NC0005088_8. CSS CAP Part 2_Appx E_FINAL_20160212

This page intentionally left blank

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Executive Summary
The goal of geochemical modeling efforts for the Cliffside Steam Station (CSS) site is to
describe the expected partitioning of constituents of interest (COls) between aqueous and solid
phases (i.e., between groundwater and soil and porewater and ash) and metals species. In
addition, geochemical modeling is performed to describe anticipated changes in phase
distributions given variations in dissolved oxygen (DO), which affects the oxidation state of
groundwater); pH; Oxidation -Reduction Potential (ORP, expressed as eH or Eh); and Total
Dissolved Solids (TDS), which affects ionic strength and ion competition at adsorption sites.
Site -specific COI evaluations were performed on a well -by -well basis using the United States
Geological Survey (USGS) PHREEQC (v3.3.3) geochemical speciation code (Parkhurst and
Appelo 2013) and PhreePlot (Kinniburgh and Cooper 2011), a companion plotting package that
utilizes looping PHREEQC with a hunt and track approach to determine stability boundaries. By
using a single -well approach, wells can be evaluated or grouped later based on geochemical
characteristics. The single -well approach also allows fine resolution of geochemical constituents
and subtle differences between wells that have a significant bearing on the overall geochemical
characterization. COls evaluated for the CSS site were antimony, arsenic, barium, beryllium,
boron, chromium, cobalt, iron, lead, manganese, nickel, pH, selenium, sulfate, TDS, thallium,
and vanadium. Calculations were developed based on measured concentrations of COls and
other analytes such as ORP, alkalinity, sodium, and other ions in groundwater for each of the
159 wells monitored at the CSS site.
PHREEQC calculations were performed to develop Pourbaix (Eh -pH) diagrams. These
diagrams display the dominant geochemical forms (i.e., species) that would be expected in
groundwater in the absence of adsorption under equilibrium conditions and allowing for most
probable mineral precipitation, where appropriate. Measured ORP and pH values for each well
were plotted on a Pourbaix diagram for each COI to evaluate the likely distribution of species at
the CSS site. Additional PHREEQC calculations were performed to simulate anticipated
geochemical speciation that would occur for each COI in the presence of adsorption to soils.
Further simulations were performed to evaluate model and COI response to changes in DO, pH,
and TDS in the presence of sediment adsorption. Adsorption to soils was represented using a
surface complexation theory approach with hydrous ferric oxides (HFO) and hydrous aluminum
oxides (HAO) representing weak and strong binding sites, respectively. Values for HFO and
HAO were determined from extractions from actual site sediment. These extractions were also
the basis for measured distribution coefficients (Kd values) for CSS soils determined from
adsorption experiments conducted by the University of North Carolina Charlotte (UNCC).
To geochemically simulate changes to aquifers or test potential remediation strategies,
simulations were utilized in which DO, pH, reduction -oxidation (redox), and TDS were varied.
These geochemical simulations are termed titrations in this report. Each set of titrations
provides an estimate of the percentage of each COI that would be adsorbed as a function of
changing DO, pH, redox, or TDS along with relevant changes to the dominant species across

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
the gradient. For these titrations, TDS was evaluated as the addition of a select set of cations
and anions known to be common in soils and sediment at the CSS site to include sodium,
calcium, chloride, potassium, and sulfate. Changes to DO, pH, and TDS were utilized for
titrations due to the affinity for numerous COls such as metals to exist primarily as anionic or
cationic species and their adsorption coefficient variations to mineral surfaces, soils, sediment,
rock and ash. The titration method in geochemical modeling can also account for mobility
changes due to redox threshold changes and potential mineral precipitation, indicated by
saturation indices in outputs. Adsorption of anionic species is typically greater at lower pH, in
which anions are more strongly attracted to positively charged surfaces (and vice versa
regarding cationic species). Similarly, the solubility of mineral phases is pH dependent and
lower pH values tend to favor formation of more soluble cationic species for most alkali
elements, alkali earth elements, and transition metals. Titration simulation results for the CSS
site suggest that COI speciation and adsorption are more strongly influenced by changes in DO
(which influences Eh) and pH than by changes in TDS, indicating potential factors to consider
during remediation strategy discussions.
Probability plots were completed using site data based on a 3 x 3 matrix. This matrix included
the upgradient, source, and downgradient wells along with a comparison of shallow, deep, and
bedrock wells. The plots were evaluated to determine if a pattern in the data is present.
Additional information obtained directly from the PHREEQC model included the Saturation
Index. This information was used to determine the potential species of metals that may
precipitate within the aquifer.
A study of the potential sorption capacity of the soils for the COls was completed as a precursor
to Monitored Natural Attenuation (MNA) Tier III site evaluations of acceptable groundwater
concentrations. The sorption potential of soils was evaluated by using PHREEQC to numerically
titrate a mixture of COls to evaluate how aqueous and sorbed phase concentrations vary in
response to COI additions. In these titrations, the ratio of Cols added was calculated as the
geometric mean of COI concentrations measured in source wells for the CSS site. Cols in the
titration mixture were antimony, arsenic, boron, cobalt, chromium, iron, manganese, and
vanadium. This mixture was incrementally added to the mixture of COls present in groundwater
of each downgradient well at the CSS site. In general, soil sorptive capacity for COls such as
boron is typically small and even a small addition of boron to groundwater is expected to result
in increased aqueous concentrations of boron. In contrast, soil sorptive capacity for COls such
as arsenic or chromium is much higher and a relatively larger amount of arsenic or chromium
could be added to soils in downgradient areas of the CSS site without resulting in aqueous
concentrations that exceed groundwater standards. These findings indicate substantial
differences between the potential adsorption of various COls to sediment across site conditions.
1. Introduction
A geochemical modeling effort was undertaken to describe the speciation of constituents of
interest (COls) and other groundwater constituents across the spectrum of groundwater
conditions measured at the Cliffside Steam Station (CSS) site. The goal of this effort is to
describe the expected partitioning of COls between aqueous and solid phases (i.e., between
groundwater and soil and pore water and ash) and anticipated changes in phase distributions
given variations in dissolved oxygen (DO), pH, and Total Dissolved Solids (TDS). Changes in

