NC0022406_60 BOD comments response-signed_20170720

DUKE
ENERGY,
July 20 2017
Mr. S. Jay Zimmerman, Director
Division of Wawr Rrwsoarc(-s
North Carolina Department of E.n,,ironmental wuality
1611 Mail Service Center
Raleigh, North Carolina 2 t 5iig- 15 i i
Subject. 60% Basis of Design Report Comments
Bulews Creek Steam Station
Dear Mr. Zimmerman -
sc6 south Criurcri street:
Wil C. de EC13K
GtlariW% Moroi Carcina enzu2
yErU-3 ! 3-3 (1 y
In a letter dated June 30, 2017, DEu documented comments from their review of the 60% Basis
of Design (BOD) package for the Bulaws Crek Interim, Action Plan. Duke Energy with
assistance from our cunsultant, has prepared the enclosed response to clarify our interpretation
ul the ivmmcnt3 and herr vve intend to addru33 unch cummcnt.
if you Have comment,,- anti/or questions, please direct them to me at 980-373-3719.
Sincerely,
Z).
D. Edwin M. Sullivan, PE
Director — Waste & Gruondwdter Prgrams
Environmental, Health & Safery
Enclosure: Interim Action Plan Schedule Updates - Belews Creek Steam Station
Cc: Shet ry Knight — WSRO Regional Office Sopervisur
Shuying Wang — WSRO Regional Office Hydmgualuyista
Striven M. Lantur — DW R Hydmgaolugist
Jim Wells — Dake Energy
M�lunia Martin — Duke Energy
Kevin Klrkley — Doke Energy
Encluaed- Ra5pon5a to Comma, 45

Response to Comments in S. Jay Zimmerman Letter to Ed Sullivan, dated June 30, 2017
60 Percent Basis of Design Report
Belews Creek Steam Station
Basis of Design 60% Submittal Comments
1) Based on the Settlement Agreement signed between DEQ and Duke Energy signed on
September 29, 2015 and test results to date, the primary objective of the interim action
conducted at the facility should be the reduction of groundwater migration of source area
constituents from the ash basin. In Section 1 .1 .4 Scope and Objectives of the 60% Basis of
Design Report, the primary objective is further clarified as being "to reduce groundwater
migration of source area constituents from the ash basin towards the 2.23 -acre parcel and remove
constituent mass from the area of highest constituent concentrations."
The information presented in the 60% BOD Report indicates the saprolite and transition zone
along the extraction well network will be dewatered, causing a gradient reversal, and
provide hydraulic control of source area constituents. Reduction of constituent mass is
unlikely due to changes in flow direction and the likely intermittent operation of the
extraction wells; however, the objective of boundary control should be achieved. Reduction
of constituent mass in groundwater should not be considered as a primary metric for overall
system effectiveness. Instead, reduction of plume distribution, especially proximal to the
extraction well network, along with confirmation of boundary control, should be the focus
for monitoring system performance. These concepts concerning the primary objective of
extraction system operation and the anticipated resulting site conditions should be reflected
in the Final BOD Report.
Response: We agree with the summary provided in this comment and that the primary
function of the interim action is the reduction of plume distribution and boundary
control. While reduction of constituent mass may occur in the target area, this is not
the primary function of this plan. The final Basis of Design (BOD) Report will be
revised to indicate this clarification of the system's function.
2) Acknowledging the limitations of testing to date, the Division encourages additional
characterization and testing following Phase I implementation to evaluate concerns related to
potential preferential pathways in the study area, variable flow directions that may develop
because of pumping, and confirmation of boundary control. In summary, the Division
believes additional work is likely needed to facilitate optimization of system operations as
mentioned in the 60% BOD Report. Related information is expected in detail as part of the
Final BOD Report.
Response: After system start-up, monitoring of water quality and water levels will continue.
As these data are processed and analyzed, it may become necessary to conduct a
focused investigation to address potential questions. These will be reported to the
Department along with a suggested approach for investigation and resolution. It is
not possible to describe this process in any more detail in the Final BOD report, but
deviations from the plan will be dealt with promptly.
Page 11

