 13, 2014, which requires that Duke Energy submit a workplan to the Division
of Water Resources by September 26, 2014 that documents proposed site assessment
activities at 14 Coal Ash facilities (including the Riverbend site) that will be performed to
complete a comprehensive site assessment (CSA) for each facility Although the previously
submitted groundwater reports were not submitted with the intent of addressing the
requirements of a CSA pursuant to 15A NCAC 2L .0106(g), we are providing you the following
comments on your previous reports These comments should not be interpreted as limiting the
scope of the CSA workplan required in Mr Skvarla's August 13, 2014 NORR It is our hope that
these comments will provide helpful feedback on the previous site evaluation activities you have
performed and that you can give them consideration as you move forward to prepare the CSA
workplan.
Division of Water Resources / Mooresville Regional Office
610 East Center Avenue, Suite 301, Mooresville, North Carolina 28115
Phone, 704-663-16991 Fax [phone]
Internet www ncwaterguality orq
An Equal Opportunity 1 AfFirmabve Action Employer

Duke Energy Carolinas, LLC — Riverbend GW Assessment
September 8, 2014
Page 2 of 5
The comments are organized following the each report's format starting with the May 2013
groundwater assessment report
4 1 Site Conceptual Model
• Better definition of groundwater flow across the entire site is necessary due to potential
hydraulic mounding from the ash basins This is acknowledged for water levels in select
wells but water level data indicate this may also affect water levels in wells at clusters 7 and
8. Hydraulic mounding would affect groundwater conditions near the assumed background
well cluster 7 and upgradient well cluster 8, which in tum could affect the evaluations of
concentration data in background and upgradient wells against compliance wells A slight
upward hydraulic gradient in well cluster 7 is somewhat unusual for wells in a topographic
high area based on the site conceptual model The well cluster 7 may in fact be on a
groundwater divide and reflect background conditions for the site, but the potential effects of
mounding 'on groundwater flow to well clusters 7 and 8 need to be assessed. This could be
accomplished by the installation of additional drivepoint piezometers or monitoring wells
between the ash basins and the well clusters 7 and 8 or by other means Water level data
between the ash basins and well clusters 7 and 8 would help clarify groundwater flow
conditions in this area. It would be appropriate to create cross sections that extend through
the cells to evaluate groundwater conditions as well as generating water table /
potentiometric maps that provide coverage across the entire site. It would also be
appropriate to evaluate groundwater levels over time using potentiometric maps to address
the effects of mounding on groundwater flow, particularly its potential effects on well clusters
7 and 8 Any groundwater pumping at or near the site also needs to be verified
Several additional monitoring wells are also needed to assess the complete vertical and
horizontal extent of the exceedances detected in existing site monitoring wells as well as
develop a more complete site conceptual model with a thorough understanding of the
subsurface geology, hydrogeology, and interactions between the surficial aquifer, transition
zone and fractured bedrock It is essential that groundwater exceedances be completely
delineated and a thorough understanding of the source areas, geology, hydrogeology, and
geochemistry of the site be completed before proposing arguments as to whether some
component of the detected groundwater exceedances may be attributable to natural
background conditions.
4 3 Ash Basin Water Quality
• We recommend collection of samples from the primary cell in addition to the secondary cell
to facilitate comparison to data from well clusters located west of the primary cell and to help
evaluate against background conditions.
• Regarding comparison of ionic concentration data of secondary cell (TOWER) with
monitoring wells, include data on ion balance to indicate the accuracy of the chemical data.
Although the ionic composition of background and upgradient wells are different than the
TOWER sample, hydraulic mounding may affect the chemical signature of these wells
through ion exchange reactions resulting in a different type of water than the downgradient
wells. Assessment of additional water level data as mentioned above in 41 above and
chemical composition of the primary cell would help clarify the spatial variation in ionic
composition of groundwater at the site

-1 S
` Duke Energy Carolinas, LLC — Riverbend GW Assessment
September 8, 2014
f Page 3 of 5
a
4 4 Review Mountain Island Lake Water Level Data
• Water level data in MW -13 show strong correlation to changes in lake elevation. We
recommend similar monitoring and evaluation of water levels in select other wells such as
MW -9, MW -10, MW -11, MWA4, or MWA5 to provide better understanding of groundwater
and surface water interaction around the perimeter of the site An assessment of the effects
of dilution and redox conditions on groundwater due to Interaction with surface water would
help clarify if notably elevated concentrations are restricted to areas downgradient of the
cells or if concentrations at other areas are less affected by groundwater -surface
interactions
4 5 Review Location -Number of Background Monitoring Wells
• Water levels in clusters 7 and 8 are below those of the primary cell and may be affected by
hydraulic mounding associated with the ponds and, in the case of cluster 8, the adjacent ash
storage area Cross sections and maps mentioned in 4 1 above will help evaluate cluster 7
and the previous use of it as a background location Duke will need to install additional
background well(s), possibly located to the east in similar geology, outside the hydraulic
influence of the ponds and other site ash handling activities Similarly, the evaluation of
cluster 8 as being upgradient may need to be revisited.
4.7 Evaluate Exceedances Against Background Well Results
• As per comments associated with 4 1 and 4 5, a complete geologic and hydrogeologic
assessment of the site which includes delineation of exceedances will be needed to verify
that the background wells are not unduly influenced hydraulically by the ponds or nearby
ash handling activities.
4.8 Evaluate Exceedances Against Turbidity Values
Continue to record turbidity values during sampling and keep track of the final turbidity
values pnor to collecting samples. Elevated turbidity values noted in MW -81, MW -8D, and
MW -9 should be monitored closely.
4.11 Collect Reduction/Oxidation Field Parameters
• Continue to obtain DO and ORP measurements during sampling as these and other field
parameters compliment analysis of the situation For instance, downgradient wells contain
lower DO and ORP and higher conductivity compared to MW -7, which may indicate the
effects of leachate from the ash ponds.
4.11A Collect Soil Samples
• We note that iron and manganese concentrations in soils are largely comparable to limited
available soil data for Lincoln and Gaston counties, but the highest values, in the vicinity of
MW -13 are higher than normally expected for the region The workplan for a CSA should
include evaluating soil and groundwater data and incorporating it with groundwater
modeling
5 Summary and Conclusions
• The summary should include results from statistical analyses — the wells and the parameters
that show statistically significant increases Interpretations on whether a constituent is