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
DO affect the oxidation state of groundwater as measured by Oxidation -Reduction Potential
(ORP), which is generally expressed as Eh, eH, or electron activity (pE). Changes in pH change
the acidity of groundwater, and thus affect Eh. Changes in TDS affect ionic strength and ion
competition at adsorption sites.
COls evaluated for the CSS site were antimony (Sb), arsenic (As), barium (Ba), beryllium (Be),
boron (B), chromium (Cr), cobalt (Co), iron (Fe), lead (Pb), manganese (Mn), nickel (Ni), pH,
selenium, sulfate (S042-), TDS, thallium (TI), and vanadium (V). Site -specific evaluations were
performed on a well -by -well basis using the United States Geological Survey (USGS)
PHREEQC geochemical speciation model (Parkhurst and Appelo 2013) and PhreePlot
(Kinniburgh and Cooper 2011), a companion plotting package based on PHREEQC.
Calculations were developed based on measured COI concentrations (including pH), ORP,
alkalinity, sodium, chloride, and other ions in groundwater for each of the 159 wells monitored at
the CSS site. Adsorption (surface complexation) data for thallium do not exist in the databases
used for PHREEQC simulations; adsorption cannot be simulated without data to parameterize
reactions with mineral phases. For COls without adsorption data, laboratory studies would be
needed to evaluate their interactions with hydrous ferric oxides (HFO) and hydrous aluminum
oxides (HAO).
PHREEQC simulations were used to:
1. Develop Pourbaix (Eh -pH) diagrams to display the most likely geochemical forms (i.e.,
species) that would be expected in groundwater in the absence of adsorption to soil.
Measured ORP (as Eh) and pH values for each well were plotted on a Pourbaix diagram
for each COI to evaluate the likely distribution of species on a site -specific basis (Section
2).
2. Evaluate geochemical speciation for groundwater condition in conjunction with surface
complexation to HFO and HAO, which are expected to be representative of soils, rock,
and ash (Section 3).
3. Simulate anticipated geochemical speciation that would occur for each COI in the
presence of adsorption to soils and in response to changes in DO, pH, and TDS (Section
4).
Simulations in which DO, pH, and TDS were varied are termed titrations. Each set of titrations
provides an estimate of the percentage of each COI that would be adsorbed as a function of
changing DO, pH, or TDS. For these titrations, TDS was evaluated as the addition of cations
and anions known to be present in soils and groundwater at the CSS site: sodium, calcium,
chloride potassium, and sulfates. DO, pH, and TDS were chosen for titration because metal
COls exist primarily as anionic or cationic species and their adsorption to mineral surfaces can
change with variations in pH and interactions with minerals present in soils, rock, and ash. In
simulations beyond development of Pourbaix diagrams, adsorption was represented as surface
complexation to HFO and HAO (e.g., the mineral gibbsite) and was determined from extractions
that are the basis for distribution coefficients (Kd values) for the CSS site soils determined from
adsorption experiments conducted by the University of North Carolina Charlotte (UNCC).
However, these experiments to determine Kd values, are not easily applicable where significant
site heterogeneity exists across ranges of pH, Eh, and DO values, as compared to utilizing
surface complexation methods of adsorption (Bethke and Brady 2000).

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Groundwater data used for analysis includes samples collected for the initial investigation
completed in June 2015, and several subsequent rounds of sample collection through
September 2015. Additional results are provided for background (BG) wells sampled in
November 2015. These data sets were separated into a matrix for calculation of average
concentrations of COls. Data were separated by several matrices including location and depth.
The location matrices include the upgradient, source or downgradient location, and the depth
matrices as shallow, deep, or bedrock. In the event there was an issue with a particular data
point, the model was developed to ignore the anomalous data point (e.g., dissolved oxygen = 0
or >10.0 mg/L, which is supersaturated). The data set was used to calculate the minimum,
maximum, and median of the data to use for the PHREEQC modeling as discussed further in
Section 3.
2. Pourbaix Diagrams
To gain an understanding of the aqueous chemical species of each COI, Pourbaix diagrams
were generated using PhreePlot (Kinniburgh and Cooper 2011). PhreePlot uses PHREEQC
(Parkhurst and Appelo 2013) to produce graphics of select PHREEQC outputs more readily
than can be generated by other means. To perform these simulations, the WATEQ4F database
was utilized since it provides a consistent set of thermodynamic constants to represent surface
complexation as described below. Constants for As, Be, Co, Sb, TI, V, and Cr are not present in
the WATEQ4F database. The values for these constituents were retrieved from the MINTEQ
database and verified for consistency. Both databases are standard components of PHREEQC.
This augmented database of thermodynamic constants was used for all geochemical speciation
simulations for consistency.
The Pourbaix diagrams display the geochemical composition (i.e., predominant phase
distribution of species) that would be expected in an aqueous solution in the absence of surface
complexation to sediment or soil. Dominant species are shown in terms ranges of their
existence across ranges of Eh and pH. Diagrams for each metal COI were generated based on
a 500 parts per billion (ppb) (pg/L) concentration of each COI (Figures 1 through 15; Table 1). A
background electrolyte of 0.5 milligrams per liter (mg/L) of sodium and calcium were added in
conjunction with a balance of chloride for speciation and base ionic strength, based on most
common measured groundwater electrolytes. These concentrations are generally higher than
concentrations measured in the CSS site groundwater samples. If precipitation (solid phases) is
not anticipated for Eh -pH regions of interest, it is not expected to occur for less saturated, lower
concentration conditions measured at the CSS site. On these diagrams, predominant aqueous
species are shown with blue shading and predominant precipitated solid phases (if any) are
shown with brown shading. Precipitation or solid phases are designated with an (s), the
amorphous solid/liquid phase is designated as an (a), and the aqueous form of the metal is
provided as an anion (-) or cation (+). The symbols on the Pourbaix figures indicate an
upgradient, source, or downgradient well and shallow, deep, or bedrock well depths. The
symbols are provided to assist with a visual description of precipitate formation based on depth
or other factors.

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Table 1 Concentrations of COls used to generate Pourbaix diagrams
Species
Concentration
ppm, m /L
Molar Conversion
/mol
Concentration
mol/L
Antimony
0.5
121.76
4.11 E-06
Arsenic
0.5
74.92
6.67E-06
Barium
0.5
137.33
3.64E-06
Beryllium
0.5
9.01
5.55E-05
Boron
0.5
10.81
4.63E-05
Chromium
0.5
52
9.62E-06
Cobalt
0.5
58.93
8.48E-06
Iron
0.5
55.84
8.95E-06
Lead
0.5
207.2
2.41 E-06
Manganese
0.5
54.94
9.10E-06
Nickel
0.5
58.69
8.52E-06
Selenium
0.5
78.97
6.33E-06
Sulfate
0.5
32.06
1.56E-05
Thallium
0.5
204.38
2.45E-06
Vanadium
0.5
121.76
4.11 E-06
Notes:
1. ppm = parts per million
2. mg/L = milligrams per liter
3. g/mol = grams per mole
4. mol/L = moles per liter
Horizontal lines on Pourbaix diagrams represent transitions that depend on reduction -oxidation
(redox) conditions, but are independent of pH. Vertical lines represent transitions that depend
on pH, but are independent of redox conditions (Eh or eH). Sloping lines represent transitions
that depend both on pH and Eh. Although Pourbaix diagrams are useful to identify the
predominant (i.e., most abundant) species in solution, some limitations exist:
Only equilibrium conditions for the predominant species are shown. Information
regarding the kinetics (rates) at which reactions to form a given species occurs is not
provided. Similarly, equilibrium conditions and predominant species are presented for a
standard temperature 250 Celsius. Formation of some mineral phases and their
predominance may differ at other temperatures and under non -equilibrium conditions.
• Only reactions between a COI and base electrolytes in water are considered. Other
aqueous or mineral species may occur in groundwater under in situ conditions and
contribute to the reactivity of a secondary constituent. Secondary minerals will vary from
well to well and are not considered. A discussion for each COI may be included in the
next section if secondary mineral precipitation is indicated. The presence of other
species or phases (e.g., surface complexation) may substantially alter the in situ
distribution of a COI.

