Approach to Managing COIs for CAPs

ROY COOPER.
Governor
MICHAEL S. REGAN
Secretary
LINDA CULPEPPER
Director
Paul Draovitch
Senior Vice President
Environmental, Health & Safety
Duke Energy
526 South Church Street
Mail Code EC3XP
Charlotte, North Carolina 28202
NORTH CAROLINA
Environmental Quality
October 24, 2019
Subject: Approach to Managing Constituents of Interests for Purposes of Corrective Action
Plans
Dear Mr. Draovitch:
On September 4, 2019 Duke Energy presented a Constituent of Interest (COI) Management Plan (Plan) to
facilitate Corrective Action Plan (CAP) development required at its coal combustion residuals (CCR)
facilities. In lieu of a document to review, the North Carolina Department of Environmental Quality (DEQ)
has reviewed the content of the presentation to develop a position regarding the subject matter. The COI
Plan as presented to date is conditionally acceptable with notable revisions described in Attachment 1.
These revisions will assist Department review and ensure a consistent approach across the CCR facilities.
DEQ staff welcome the opportunity to discuss related COI Plan issues with Duke Energy, including
attending other facility -specific presentations. If you have questions or concerns regarding the
Department's position relative to the COI Plan provided in this correspondence, please contact Steve Lanter
in the DWR Central Office at [phone] and he will coordinate with the respective Regional Offices
to initiate discussion.
Sincerely,
Jtl�son, Deputy Director
Division of Water Resources
Attachments
Attachment 1 - Approach to Managing Constituents of Interests for Purposes of Corrective Action Plans
cc: WQROS Regional Offices (electronic copy)
GWRS Central File Copy
North Carolina Department of Environmental Quality 1 Division of Water Resources
�
D_E � 512 North Salisbury Street 1 1636 Mail Service Center I Raleigh, North Carolina 27699-1636
�0-ft 919.707.9000

October 22, 2019 Attachment 1- Approach to Managing COls for Purposes of CAPS
Attachment 1
Approach to Managing Constituents of Interests for Purposes of Corrective Action Plans
On September 4, 2019, Duke Energy presented a Constituent of Interest (COI) Management Plan (Plan)
for Corrective Action Plan (CAP) development required at its coal combustion residuals (CCR) sites.
The presented Plan is conditionally acceptable with notable revisions described below. These revisions
Will assist Department review and ensure a consistent approach across coal ash facilities.
Framework as Presented by Duke on 9/04/19. As described in the 9/04/19 Duke presentation, COls
that occur above the criterion (defined as the greater of 15A NCAC 02L standards [02L], interim
maximum allowable concentrations [IMACs], or background threshold values [BTUs]) at/beyond the
compliance boundary (CB) will be identified for corrective action. Depending on their
observed/modeled occurrence and distribution and a "groundwater (GW) exceedance ratio", Duke
Energy proposes to map some COIs; other CON will, as proposed, be unmapped and only listed in a
table. The typical mobility of each COI will be described along with conditions that affect its mobility.
Attenuation mechanisms will be described for each COI along with the expected long-term stability of
those mechanisms.
Framework Response by DWR with Revisions. This COI Management Framework is a process
developed to facilitate corrective action planning. The Framework helps identify the areas and COls in
need of corrective action and the potential remedies that could be effective. Corrective actions are
being implemented in conjunction with and to support and augment basin closures. When CBs are
modified or expire, and compliance has not been achieved in an area no longer covered by a CB,
corrective actions will be required in those areas. Corrective actions may or may not need to be
"active" depending on factors evaluated in the Framework such as, for example, mobility of the CON in
question, stability of attenuation mechanism(s), remediation goals for the COls, etc. Rather than
computing a maximum mean and a GW exceedance ratio, use of a lower confidence limit (LCL)95 (95%
lower confidence limit)' is a more appropriate metric to identify and document areas in need of
corrective action. For each monitoring well, Duke Energy shall compute an LCL95 for the COI in
1 See, for example, United States Environmental Protection Agency (EPA) Unified Guidance (March 2009) and ProUCL
Technical guidance (2013), including discussion of parametric and non -parametric 95% LCLs. Note that if the well
sample dataset is shown to be trending for a given COI, an LCL95 may be computed on the trendline.
Page 1 of 4