Duke Energy Carolinas, LLC — Riverbend GW Assessment
September 8, 2014
Page 4 of 5
naturally occurring or not should be based on a complete CSA with all data taken together
In summary, the groundwater assessment at Riverbend was previously Intended to provide
an understanding of groundwater flow conditions and groundwater quality for establishing
naturally -occurring background concentrations of target chemical constituents at the site and
for an evaluation of exceedances of tarcet constituents In compliance monitoring wells
The CSA which is now required for the site will also Involve completing Investigations to
thoroughly understand the geology, hydrogeology, and full delineation of the groundwater
exceedances.
Appendix C Statistical Analysis
• In general, the procedure seems appropriate, but the narrative Is not easy to follow
• Much of the analysis hinges on the validity of cluster 7 being background As noted above,
there are concerns about this conclusion that must be addressed.
• There is no discussion of outlier data and how they are handled For any statistical
analyses used in completion of the CSA, please provide rationale for keeping or removing
outlier data
• Please provide a very brief description of each statistical test used along with the purpose of
each test, the hypotheses tested, any Lnderlying assumptions and how these are met or not
met, when the test is used, any advantages/disadvantages of its use and a reference. Refer
to Chapter 8 of EPA guidance previously provided
• Section 1, Parametric and non -parametric, for the >50% NDs show or describe the
Wllcoxian Rank Sum test as an alternative to the Non -Parametric Prediction Limit test along
with selection criteria for when each test is used There is no reference to the Levene's test
for Equal Variance once the Shapiro tests for normalcy fail How was the Levene's Test
applied? Use of the Wilcoxian Rank Sum test when variances are not equal should have
been clarified.
• A discussion should have been provided of the use of Sen's Slope Test to determine trends
and how applied if both background and compliance wells exhibit trends
e Section 2 Inter -Well Prediction Limits, a description of possibility, as shown in flow chart, of
use of Kruskal Wallis method if variances prove to be equal should have been provided
• Rationale should be provided for using Prediction Limit test when data are trending
according to Sen's Slope test
• Section 4 Statistics Summary, clarify what happens when data are trending up or down
• Table; Chemstat Initial Statistics, state clearly what is being tested and fudged to be true or
false. Check the column heading for 99% LCL Should this be 99% Confidence Interval?

4 .
e
Duke Energy Carolinas, LLC — Rwerbend GW Assessment
September 8, 2014
!' Page 5 of 5
• It would be good to provide stacked concentration vs time graphs stacked well by well
showing the concentration of a parameter over time in each well Ideally, group these for
each parameter.
• Provide a statement that makes clear that all constituents in all compliance wells are tested
• Clarify the frequency of formal statistical testing into the future (i a after sampling events,
annually, etc) Discuss any re -testing procedures and how false -positives are handled
• Describe how the well system design and sampling, hydrological,"geochemical and
analytical factors have been ruled out as potential sources of error in the statistics (See
4 3 1-4 3.6 in EPA unified guidance)
2012 Supplemental Groundwater Monitoring Report
The models developed for three cross-sections across target monitoring wells MW -9, MW -10,
and MW -13 provide an initial understanding of groundwater flow and contaminant transport at
these selected locations. The models broadly characterize the groundwater flow conditions at
these cross-sections and roughly predict the contaminant concentrations at the projected
compliance boundary However, the models lacked sufficient details to capture the complex
groundwater flow normally found in a regolith -bedrock aquifer system. Asa part of
understanding groundwater flow and contaminant transport at the site, the proposed workplan
and subsequent CSA will need to include broader groundwater modeling efforts that include
predicting groundwater flow within and between the various aquifer zones (including bedrock
fractures), predicted future constituent distributions in aquifer zones across the site and beyond
site property boundaries, and contributions and loadings to surface waters
Should you have questions regarding these comments, feel free to contact me at 704-235-2180
Sincerely,
W, �
Andrew H Pitner, P.G
DWR Assistant Regional Supervisor
CC. Jay Zimmerman, NC DENR DWR Central Office, Raleigh
Debra Watts, NC DENR DWR Central Office, Raleigh
Sergei Chernikov — NC DENR DWR Water Permitting Section, Raleigh
Bill Miller, HDR Engineering, 440 S. Church St., Suite 1000, Charlotte, NC 28202