Response to Comments in S. Jay Zimmerman Letter to Ed Sullivan, dated June 30, 2017
60 Percent Basis of Design Report
Belews Creek Steam Station
3) In -Situ Chemical Immobilization was evaluated as an effective approach in Appendix B but
was not mentioned in Section 3.2. Please expand on why this remedial alternative would not
be effective or should not be considered for the area up -gradient of Parcel A in the Final BOD
Report.
Response: The geologic, lithologic and hydrogeologic factors that preclude several of the
remedial options discussed in Section 3.2 also preclude the use of in-situ chemical
immobilization. The final BOD Report will be revised to include this technology in
the discussion in Section 3.2.
4) The report indicated that the initial extraction system will be monitored for a six-month
period after initial startup to determine system efficiency. Please provide additional details
regarding the effectiveness monitoring program (e.g. monitoring wells, monitoring
frequency, parameters, etc.) in a table in the Final BOD Report.
Response: These details will be provided in a table in the Final BOD report.
5) Please provide well specifications such as depth, diameter, etc., for the two proposed
observation wells and other wells if determined to be needed in the Final BOD Report.
Response: These details will be provided in the Final BOD report.
6) Regarding Appendix C, on page 4-1, the second to last paragraph states, "The lower limit of the
model domain coincides with an assumed maximum depth of water -yielding fractures in bedrock. This
was estimated to be approximately 710 feet below the base of the transition zone." What is the basis
for this specific depth selected as the base of water yielding fractures in the model domain?
Response: A conservative thickness of 710 feet was used, as provided in the site wide model
constructed by HDR, based on the premise that the geology in the Piedmont is
complex and the fracture density decreases with depth. According to the HDR CAP -
Part 11, "The bedrock hydrostratigraphic layers were extended vertically in the
refined model to correlate to the deepest reported off-site private water supply well,
as obtained from the completed questionnaires received from adjacent well owners
during the receptor survey performed in 2015 and/or the NC State/Stokes County
geographic information system(GIS) database".
7) Table 1-1 should list all the constituents of interest (COIs) detected in groundwater within the
study area, not just boron, chloride, and selenium. In Section 2.2.2 Summary of HDR Recent
Groundwater Quality Results, exceedances of 02LStandards provided are confined to the
upper two zones, not including reporting exceedances present in the bedrock. Elevated
concentrations of antimony, arsenic, chromium, and vanadium were detected in
downgradient bedrock well GWA-2713R. These constituents were also detected above
02LStandards in bedrock well GWA-20 BR located between the ash basin and Parcel A.
However, there are no monitoring wells in bedrock within Parcel A, which does not mean
02L exceedances are not present in bedrock. This should be considered as part of the
modeling effort.
Response: Table 1-1 and the text will be modified to reflect these comments.
8) Only boron, chloride, and selenium are modeled for the interim action while there are several
other COIs that impact groundwater in the study area, including antimony, arsenic,
Page 12

Response to Comments in S. Jay Zimmerman Letter to Ed Sullivan, dated June 30, 2017
60 Percent Basis of Design Report
Belews Creek Steam Station
beryllium, chromium, hexavalent chromium, cobalt, thallium, and vanadium. These latter
COIs should also be simulated to provide predictions of site conditions resulting from this
interim action.
Response: Constituents were selected based on (1) ash pore water concentrations being
significantly greater than background concentrations, and (2) a sufficient number to
downgradient wells having concentrations greater than a 2L standard so that a
plume is discernable. Site specific constituents selected for the modeling process
were boron, arsenic, beryllium, chloride, chromium, cobalt, hexavalent chromium,
selenium, and thallium for the 2016 transport simulations. Boron was chosen as a
main tracer for the ash basins for three reasons: 1) boron is present in coal ash, 2)
there is typically a low background concentration, and 3) boron is a mobile and
dominant constituent.
Antimony, arsenic, and chloride are not detected in enough wells above 2L to
discern a plume. Hexavalent chromium and vanadium have elevated background
concentrations making it difficult to discern a plume. However, beryllium, cobalt,
and thallium had a sufficient number of wells with results greater than the 2L
standard within the Area of Concern (AOC) and will be incorporated into the
groundwater fate and transport model in the Final BOD report.
Appendix D Updated Geochemical Modeling Report Comments
1. General Comments
a. The focus of the review comments provided here address Attachment A Geochemical
Model Development Focused Geochemical Modeling Report. These comments are
intended to provide a basis for revised Focused Area Model development and reporting
for Parcel A
Response: We appreciate the reviewer's careful consideration and helpful comments for
improving this report. The comments above indicate the need to focus on qualitative
discussions of the geochemical behavior of the constituents of interest with a
primary focus on using site-specific data. We have included site-specific analyses in
our modeling approach as well as a "global" modeling approach considering a range
of geochemical conditions to evaluate how variations in geochemical conditions
could result in the mobilization of constituents of interest. We agree with the
reviewers comment that the discussion should be primarily qualitative.
Geochemical modeling, particularly including sorption reactions, requires several
assumptions and has several limitations that limit our ability to develop a
quantitative model. Empirical distribution coefficients cannot be used in our model
because such coefficients do not consider changes in geochemical parameters (pH,
Ex, ionic strength) or changes in chemical speciation of the constituents. Therefore,
we have employed a surface complexation based modeling approach that
approximates ion sorption reactions using a thermodynamic construct similar to that
used for aqueous speciation modeling (e.g., Davis et al., [1]). While surface
Page 13