October 22, 2019 Attachment 1- Approach to Managing COls for Purposes of CAPs
question by using data from all sample events at the wel12. If the computed LCL95 exceeds the
applicable criterion then that well -COI would be identified in the CAP as a localized area in need of
corrective action. If the localized area for the identified COI is isolated and does not represent a larger
scale plume, it may be mapped accordingly in the CAP by simply showing a large scale plan view map
with the well that contains the exceeding LCL95, along with the LCL95 value, representative pH and Eh
values, Kd, ratio of species concentrations for the C01 in question (if applicable and assuming
speciations have been measured/computed), and a dashed line containing the area in need of
corrective action. As described in the 9/4/19 presentation, a table(s) will also be provided containing
the list of CON, and their corresponding typical speciation (anion, cation, neutral), mobility under
acidic/alkaline conditions, mobility under reducing/oxidizing conditions, localized conditions that affect
mobility, propensity for sorption, ion exchange, and (or) precipitation, and expected long term stability
of the attenuation mechanisms. Influence of hyporheic zone on geochemical conditions and COI
mobility, if applicable, should also be considered/discussed, as should the influence of potential
surface water mixing on geochemical conditions and mobility during storm -induced rises in surface
water levels that can, in some cases, reverse groundwater gradients. The extent of boron above
background is to be shown on all maps as an approximate extent of hydraulic influence from the basin.
If the modeled boron plume has not yet stabilized (is continuing to move in time) then the extent of
boron above background at future year 2120 should also be shown on the maps to indicate the
predicted future extent of basin influence.
Transects referred to in the CAP shall be shown (a) in a plan view map along with the observed head
contours and corresponding flow lines, (b) in a plan view map along with modeled head contours, and
(c) in cross section with modeled head contours and velocity vectors.
Dissolved Groundwater Concentrations. Unfiltered (total) concentrations of constituents are
measured for most groundwater samples. However, for geochemical modeling purposes, dissolved
concentrations must be used in the input file of the computer code. For each CCR basin, a conceptual
2 Rather than using only data from 2018 to 2019 as presented by Duke, data from all sample events should be used. If
a technical reason exists to omit a portion of the historic dataset, an appendix may be provided that includes the well,
all values in the historic record for the COI in question, the values that should be omitted, and rationale for the
omission (e.g. early break-in issues, COI concentration -time trends, pH or turbidity issues, etc.). Future monitor wells
would also undergo LCL95 computation to identify additional areas in need of corrective action.
Page 2 of 4

October 22, 2019 Attachment 1- Approach to Managing Cols for Purposes of CAPS
geochemical model will be developed to represent current conditions and estimate how COI
concentrations may change in the future in response to changes in environmental conditions, such as
redox change due to decanting/dewatering. The results of ion speciation and mineral equilibrium
calculations from groundwater data along flowpaths from the source areas to downgradient locations
will be used to develop the geochemical conceptual site models. Dissolved concentration data for all
parameters (major/minor ions and COls) must be collected from the monitoring wells along the
flowpaths to develop these models. This will also be done for areas where anomalous geochemical
conditions occur such as the low pH area at Allen. In most cases, dissolved sampling conducted under
the Interim Monitoring Plans will be sufficient for modeling purposes.
Valence State Measurements. Several of the potential CON are redox-sensitive and occur in more
than one valence state [e.g., As (III,V), Se (-II, 0, IV, VI), Fe {II,III), Mn (II, III, IV)]. Because of the
perceived difficulty of preserving samples in the field for redox species measurement in the laboratory,
redox speciation is being calculated from the measured pH and Eh using a geochemical modeling code.
This method assumes redox equilibrium and may not always be appropriate. In situations where
anomalous groundwater concentrations of a redox-sensitive COI are present, it would be beneficial to
conduct sampling and laboratory analyses for the redox species of the COI to determine if speciation is
a factor leading to the anomalous behavior. This would require appropriate preservation of water
samples in the field for lab measurements of the specific redox species.
Additional sampling and analysis of redox species in selected wells would help to demonstrate that the
modeled speciations that have been calculated under an assumption of equilibrium conditions are
appropriately determined. The number and location of wells used for this purpose should be
appropriate to demonstrate confidence in the modeling approach, input data, and results.
COI Identification. The Plan process discussed in the meeting included a comparison of groundwater
concentrations to relevant regulatory criteria in order to select Cols based on exceedances of their
respective criteria. Consideration should also be given for those constituents that do not currently
exceed their criteria but may feasibly exceed that criteria in the future if environmental conditions
change. For example, if the arsenic criterion is 10 µg/L and the measured groundwater concentration
is 5 µg/L, then arsenic would be included in predictive geochemical modeling to determine if corrective
Page 3 of 4

October 22, 2019 Attachment 1- Approach to Managing COls for Purposes of CAPs
actions produce conditions that elevate the arsenic concentration above its criterion. For practical
purposes, constituents that are currently measured in a groundwater well beneath or downgradient of
a basin at an LCL95 concentration at or above 50% of the criterion (i.e. LCL95 >= COI criterion x 0.5)
would be included in the modeling of future conditions to estimate whether or not those future
conditions increase the groundwater level to a concentration greater than the criterion.
Conclusions and Discussion in CAPS. Findings and conclusions presented in the CAPS should pertain
to a specifically identified local area beneath and (or) downgradient of a basin. Each area identified for
corrective action, whether it be a plume, an isolated, localized area, or an anomalous area, should be
discussed individually and specifically. For consistency in the CAP, discussions and tables related to COI
management generally should refer to the LCL95 (rather than the mean or geomean) and the COI
criterion (rather than 2L, IMAC, or background). Where the CAP discusses performance or
effectiveness monitoring that will be conducted as part of corrective action implementation, an upper
confidence limit 95% (UCL95) would be the appropriate evaluation metric rather than the LCL95 (i.e.
corrective action continues until the UCL95 is below the cleanup criterion').
s See EPA (2018) Groundwater Statistics Tool — User's Guide.
Page 4 of 4