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
1.0
0.5
w
> 0.21 atm
❑a
Sb03 _
�a
❑
aCP ° 00Cl
m 0 ° ° 00 _
CHa(g) > 1
❑ Shallow
0 Deep
© Bedrock
o Upgradient
a Source
® Downgradient
—0-5 1 1 1 1 l -1- 1
2 4 B H 10 12
pH
Figure 1 Pourbaix diagram for antimony with measured Eh and pH from Cliffside Steam
Station wells
1.
0.
v
LU
n
0
0 >0a.21 atm
H3 C4
N#AS04
A A
HAGO,_2-
i❑ I]
° Da °
A
®Q H3As 0 013 0 0 HP Sol,
CH4(g) > 1
As a
❑ Shallow
0 Deep
Q Bedrock
O Upgradient
O Source
© Downgradient
2 4 6 S 10 12
pH
Figure 2 Pourbaix diagram for arsenic with measured Eh and pH from Cliffside Steam
Station wells
6

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
1.0
0.5
W
M
a 0.21 atm
AA Bat+
6
O O O
0 00 0
a
�I 0 ° ° E3 PdCi4)34aq_
�>l
❑ Shallow
0 Deep
0 Bedrock
© Upgradient
0 Source
O Downgradient
—0.5 1 1 1 1 1 1
2 4 5 H 10 12
pH
Figure 3 Pourbaix diagram for barium with measured Eh and pH from Cliffside Steam
Station wells
1.0
0.5
v
W
Ur
❑A
Bee-
AA
0.21 atm
■ �,r, ■ •a ■■ ■ • • •
CH4(g) >
❑ Shallow
0 Deep
0 Bedrock
0 Upgradient
0 Source
} 0 Downgradient
—0.5 1 1 1 1 1 1
2 4 5 S 10 12
pH
Figure 4 Pourbaix diagram for beryllium with measured Eh and pH from Cliffside Steam
Station wells

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
1.0
02(g) > 0.21 atm
0.5
s
W
M
❑ Shallow
0 Deep
Q Bedrock
© Upgradient
• Source
O Downgradient
—0.5 1 1 1 1 1 1
2 4 6 H 10 12
pH
Figure 5 Pourbaix diagram for boron with measured Eh and pH from Cliffside Steam
Station wells
1.0
0.5
v
LU
Ur
05
HCrO4- Oafg) > U-21 atm
❑+, CrOd2-
Cr3�
OH)2 NaC'C
�❑ 00
❑ 0
�9—
CIg} > 1 Ja
❑ Shallow
0 Deep
O Bedrock
d Upgradient
O Source
© Oowrgradient
2 4 6 8 10 12
pH
Figure 6 Pourbaix diagram for chromium with measured Eh and pH from Cliffside Steam
Station wells
8

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
1.0
0.5
W
M
a 0.21 atm
❑AL �.
❑ Co2+ C0304(S)
❑
'' ❑ o
64- ❑ ©0 ❑
❑ O ❑13 d
CH4(g) s 1
C00O3(S)
❑ Shallow
0 Deep
© Bedrock
o Upgradient
Q Source
4 Downgradient
—0-5 1 1 1 1 1 1
2 4 5 8 10 12
pH
Figure 7 Pourbaix diagram for cobalt with measured Eh and pH from Cliffside Steam
Station wells
1.0
F
FFOHr7�
09(q) > 0.21 atm
0.5
v
W
Xr
■■_: ■ ■ ■
■■ E.. ■
CH4(g) > 1
❑ Shallow
0 Deep
Q Bedrock
O Upgradient
O Source
© Downgradient
—0 51 1 1 1 1 V. 1
LV2 4 6 S 10 12
pH
Figure 8 Pourbaix diagram for iron with measured Eh and pH from Cliffside Steam
Station wells
9

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
1.0
0.5
WONT -I
02(9) > 0.21 am
❑e
Pb2+
�e
46 %4-- C3 E3 C3
❑ Pb(C[D3g2`
O ❑ ❑ OP OH), S1
CH4(g) � 1
❑ Shallow
0 Deep
O Bedrock
© Upgradient
D Source
a Downgradient
2 4 B H 10 12
pH
Figure 9 Pourbaix diagram for lead with measured Eh and pH from Cliffside Steam
Station wells
1.0
0.5
MCI
❑ Shallow
I] Deep
O Bedrock
© Upgradient
• Source
O Downgradient
-0.5 1 1 1 1 1 --. 1
2 4 6 8 10 12
pH
Figure 10 Pourbaix diagram for manganese with measured Eh and pH from Cliffside
Steam Station wells
10

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
1.0
P2(g) a 0.21 aim
ee
0.5
.—. Ni2+
s e
W 4e► 1] M
0.0 W
° ° ° °
CH4(g) � 1
❑ Shallow
I] Deep
Q Bedrock
© Upgradient
• Source
o Dawngradient
—0.5 1 1 1 1 1 1 1
2 4 5 H 10 12
pH
Figure 11 Pourbaix diagram for nickel with measured Eh and pH from Cliffside Steam
Station wells
1.0
LU
0.
MEN
0
02(g) a 0.21 aim
H Se 3
e
e Se042-
HSeG3
❑
© SEiT320
O O
CH4(g) 1
❑ Shallow
I] Deep
O Bedrock
© Upgradient
• Source
O Downgradient
— s�2 4 6 S 10 12
pH
Figure 12 Pourbaix diagram for selenium with measured Eh and pH from Cliffside Steam
Station wells
11

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
1.0
0.5
W
2(g) a 0.21 atm
A ❑ S042_
4A% O 13 O
d WO
00 U
� O ©O 0 Oa
CH4(g) >
❑ Shallow
E3 Deep
O Bedrock
O Upgradiient
O Source
O Downgradient
-0.50. i 1 i 1 i 'L� i
V 2 4 B S 10 12
pH
Figure 13 Pourbaix diagram for sulfate (as sulfur) with measured Eh and pH from
Cliffside Steam Station wells
1.01 TI I
W
0.5
Ur
Op(g) a 0.21 atm
❑❑
I
A
❑ TI+
4�%6 %4- D E3 O
o OQ 13
6
O D 0 ® O
CHA) > 1
❑ Shallow
E3 Deep
O Bedrock
O Upgradient
O Source
© Downgradient
-0.51 1 i 1 i i
2 4 6 S 10 12
pH
Figure 14 Pourbaix diagram for thallium with measured Eh and pH from Cliffside Steam
Station wells
12