Response to Comments in S. Jay Zimmerman Letter to Ed Sullivan, dated June 30, 2017
60 Percent Basis of Design Report
Belews Creek Steam Station
complexation modeling approaches do consider changing geochemical conditions
and chemical speciation, there are specific limitations to consider including:
The thermochemical sorption constants (i.e. surface complexation constants) are
only available for a relatively small group of minerals. In this work, we have
primarily used the self -consistent reactions describing ion sorption to ferrihydrite
and gibbsite compiled by Dzombak and Morel [2] and Karamalidis and
Dzombak [3]. Including ion sorption to other minerals in the model cannot be
done unless a similar set of self -consistent surface complexation constants are
available for site-specific minerals and generally this is not the case. Thus, the
models are not representing site-specific mineralogy and are using ferrihydrite
and gibbsite as representative sorbing surfaces.
2. Incorporation of surface complexation based sorption models requires that a
modeler make assumptions regarding the concentration of reactive sites on the
mineral surface where sorption of ions can occur. We have followed the same
site density assumptions used by Dzombak and Morel [2] and Karamalidis and
Dzombak [3] to constrain the surface site concentrations in our modeling
approach and used measurements of extractable iron and aluminum
concentrations from well -bore samples to constrain the model. However, this
approach follows an inherent assumption that the extractable iron and aluminum
represents available mineral surfaces.
The use of gibbsite and ferrihydrite as representative minerals due to the
availability of surface complexation reactions for these minerals necessitates that
all coal -ash disposal sites are modeled using these same minerals. Therefore, the
only differences between the sorption behavior at each site will be due to 1)
differences in the pH, Ex, and ion concentrations at each site, and 2) differences
in the extractable iron and aluminum concentrations from site specific solids.
However, our "global" analysis of 7 coal -ash disposal sites indicated a relatively
similar distribution in the pH, EH, and extractable iron and aluminum
concentrations despite the fact that the seven sites have different geologic
characteristics. Therefore, we have taken a global modeling approach to provide
a qualitative understanding of how changes in pH, EH, and ion concentrations
can influence the mobility of constituents of interest. Then site-specific transects
are modeled using site-specific groundwater data and the model output is
evaluated with consideration of the overall behavior of each constituent.
In the Final BOD report, we will add a discussion similar to what is presented
above to address comment 1b and also address comment 1c by adding additional
details to our discussion of the site-specific transect model with regards to
constituent mobility that may result from implementation of the interim action.
Page 14

Response to Comments in S. Jay Zimmerman Letter to Ed Sullivan, dated June 30, 2017
60 Percent Basis of Design Report
Belews Creek Steam Station
b. Project purpose, scope, and objectives of the geochemical modeling effort to support the
Basis of Design Report should be clearly defined in the introduction of the revised
Focused Geochemical Model Parcel A. The discussion in subsequent sections, including
the Summary section, should reflect how the geochemical model provides information
for the evaluation of remediation options that address task objectives.
Response: Response to this comment included in the above discussion regarding comment 1a.
c. Geochemical modeling conducted to evaluate remedial alternatives should focus on
providing qualitative assessment of geochemical conditions and constituent mobility that
may result from implementation of the interim action using site-specific data rather than
making projections from data from other sites, particularly those data obtained from sites
with different geologic and geochemical settings.
Response: Response to this comment included in the above discussion regarding comment 1a.
2. Specific Comments concerning Attachment A Geochemical Model Development Focused
Geochemical Modeling Report
a. Page A-1. "Equilibrium constants for aqueous speciation reactions were taken from the
USGS WATEQ4F database. This database contained the reactions for most elements of
interest except for Co, Sb, V, and Cr. Constants for aqueous reactions and mineral
formation for these elements were taken from the MINTEQ v4 database which is also
issued with PHREEQC." Why not just use the MINTEQ v4 database that contains
reactions for all the elements of interest?
Response: The WATEQ4F database is the foundation of the standard database deployed with
PHREEQC (phreegc.dat) and therefore was selected for use as the primary database.
Both WATQ4F and MINTEQ v4 databases use a substantial portion of the NIST v46
"Smith and Martell" database[4] as well as additional reactions from Ball and
Nordstrom [5]. Therefore, these databases are relatively consistent. We have also
noted some issues with the MINTEQ v4 database incorporation into the PHREEQC
format (i.e. the MINTEQv4.dat database that is deployed with PHREEQC). For
example, the reaction: Fe3+ + 3H2O q 3H+ + Fe(OH)3(aq) has a log K value of -15 in
the actual visual MINTEQ program. However, the value is listed as -12.56 in the
MINTEQv4.dat file deployed with PHREEQC. The later value is consistent with the
phreeqc.dat database. To confirm this, we will run and report selected reaction
systems with the MINTEQv4 database to demonstrate the outputs are comparable
for the Final BOD report.
b. Page A-1. "(Adsorption) Constants for Co, V, Cr, and Sb were added to the modified database as
well as all constants involving ion sorption to HAO." What were the sources of these
constants for HFO and HAO? The MINTEQ v4 database has sorption constants for HFO,
but not HAO.
Response: These constants were taken from the compiled databases described by Dzombak
and Morel [2] and Karamalidis and Dzombak [3].
c. Page A-3. "Therefore, attempting to delineate between mineral surfaces, let alone strong and
weak sites on such surfaces, would add unnecessary uncertainty andfitting parameters to the
Page 15

Response to Comments in S. Jay Zimmerman Letter to Ed Sullivan, dated June 30, 2017
60 Percent Basis of Design Report
Belews Creek Steam Station
models. Therefore, sorption to only one site on both HFO and HAO is considered." What type
of site was considered and why?
Response: We used a one -site 2-pKa model as described in Table A-1 of Appendix A. The
surface complexation constants for ion interactions with weak sites from Dzombak
and Morel [2] and Karamalidis and Dzombak [3] were used. We will include
additional text describing this modeling approach in the revised report.
d. Page A-4. "This approach uses the total Fe and Al solid phase concentrations in specific wells to
identify how much sorbent is available. "
i. What is meant by "total Fe and Al solid phase concentrations"? Was XRF analysis used
to measure Fe and Al in the solid phase? Was an extraction technique used to
dissolve Fe and Al from the solidphase? If so, what was the extraction method?
Response: The total Al and Fe was determined by digestion and elemental determination of
many samples from the site using EPA method
ii. What is the rationale for assuming that all of the solid Fe and Al is present as
ferrihydrite and gibbsite? This is unlikely to ever be the case in a natural system.
Response: We completely agree that the Fe and Al are likely not present as ferrihydrite and
gibbsite. However, the necessary surface complexation constants that are required
to complete this type of sorption modeling are not available for all of the ions and
minerals of interest at this site. The compilations of reactions and constants
described for ferrihydrate and gibbsite are two of the few self -consistent
thermochemical sorption database examples which are currently available.
Therefore to utilize these available databases, we had to assume that ferrihydrite
and gibbsite are present in these systems and the Fe and Al values were used to
calculate how much was available. While this is a large assumption, we reiterate
the intent of this modeling exercise to be a qualitative assessment of the
geochemical behavior of several constituents of interest and evaluate potential
changes in constituent mobility that may result from implementation of the interim
action. So these two mineral surfaces are essentially representative surfaces to
account for sorption sites during this modeling effort.
e. Page A-4. "The utility of the EQUILIBRIUM—PHASES command, in this case is that it can
provide the total number of sorption sites to produce HFO and HAO, and it is linked to the
stability of the gibbsite and ferrihydrate solid phases which are used to approximate the sorption
sites." The spelling of the iron adsorbent ferrihydrite is misspelled as ferrihydrate in this
sentence and all subsequent uses of the word.
Response: Thank you for identifying this mistake. We will make this correction in the revised
report.
f. Page A-4. "The utility of the EQUILUIBRIUM_PHASES command, in this case, is that it can
provide the total number of sorption sites to produce HFO and HAO, and it is linked to the
stability of the gibbsite and ferrihydrate solid phases which are used to approximate the sorption
sites. So in effect, the model is calculating the aqueous concentrations of FWD, Fe(III), and AKIII)
as well as moles of gibbsite and errih drate by combining the total aqueous Fe and Al
Page 16