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
W
1.0
0.5
—0.5
V02+
09(9) > 0-21 atm
H3V2{ 7
6e
H2VOa
Vol+ ❑
❑ HV042-
o0
2+ 1 O 0 0 ® 0074
V305(s)
CHejg) = t
2 4 6 8 10 12
pH
❑ Shallow
© Deep
© Bedrock
© Upgradient
• Source
O Downgradient
Figure 15 Pourbaix diagram for vanadium with measured Eh and pH from Cliffside Steam
Station wells
Site -specific measurements of Eh and pH for the 159 wells monitored at the CSS site are also
shown on the Pourbaix diagrams (Figures 1 through 15). These overlays of site measurements
on the predominant phases shown on the Pourbaix diagrams provide an indication of the
expected geochemical forms in which a COI may exist in CSS groundwater in the absence of
interactions and surface complexation with soils, rock, or ash. In these diagrams, the symbology
highlights categories for well type (upgradient, source, and downgradient) and well depth
(shallow, deep, and bedrock).
Antimony
Antimony is a non-metallic element having similar chemical properties to that of arsenic.
However, antimony is only about 1/10th abundant in minerals (Hem 1985). For Eh -pH
conditions measured in the CSS site groundwater, antimony is most likely to exist predominately
as Sb(V) in the form of Sb03 and as Sb(III) in the form of Sb(OH)3(aq) to a lesser extent under
most redox conditions. Numerous potential minerals of antimony exist, with potential
precipitation likely of SbxOx phases most commonly having the highest saturation indices.
Arsenic
Arsenic is a common element found in rocks and sediment in low concentrations across the
United States. Low levels of arsenic in the aqueous phase have been identified to cause
impacts to human health. For Eh -pH conditions measured in the CSS site groundwater, arsenic
is most likely to exist as As(V) in the form of conjugate bases of arsenous acid: H2As04 and
HAs04 Z. Under lower redox potentials, Eh < 0.1, arsenic exists as As(III) in the form of H3AsO4.
13

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
At the highest pH values measured, arsenic may occur in solution as NaAs04 2. However, other
highly aqueous phases such as thioarsenate are not considered in these diagrams that have
been identified in many aquifers under sulfate reducing conditions. Given these circumstances,
it is unlikely that significant arsenic mineral precipitation of formation would occur; rather,
adsorption would be more likely (Hem 1985).
Beryllium
For Eh -pH conditions measured in the CSS site groundwater, beryllium will exist as B(II) and will
be predominantly in the form of Bee+ (aq), Be(OH)+ (aq), or Be3(OH)3+ (aq). For those wells
where pH values range between approximately 8.25 and 9.5 SU, beryllium will exist as Be(OH)2
(s).
Boron
In the CSS site groundwater, boron exhibits relatively simple chemistry existing and
predominantly exists as B(III) in the form of neutrally charged (protonated) boric acid, noted in
the literature as either B(OH)3 or H31303, or as a borate anion H2BO3- (or noted as B02) which
occurs when the pH value exceeds 9 SU.
Chromium
Under the Eh -pH conditions of the CSS site, chromium will predominantly exist as Cr(III) in the
form Cr203 (s), which will likely precipitate. However, under highly oxidizing conditions, Cr(VI)
may become dominant. This could occur due to increases in alkalinity leading to increased pH
or additions of DO from recharge or anthropogenic changes, leading to increased redox
condition changes in the aquifer. Cr(VI) is highly mobile as an aqueous complexed species and,
in situations where oxygen bearing recharge have been identified, has been known to persist
above acceptable limits in widespread areas from point sources in numerous studies (Hem
1985).
Cobalt
In the CSS site groundwater, cobalt is likely to predominantly exist as Co(II) in the form of
aqueous Coe+ as Co(III) is thermodynamically unstable under typical groundwater conditions
(circumneutral pH). For those wells where pH values greater than approximately 8 SU occur,
cobalt is likely to partially oxidize to Co(II,III) and precipitate as C0304 (s), as cobalt has a very
low solubility in natural water.
Iron
Under the Eh -pH conditions measured in the CSS site groundwater, iron is likely to primarily
exist as Fe(II) in the form of Fee+ (aq). For those wells where the pH value is higher than
approximately 8 SU, iron will occur as Fe(III) in the form of Fe203 (s) or alternatively as Fe(OH)3
(s), and is likely to precipitate.
Lead
In the CSS site groundwater, lead is likely to occur predominantly as Pb2+(aq). As the pH is
increased above a pH of approximately 8 SSU, lead will precipitate as Pb(OH)2.
14

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Manganese
Given measured Eh -pH conditions in the CSS site groundwater, manganese is likely to exist
predominantly as Mn(II) in the form of Mn2+(aq).
Nickel
Under the Eh -pH conditions measured in the CSS site groundwater, nickel is likely to exist as
Nit+ (aq).
Selenium
In the CSS site groundwater, selenium is likely to occur predominantly as Se(0) in the form of an
uncharged, solid state, Se (s).
Sulfur/Sulfate
Under Eh -pH conditions in the CSS site groundwater, sulfur is likely to exist entirely as S(VI) in
the form of sulfate (S042-).
Thallium
In the CSS site groundwater, thallium is likely to exist as TI(I) in the form of TI+(aq). However, in
most groundwater samples from the CSS site wells, thallium concentrations were often reported
as being below quantitation limits or as estimates at or near the detection limit. Thallium was
above quantitation limits in just eight of 187 sample analyses.
Vanadium
In the CSS site groundwater, vanadium is expected to predominantly exist as V(III) in the form
of V(OH)3 (s) precipitate. However, in most groundwater samples from the CSS site wells,
vanadium concentrations were typically reported as being below quantitation limits. Vanadium
was above quantitation limits in 38 of 187 sample analyses.
3. Adsorption Model Development
To be useful in modeling electrolyte adsorption, a theory needs to be developed to describe
hydrolysis and the mineral surface, account for electrical charge (electrostatic adsorption), and
provide for adequate mass balance on the sorbing sites. In addition, an internally consistent and
sufficiently broad database of adsorption reactions should accompany the theory. Of the
approaches available, a class known as surface complexation models (Adamson 1976; Stumm
1992) reflects such an ideal most closely and accounts for the parameters to explain overall
adsorption. This class includes the diffuse double layer model (also known as the diffuse layer
model) and the triple layer model (Westall and Hohl 1980; Sverjensky 1993; Bethke 2008).
Of the two models, the diffuse double layer model is more fully developed in literature (Dzombak
and Morel 1987; Bethke 2008) and hence currently the most useful and accepted surface model
in geochemical modeling as compared to other methods (Bethke 2008; Soltermann et al. 2014;
Stoliker et al. 2011). The constant capacitance model commonly refers to an application of
double layer theory that does not account for variation in the electrical charge on the sorbing
surface, whereas diffuse layer models do. Some researchers prefer the term two layer model to
describe an application of double layer theory that accounts for more than one type of sorbing
15