Response to Comments in S. Jay Zimmerman Letter to Ed Sullivan, dated June 30, 2017
60 Percent Basis of Design Report
Belews Creek Steam Station
concentrations from well measurements with the total Fe and Al available as solid phases
measured from wells with solid phase data available (listed in Table A-2)." What does it mean
that the model is calculating these values? Aren't they being input by the user? A
PHREEQC input file for this calculation should be provided to clarify how adsorption is
being modeled.
Response: The EQUILIBRIUM_PHASES command in PHREEQC is allowing ferrihydrite and
gibbsite to form if the input aqueous concentrations of Fe(III) and Al(III) are above
saturation levels. Then the ferrihydrite and gibbsite that are formed become the
sorbing surfaces. This essentially gives the model a method of predicting how much
sorbent will be available without additional user input. The input Al(III) and Fe(III)
levels include that measured in the aqueous phase from well samples as well as the
solid phase concentrations measured by digestion and elemental analysis. We use
this approach to compare with simulations where we estimate the concentration of
sorbent based on the solid phase concentration and using the site density standard
assumptions described by Dzombak and Morel [2] and Karamalidis and Dzombak
[3].
g. Page A-6. For partially weathered rock of the transition zone, the fracture's pore volume
available for adsorption surface reactions is decreased by lowering the porosity of the
material ... The resulting solid phase concentration of 32 kg/L accounts for the high density of
transition zone material." This method likely vastly overestimates the concentration of
adsorption sites associated with the fractures. Only the solid phased present along a
fracture is available for adsorption, not the entire mass of solid associated with each liter
of the water. Defensible assumptions should be developed concerning modeling
eochemical conditions in the transition zone and bedrock.
Response: We appreciate this comment and had multiple discussions and iterations of this
approach internally. There is no standard way to determine the available surface
area and sorption site density for a fractured solid. Therefore, we assumed a
relatively small volume due to the low porosity that effectively raised the solid phase
concentration used in PHREEQC. While this does appear high, we feel it is
representative of the apparent surface site concentration a small volume of pore
water would be exposed to in a fracture. However, we also report the model output
in terms of distribution coefficients (Ka values). Therefore, the solid phase
concentration used is normalized back to a linear Ka value to allow for a direct
comparison to a model of sediment or even a batch sorption experiment (such as
what was used in the "global" model). We will add some additional explanatory text
to the revised report to clarify our assumptions.
The remediation effort is focused on the shallow and transition zone aquifers,
therefore geochernical conditions were not modeled in bedrock.
h. Page A-9. Groundwater Data Used in PHREEQC Model. Why isn't inorganic carbon
included as a major ion? Bicarbonate is a common major ion in shallow groundwater.
Page 17

Response to Comments in S. Jay Zimmerman Letter to Ed Sullivan, dated June 30, 2017
60 Percent Basis of Design Report
Belews Creek Steam Station
Response: We have some modeling runs which included bicarbonate and it does not
significantly influence the speciation of our constituents of interest. However, it has
caused some issues with model stability due to charge imbalance. Therefore, we
elected early on to omit bicarbonate. In the Final BOD report, we will provide
examples where inclusion of bicarbonate in the model does not influence our model
output.
i. Page A-9. "A second modeling effort was undertaken to examine general behavior of several
constituents of interest under changing geochemical conditions. This modeling effort is referred to
as a global model because it considers the range of ion concentrations, pH, Eh, and extractable Fe
and Al concentrations from seven coal ash disposal sites in NC. The results of this model remain
applicable to Belews Creek because nearby sites, DEP Roxboro and DEP Mayo, used in the Global
Model have similar onsite geology and geochemistry as Belews Creek." The global model is
setup in Section A.3 and parameters are discussed in Section A.4, which includes data
from Weatherspoon, Sutton, and H.F. Lee facilities located in the coastal plain in addition
to data from the Cape Fear facility that is in a Triassic Basin. Data from these four sites
do not represent those site conditions representative of the geology and geochemistry of
the Piedmont crystalline bedrock and do not enhance model results. Geochemical model
development at this point should be focused on site-specific data.
Response: As discussed above related to the general comments, our global model using all of
the sites is meant to provide a qualitative description of the behavior of several
constituents of interest and identify the extremes in pH and Ex where mobilization
or retardation of the constituents is expected to occur. We chose to include this
global model in this report to illustrate the potential mobility of selected constituents
across a range of conditions. The discussion of the global model is meant to
supplement the more relevant model using site-specific data along a transect of
interest to this remediation effort.
j. Page A-11. This is the approach used in the global model and the assumed solid phase
concentration was 50 g-solid/L- volume." What is the source of this value? Why aren't
the calculated values of 8 kg-solid/L-solution (unconsolidated material) and 32kg-solid/L-
solution (transition zone fractured material) on page A-6 used in this equation?
Response: The global model used a much lower solid to liquid ratio because it was originally
intended to compare the output with a series of laboratory batch sorption
experiments using solids from the seven sites included in the global model. Those
laboratory experiments used a 50 g/L solid phase concentration so we used that same
value for the global model. However, we also report the model output in terms of
distribution coefficients (Ka values). Therefore, the solid phase concentration used is
normalized back to a linear Ka value to allow for a direct comparison to a model of
sediment or even a batch sorption experiment (such as what was used in the
"global" model.
k. Page A-12. Note the large differences (>10x) in Al and Fe concentrations in Table A-6
compared to Table A-2. Why are total concentrations used for Model 1 Belews Creek
Page 18

Response to Comments in S. Jay Zimmerman Letter to Ed Sullivan, dated June 30, 2017
60 Percent Basis of Design Report
Belews Creek Steam Station
adsorption site concentrations, while extractable concentrations were used in the global
model? Comparing Table A-3 and A-6, the Al and Fe site concentrations for Model 1 are
on the order of 100,000 times greater than those for the global model. How were the
"extractable" concentrations of Fe and Al measured?
Response: The extractable concentrations of Fe and Al in Table A-6 were measured using a
standard soil analysis approach described by Chao et al., [6]. These values were
determined during part of an initial laboratory based study that included the batch
sorption work discussed above regarding the global model. Since then, total
digestion analysis of solids from site-specific wells have been obtained. Those values
are listed in Tables A-2 and A-3. As expected, the values from the total digestion are
considerably higher. For the site-specific model, our interest was in comparing the
model output to specific well samples along a transect and we do not have
extractable iron or aluminum concentrations for those samples. Therefore, we have
elected to work with total iron and aluminum concentrations. However, this
provides only a IN difference. The reason for the 100,000x different site
concentrations are from the assumed 50 g/L solid phase concentration in the global
model versus the 8 kg/L or 32 kg/L assumed for the site-specific model. However, we
emphasize again that returning the model output to a Kd value effectively
normalizes the selection of the solid phase concentration and makes the output
comparable.
1. Page A-18. "The concentrations of major ions (e.g., Ca2+, Na+, Fe (II/III), Cl-, SO42-) were
varied to consider the range of potential values. The concentrations of several trace ions and
constituents of interest were not varied so that the model could examine the potential for
competition for sorption sites between the varying major ion concentration conditions and a fixed
condition for the trace elements (Table A-8)".
i. Trace ion adsorption may be more competitive for the COI than major ion
adsorption.
Response: We completely agree with the reviewers comment. Thankfully, the thermochemical
construct of the speciation and surface complexation modeling approach we have
adopted for this effort accounts for such ion competition. Therefore, by varying one
parameter, we can see the influence of the varied molar ratios of the COIs and
background major ion concentrations.
ii. Also, why not use the actual trace concentrations for Belews Creek, instead of an
average for seven other sites?
Response: This model was part of our "global" modeling was done to identify the extremes in
pH and EH where mobilization or retardation of the constituents is expected to occur.
We compare the sorption extent and behavior across this wider global range to more
site-specific values when modeling transects at Belews Creek.
m. Page A-22. "To gain an understanding of the aqueous chemical species of each constituent of
interest, Pourbaix diagrams were generated using Geochemist Workbench (GWB) version
10.v10." Instead of using generic groundwater data for missing lots, why not use
Page 19

Response to Comments in S. Jay Zimmerman Letter to Ed Sullivan, dated June 30, 2017
60 Percent Basis of Design Report
Belews Creek Steam Station
Belews Creek data?
Response: These Pourbaix diagrams were part of our "global" modeling effort done to identify
the extremes in pH and Ex where changes in constituent speciation may occur. Since
many of the constituents do not form complexes with background ions, the changes
in speciation will be minimal and the effects of changes in pH and Ex are more
important. Therefore, we felt it was appropriate to use the average global data. For
the Final BOD report we will select or average site-specific data to calibrate new
Pourbaix diagrams to represent site-specific geochemical conditions.
Page 110