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
site on a single surface, although that distinction is not given in this section (Bethke 2008). In
addition, a Charge Distribution Multi -Site Complexation Model (CD -MUSIC) with triple layer
adsorption and competitive sorption between two or more Iigands was not applied to the model
given the complexity, unknown charges to each planar surface, and inability to consider
amorphous solids using this model (Merkel and Planer -Friedrich 2008).
The use of mechanistic geochemical models has advanced geochemical modeling from a
theoretical approach to a highly accurate method used for nuclear repository and contaminated
site remediation research by the USGS and the bulk of academic researchers (Dulnee et al.
2013; Hyun et al. 2009; Stoliker et al. 2011; Xie et al. 2016). Mechanistic surface complexation
models rely on oxides (iron, aluminum, manganese, humic substances, silica, etc.) on surfaces
to account for the phenomenon of surface complexation. These proxies used in models are
determined by quantifying oxides of these metals in field collected sediment, thus accounting for
the bulk of reactive surfaces with which a species can interact. In doing so, these models excel
over simple Kd approaches in accounting for variable redox, DO, pH, COI concentration (given
likely Langmuir adsorption), and surface site competition (Bethke 2000; Stoliker et al. 2011).
Typically, Kd experiments are also conducted under oxic conditions, which may not reflect actual
site conditions. Consequently, Kd values from laboratory experiments should be interpreted and
used with caution because they may not account for the full range of sorption conditions that
occur in heterogeneous, natural soils.. Experimentally determined Kd values are also highly
dependent solid/liquid ratios, which may not be consistent across lab tests and which may not
reflect actual environmental conditions (Bethke 2000). Surface complexation methods in
geochemical models provide a more consistent means to evaluate how Kd values may change
with COI concentration changes and varying geochemical composition. For this reason,
adsorption reactions were modeled using a double layer surface complexation model. To
ensure consistency in the modeling effort, a single database of constants was used as opposed
to searching for individual constants from the literature. The diffuse double layer model
describing ion adsorption to HFO and HAO or gibbsite by Dzombak and Morel (1990) and
Karamalidis and Dzombak (2010) was selected for this effort.
Given the widespread abundance of surface complexation coefficients in literature, utilizing
HAO and HFO as proxies for adsorption will account for adsorption to dominant facies
controlling surface complexation in this system more comprehensively than other approaches.
Given the heterogeneity of the site, single reaction approaches would not be appropriate or
accurately describe surface complexation across all COls as compared to an HFO/HAO
approach. Many surface complexation reactions for ions of interest have already been added to
the PHREEQC database. Equilibrium constants for aqueous speciation reactions were obtained
from the USGS WATEQ4F database. This database contained the reactions for most elements
of interest except As, Be, Co, Sb, TI, V, and Cr. Constants for aqueous reactions and mineral
formation for these elements were obtained from the MINTEQ v.4 database, which is also
issued with PHREEQC. However, no surface complexation data in either database was
included for thallium adsorption to HFO or HAO. The constants utilized were all checked to
provide a consistent incorporation into the revised database (Martell and Smith 2001).
Densities of HFO and HAO were determined based on extractable aluminum and iron content
from sediment samples collected from the CSS site. The extractable aluminum and iron
16

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
concentrations were determined by UNCC and are shown in Table 2. Based on the modeling
approach described by Dzombak and Morel (1990) and Karamalidis and Dzombak (2010), the
sorption site densities were calculated utilizing 0.2 moles of sorption sites per mole of
extractable iron and 0.41 moles of sorption sites per mole of extractable aluminum. The
extractable aluminum and iron concentrations in the solids were variable. Therefore, a range of
sorption site conditions was calculated in Table 2. Sorption site concentrations were calculated
as follows and based on minimum, mean, and maximum extractable aluminum and iron
concentrations for all tests completed. This approach was utilized as 0.05 and is assumed to be
a fraction of extractable iron available for sorption sites (consistent with Dzombak and Morel
[1990]) and 0.12 of aluminum (Karamalidis and Dzombak [2010]). The extractable iron and
aluminum content for a number of solid samples has been measured and batch Kd
measurements have been performed on those solids by UNCC (refer to Corrective Action Plan
Part 1, Appendix D; HDR 2015). These data indicate an average extractable iron content of
445.5 mg Fe/kg-solid and an average extractable aluminum content of 347.1 mg AI/kg-solid.
This was converted to moles of iron or aluminum per g of solid using the appropriate molecular
weight of HFO and HAO, then converted assuming site density as described previously for iron
and aluminum based on literature:
[-FeOH/-AIOHO] _ [Solid] * [Hf(a)o] * * mol Hf�L* 0.2 mol FeOH (0.41 mot==_AIO A
1000 55.845g (26.98g) mol Hf(a)o
Where:
[=FeOH/=AIOH°] is the concentration of iron adsorption sites in mol/L for model input
[Solid] is the concentration of solid in g/L
[HF(A)O] is the amount of extractable Fe or Al in mg per kg of solid
Mol HF(A)O is the moles of extractable Fe or Al
0.02 or 0.41 is the assumed fraction of extractable Fe or Al adsorption sites
17

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Table 2 Extractable iron and aluminum contents for CSS soil samples
Sample Name
Depth
HFO
HAO
HFO
HAO
HFO
HAO
ff.
mWKg
mg/Kg
moVkg
moVk
mol(sites)/k
mol(si es)/kg
AB - ID
109
153.3
124.8
1.43E-03
1.60E-03
2.87E-04
6.56E-04
AB - 2D
59 - 63
224.8
195
2.10E-03
2.50E-03
4.21E-04
1.02E-03
AS - 3 BR
54.5 - 55.5
230.8
530
2.16E-03
6.79E-03
4.32E-04
2.79E-03
GWA - 55
19 - 28
283
232.8
2.65E-03
2.98E-03
5.30E-04
1.22E-03
GWA - 1 BRU
21 - 22
1534.5
406
1.44E-02
5.20E-03
2.87E-03
2.13E-03
GWA - 13 BR
28 - 29
280.5
183.8
2.62E-03
2.36E-03
5.25E-04
9.66E-04
GWA - 30 BR
27.4
Bedrock - NA
GWA - 2D
28 - 29
809
557.8
7.57E-03
7.15E-03
1.51E-03
2.93E-03
GWA- 1 OD
46.2 - 46.8
600.3
276.3
5.62E-03
3.54E-03
1.12E-03
1.45E-03
GWA - 23D
20.4 - 21.4
302
138.4
2.83E-03
1.77E-03
5.65E-04
7.27E-04
GWA - 20 S
17 - 19
265.75
179.8
2.49E-03
2.31E-03
4.97E-04
9.45E-04
GWA - 24 S
20 - 25
509
454
4.76E-03
5.82E-03
9.53E-04
2.39E-03
IB - 3 SB
49 - 51
531.3
134.8
4.97E-03
1.73E-03
9.94E-04
7.09E-04
MW - 22 BR
41 - 43
696.5
402.5
6.52E-03
5.16E-03
1.30E-03
2.12E-03
US - 3D
40 - 45
1 176.3
204.5
1.65E-03
2.62E-03
3.30E-04
1.07E-03
US - 6 D
34
292.5
427.3
2.74E-03
5.48E-03
5.47E-04
2.25E-03
US - 6 D
50
496
950.8
4.64E-03
1.22E-02
9.28E-04
5.00E-03
IB - 1 D
35 - 40
742.5
767.5
6.95E-03
9.84E-03
1.39E-03
4.03E-03
US - 45
50
676.3
781.3
6.33E-03
1.00E-02
1.27E-03
4.11E-03
AS - 55
39 - 44
491
310.5
4.59E-03
3.98E-03
9.19E-04
1.63E-03
US - 5D
38
673.8
3.4
6.31E-03
4.36E-05
1.26E-03
1.79E-05
AS - 2 S
30
571.5
431.5
5.35E-03
5.53E-03
1.07E-03
2.27E-03
GWA- 33 D
28.1 - 29.1
347.8
222.5
3.25E-03
2.85E-03
6.51E-04
1.17E-03
The basic speciation model was developed by varying the amount of [=FeOH] or [=AIOH°], and
forcing actual measured pH and redox potential with utilization of the fixed ion concentrations
listed in Table 1. The iron adsorption site densities were varied from minimum to the highest
measured level to simulate batch adsorption experiments for these same samples. These
values corresponded to soil suspension concentrations of 100 g/L mass to liquid ratio, with a
site density of 0.2 for HFO, and 0.41 for HAO based on moles of iron or aluminum as previously
described. In titrations, the redox potential was varied between a pE of -5 to 15, corresponding
to redox potential measurements of -5 to 255 mV. Titrations of pH were designed to titrate to
ending points where values of a pH of 2 to 12 were achieved for all =FeOH or [=AIOH°], sites
and redox potential values listed above. Titrations evaluated adsorption as a factor of percent
species adsorbed. In this way, changes in liquid/sediment ratios would not impact final results;
however, Kd values have been calculated and stored across the titrations ranges and are easily
convertible to other sediment/liquid ratios. The range of pH and Eh values were selected taking
into account sites measured values from the CSS site so that the breadth of water chemistries
could be evaluated and potential remediation strategies that would vary these geochemical
parameters. Chemical speciation was evaluated after all [=FeOH] or [=AIOH°] concentrations,
pH values, and redox potentials were run, particularly with respect to changes in redox
speciation, mineral precipitation, and changes to the percent of adsorbed species.
The speciation model outputs were compiled by calculating an average Kd value for model runs
at minimum, mean, and maximum HFO/HAO adsorption at fixed Eh and pH values. For each
category, a median value was utilized to minimize skew from site outliers. The variable surface
site concentrations with constant pH and Eh conditions for each experimental run allowed for
18

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
the evaluation of changes in the predicted Kd value due to competition between ions for
adsorption sites which may be depleted in models using a low suspended solids concentration.
The Kd values calculated for the minimum, mean, and maximum pH and/or Eh field
measurements are shown in Table 3.
Table 3 Minimum, mean, and maximum calculated Kd values
C01
Min
Mean
Antimony
8.33E-06
5.90E-05
Arsenic
3.00E+01
3.76E+02
Boron
1.34E-03
3.75E-02
Barium
4.69E-06
1.06E-05
Beryllium
3.47E+03
9.08E+03
Cobalt
8.63E-02
9.22E-01
Chromium
1.41E+06
3.72E+06
Iron
1.89E-02
6.46E-02
Lead
2.79E+01
3.02E+02
Manganese
2.13E-02
5.25E-02
Nickel
7.39E-02
2.51E-01
Selenium
1.15E+00
2.13E+01
Sulfate
2.54E-03
4.14E-01
Thallium
n/a
n/a
Vanadium
2.64E+00
2.76E+02
MM-
2.63E-04
1.76E+04
1.12E-01
2.43E-05
3.55E+04
2.70E+00
1.67E+07
2.96E-01
9.16E+02
1.32E-01
8.58E-01
1.64E+02
1.27E+00
n/a
7.24E+03
Generally, the experimental data are captured by the minimum, mean, and maximum model -
predicted Kd values. However, it should be noted that predicted Kd values can be linearly scaled
with the assumed fraction of surface reactive iron/aluminum (assumed to be 5% or 12% of the
total extractable iron or aluminum, respectively, in this model as discussed above) and soil to
liquid ratios can differ, leading to changes in Kd values of up to 2 orders of magnitude. For
comparison, experimental data from batch adsorption experiments are provided where
available. While the modeled and experimental Kd values are not exactly the same, the trends
describing the adsorption strength of ions relative to each other are well demonstrated in these
predicted Kd values and the observed trends discussed with regard to the site measurements.
Results of the PHREEQC modeling indicate very low adsorption to HFO/HAO is observed for
the COls antimony, barium, and boron. Adsorption of thallium is unable to be determined as no
adsorption constants exist in either thermodynamic database for thallium adsorption to
HFO/HAO. In several wells, minimal adsorption of these COls is likely occurring due to low
concentrations of the COI in groundwater (e.g., antimony). Major groundwater constituents
calcium and carbonate exhibit zero adsorption to HFO/HAO in all well cases and are not
included in Table 3.
Information from the model was used to develop the Pourbaix figures shown in Section 2 and to
develop the titration output addressed in Section 4. In this way, the utilization of "percent
adsorbed" over a range of pH and other geochemical ranges may be more informative of COI
behavior through the CSS site and in response to potential remediation efforts.
19

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
3.1 Probability Plot Evaluation
Probability plots are provided in Attachment A. These plots provide a comparison of results
across well type (upgradient, source, and downgradient) and well depth (shallow, deep, and
bedrock) categories. The plots provide a graphical evaluation of groundwater chemistry by
illustrating where results for different well type and depth categories occur within the overall
distribution of results. For example, if results for upgradient, source, and downgradient wells are
equally spread throughout a distribution, it would demonstrate that source well and
downgradient well characteristics (for a given analyte) may not differ from upgradient conditions.
In contrast, if results for a distribution show distinct groupings with source wells that are higher
than downgradient wells and with upgradient wells alone at the low end of the distribution, it
would indicate that an analyte may be migrating from source wells to downgradient areas.
Analytical results of these plots are presented without modification (i.e., potentially anomalous
values have not been eliminated). Results flagged as estimates or below quantitation limits
(e.g., values flagged with J or U qualifiers) are presented at their reported values.
Aluminum
Aluminum detections are fairly well -distributed across the CSS site in the shallow, deep, and
bedrock wells as well as the upgradient, source, and downgradient well locations. The highest
concentrations of aluminum were detected in upgradient shallow wells and shallow and deep
source wells. Aluminum does not have a 2L Standard or a USEPA Maximum Contaminant Level
(MCL).
Antimony
The greatest concentrations of antimony were detected in shallow and deep source wells at the
CSS site. Several downgradient deep groundwater wells indicate that antimony may have
migrated at concentrations that exceed the Interim Maximum Allowable Concentration (IMAC).
Arsenic
Arsenic was detected in a number of source wells above the 2L Standard. Arsenic in several
deep downgradient wells at the CSS site was in the upper percentile of detections, but source
well concentrations are evenly spread along the probability plot curve and contain the lowest
concentration detections.
Barium
Barium exceeded the 2L Standard in several deep source wells and deep downgradient wells at
the CSS site, and is well distributed along the probability plot.
Beryllium
The highest concentrations of beryllium were found in deep source wells and deep
downgradient wells at the CSS site, and were generally detected at concentrations less than the
IMAC. A large number of the results were indicated at the detection limit, which accounts for the
horizontal line in the probability plot.
20

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Boron
Boron is present in deep source wells and higher concentrations were detected in a number of
downgradient deep wells at the CSS site. The 2L Standard has been exceeded in source and
downgradient locations based on the probability plot.
Chromium
Chromium was detected in shallow and deep source wells at the CSS site above the 2L
Standard. However, several shallow, upgradient wells have higher concentrations of chromium.
Cobalt
The probability plot indicates the greatest concentrations of cobalt were detected in deep and
shallow source wells at the CSS site at the highest and lowest concentrations. The majority of
the upgradient and downgradient locations are between the source locations. It is unclear what
these results indicate and additional groundwater sampling should be completed.
Iron
Iron appears to be present in higher concentrations in source and deep downgradient wells at
the CSS site above the 2L Standard. The upper and lower results may indicate higher
concentrations of iron, but no discernable pattern is present.
Lead
Results for lead appear to be spread evenly amongst the well locations and depths at the CSS
site and no discernable pattern exists.
Manganese
There are several deep source well results at the CSS site near the upper end of the probability
plot curve. Results for manganese appear to be spread evenly amongst the well locations and
depths, and no discernable pattern exists.
Nickel
Results for nickel appear to be spread evenly amongst the well locations and depths at the CSS
site and no discernable pattern exists. The highest and lowest concentrations of nickel were
found in several shallow upgradient and downgradient locations.
Selenium
The probability plot indicates that several shallow and deep source wells at the CSS site contain
selenium along with several deep, downgradient wells at concentrations less than the 2L
Standard. The remainder of the data indicate that selenium is evenly spread out amongst the
well locations and depths.
Thallium
The probability plot indicates that several shallow and deep source wells at the CSS Site
contain thallium along with several deep, downgradient wells at concentrations less than the
21

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
IMAC. The remainder of the data indicates that selenium is evenly spread out amongst the well
locations and depths.
Vanadium
Vanadium appears to be present in the source wells at the CSS site at concentrations above the
detection limit. A number of shallow, upgradient wells indicated the presence of vanadium.
3.2 Saturation Index Evaluation
In geochemical modeling it is important to consider potential solid/mineral precipitation as both a
mechanism of sequestering constituents or later release should redox, pH, or saturation
equilibrium changes. For the purpose of this discussion, designations of (a) after amorphous
solid species signify a solid phase. Typically, allowing a solid/mineral to precipitate in a model
will decrease the aqueous concentration and allow for equilibrium to occur from a previously
"supersaturated state." In the natural environment, precipitation of pure phase minerals, while
indicated, is not exactly known and more than likely impure or mixed form minerals are likely to
occur. In addition, mineral precipitation secondary to reduction associated with microbiota may
thermodynamically differ due to microbial catalysis rather than strict thermodynamics if the
reaction is thermodynamically favorable and there is energy to be gained from the transfer of
electrons for microbial growth for adenosine triphosphate production. Such reduced or
sequestered minerals can again be dissolved in solution should for example in the case of
Cr(III), redox conditions increase due to increases in dissolved oxygen. However, these systems
in any case will find a point of equilibrium between those minerals phases and aqueous phases.
Changes to ionic strength are also important in considering minerals precipitation. At certain pH
and redox levels, natural groundwater becomes supersaturated with anions/cations and
precipitation will occur.
In PHREEQC, mineral precipitation is determined by the Saturation Index (SI) in the output of a
solution:
Saturation Index (SI) = log ([COI](aqueous) / [CO I](saturated))
The SI of a solid phase can range from extremely negative (undersaturated in solution) to
extremely positive (supersaturated in solution). An SI of zero (0) would indicate the solid/mineral
is in perfect equilibrium with the aqueous phases in solution, although this rarely occurs in
modeling results except when equilibrium with certain mineral phases is manually invoked for
calculations. Most commonly in systems that are in equilibrium, solid phases of chromium, iron,
selenium and manganese are typically indicated to likely precipitate. In these cases this is due
to redox conditions and likely dissolution could also occur given changes in aquifer conditions. It
is important to take into account SI for phases and track them through changing redox
conditions for any potential remediation strategies. Should an SI decrease, it likely indicates that
conditions are causing the increase in an aqueous form of a species, and vice versa. However,
in all cases care should be taken to thoroughly research any potential solid phase of complex
that is indicated to be supersaturated. Thermodynamic data for these species is often lab
collected and tested under pure and controlled conditions, and calculated with an infinite dilution
series (or ionic strength of zero). While supersaturation and potential precipitation of a phase
22

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
may be indicated, it may not be possible in the pure phase shown, which is kinetically too slow
to occur, or pressure and temperature may not be optimal for the reaction catalysis.
Below are some potential phases that may potentially exist in various locations at the CSS site
as identified through geochemical modeling and having an SI greater than zero:
COI POTENTIAL SOLID PHASE CHEMICAL COMPLEX
ANTIMONY
SbOZ; Sb2Se3
ARSENIC
Ba3(ASO4)2
BARIUM
BaSO4; Ba3(AsO4)2
COBALT
CoFe204; CoSe
CHROMIUM
Cr(OH)3(a); Cr203; FeCr204
IRON
Fe(01-1)3(a); FeSe2; FeCr204; Fe00H; Fe203; Fe304; FeCO3; CoFe204
MANGANESE
MnCO3; Mn02
SELENIUM
Se (s); CoSe; FeSe2; Sb2Se3
SULFATE
AI4(OH)10SO4; KAI3(SO4)2(OH)6; BaSO4
VANADIUM
V305
4. Variation of pH, Redox, Dissolved Oxygen, and TDS
A series of solutions were defined for each well at the CSS site to evaluate the impact of
changes in pH, DO, redox, and TDS. Several wells were selected within the flow path from
upgradient to downgradient to generally describe differences in groundwater flow chemistry as
groundwater flows from the upgradient through the source area and downgradient towards the
CSS site. This evaluation was completed in PHREEQC by simulating specific changes in pH,
redox, or TDS over a number of steps or simulations (Table 4). In general, reactions were
completed to equal endpoints of the range specified and the number of steps in PHREEQC to
produce an output. Increases in number of steps increases resolution, but substantially
increases computational time. While redox directly impacts pH, pH was held constant during
redox titrations to achieve a consistent, wide sampling of potential redox and pH combinations.
The impacts of increases in TDS were evaluated as the addition of sodium chloride, potassium
chloride, and gypsum (CaSO4) to address a range of potential cations and anions that are
potentially related to the detection of TDS and address the general ionic strength of the
groundwater. These additions were chosen as proxies for TDS due to their common mineral
occurrence in sediment in these types of aquifers, and are most likely to undergo mineral
weathering. Individual well pH titration results are provided in Attachment B, and redox titrations
in Attachment C. Selected DO and TDS titration results are provided in Attachment D and E,
respectively.
23

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Table 4 Titration reactants
Reaction
Range
# Steps
pH
2 to 12
500
REDOX (Eh)
-500 to +750 mV
300
TDS
10e-3 moles
300
Oxygen (DO)
5.0e-2 moles
300
Titration of these reactants using the minimum, maximum, and mean Kd values (Table 3) will
provide a general understanding of impacts due to changes in groundwater chemistry and
facilitate decisions on various groundwater alternatives and MNA activities. For example,
maintaining the pH at near -neutral conditions may prevent metal species formation such as
Cr(VI) to remain reduced as Cr(III), which is highly insoluble in this system and easily adsorbed.
A titration model was completed for each of the 159 wells on CSS for each of the COls to show
the percent of COls adsorbed, and COI speciation versus pH, DO, redox, and TDS for
minimum, mean, and maximum adsorption, resulting in 17,000 simulations. Review of titration
simulations indicates that the greatest impact to adsorption of Cols occurs when pH and redox
are modified. Changes in response to DO and TDS variations are less than an order of
magnitude and do not appear to materially alter adsorption of COls relative to variations in pH or
redox. DO changes are noted to either not substantially impact adsorption, or react quickly at
very low concentrations (below 0.1 mg/L) and further adsorption remains consistent throughout
saturation. For DO, only figures that indicate notable changes in percent species adsorbed from
solution are provided in Attachment D.
Based on review of the CSS site layout, one transect was selected for comparison based on
flow from upgradient to the downgradient locations through the source material:
Transect 1:
MW-30S/D
AS-7S/D/BR
GWA-21 D/BR
Representative results for Transect 1 are presented in Figures 16 through 21.
24

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
0.0
Cr at Site CliffsIde, Well MW-30D for max HFOlHAO values, pH Titration
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
10 1
OL. _ 110 -2
OX 1.0 2.6 3.0 4,0 5.0 6.0 7.0 8.0 9,0 1❑-0 11.0 12-0
pH (SU)
Figure 16 Simulated response of chromium absorption and concentrations in response to pH
variations for Well MW-30D based on the upper bound for surface complexation to HFO and HAO
Co at Site Clillside, Well MW-30D for max HF&HAO values, pH Titration
0.0 1.0 2.0 3A 4A 5.0 6A 7.0 8.0 9A 10.0 11.0 12A
100 10
1013
t
to
(9
n
[0
UP
0
O
7
10 2 0
7
m
O
7
■ Total_Cobalt _ug1L
0 C.M ugiL
C C-13) uugg+L
11 %C. Adenr6ed
01 - d 10 -4
0.0 1.0 2.0 3.0 49 5.0 6.0 7.0 8.0 9.0 10.0 11A0 12.0
pH (SU)
Figure 17 Simulated response of cobalt absorption and concentrations in response to pH
variations for Well MW-30D based on the upper bound for surface complexation to HFO and HAO
25

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Cr at Site Cliffside, Well AS-7S for max HFO/HAO values, pH Titration
1.0 1.0 20 3.0 4.0 5.0 6.9 7.0 8.0 9.0 10.0 11.0 12.0
1�
r
ag
70 �+
ea I
a �
50
40
I
I I
30 I
■ Tola Chmmium_LgL
o Cr(21_ug4
20 0 C,{31_ug'L
o Cro) uugg L
0 %Cr Ads rbed
10 �I
10
C
U3
10 $
10 '®
0.0 1.0 2A 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
pH (SU)
Figure 18 Simulated response of chromium absorption and concentrations in response to pH
variations for Well AS-7S based on the upper bound for surface complexation to HFO and HAO
Cr at Site Cliffside, Well AS-7S for max HFOIHAO values, pH Titration
0.0 to 2.0 3.o 4A 5.0 5.0 7.0 8.0 9.0 fo.0 11.0 12.0
loci 10
10 .7
all
d 10 '2
■ Tou
® Cr12
0 Gr{3
O ■ G Z?l
w
v
ro
n
10 -3 y.
0
0
a
10 4
a
Gt
fo _5 O
a
C
i0
10
r
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9A 10.0 11.0 12.0
pH (SU)
Figure 19 Simulated response of cobalt absorption and concentrations in response to pH
variations for Well AS-7S based on the upper bound for surface complexation to HFO and HAO
26

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Cr at Site Cliffside, Well GWA-21S for max HFOIHAO values, pH Titration
0.0 1.0 20 3.0 4.0 5.0 6.0 7.0 8.0 9.0 f0.0 1110 12.0
10
2
10
9.0 10.0 11.0 12.0
pH (SU)
Figure 20 Simulated response of chromium absorption and concentrations in response to pH
variations for Well GWA-21S based on the upper bound for surface complexation to HFO and HAO
Coat Site Cliffside, Well GWA-21S for max HFO/HAO values, pH Titration
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 &0 9.0 0.0 11.0 12.0
inn n 9
■ Tot
E8 Co(.
G0 %
a °b
10 2
w
v
n
1 q 1 14
0
0
a
0
ro
a
10 o
m
0
a
C
i0
10 1
0.0 1.0 2.0 3.0 4.0 5.0 6.0 TO 8.0 9.0 10.0 11.0 12.0
pH (SU)
Figure 21 Simulated response of cobalt absorption and concentrations in response to pH
variations for Well GWA-21S based on the upper bound for surface complexation to HFO and HAO
27

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Summary of Transect 1
MW-30S/D
AS-7S/D/BR
GWA-21 D/BR
The general shape of adsorption curves is similar from well to well for an individual COI for the
pH titration but can differ substantially between COls. A brief discussion for the maximum Kd
value is provided for each COI based on the upgradient, source or downgradient location
through the transect.
MW-30S/D (Upgradient)
Antimony
Antimony does not change with pH. The adsorption line is not changed by pH.
Arsenic
The shallow and deep soils behave similarly. Arsenic is generally 100% adsorbed onto soils at a
pH of 2 SU. Arsenic (5) is generally the least sorbed at a pH of 6 SU, but resorbs as the pH is
increased or decreased.
Beryllium
Shallow and deep soils behave similarly for the adsorption of beryllium. In general, beryllium is
sorbed at pH levels between 5 SU and 10 SU.
Boron
Boron does not change with pH, the line for both shallow and deep soils is flat.
Chromium
Based on titration results for chromium, the percentage of chromium in a sorbed phase reaches
a peak near acidic to neutral pH, indicating the greatest binding to HFO/HAO sites and/or
precipitation to a solid phase. As pH increases, chromium is expected to desorb would be and
released to groundwater as Cr(VI). The peak adsorption in the deep well is around a near
neutral pH.
Cobalt
Shallow and deep soils behave similarly for cobalt. The maximum adsorption occurs at a pH of
7 SU at which point it begins to desorb as Co (3).
Iron
Iron does not change with pH. The line for both shallow and deep soils is unchanged with
changing pH.
28

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Manganese
Manganese indicates its maximum adsorption between a pH of 9 SU and 10 SU. Manganese
(2) desorbs at a pH of 7 SU and is least sorbed at pH of 9 SU and 10 SU.
Selenium
In CSS groundwater, selenium is likely to occur predominantly as Se(0) in the form of an
uncharged, solid state, Se(s).Thallium
Thallium (1) is primarily sorbed at all pH ranges, but indicates a greater sorption of TI (3) at a pH
of 5 SU.
Vanadium
Vanadium indicates the greatest adsorption at a pH of approximately 6 SU. Other vanadium
species desorb at a pH of 4 SU.
AS-7S/D/BR
The adsorption of antimony, arsenic, beryllium, boron, cobalt, iron, selenium, sulfate, thallium,
and vanadium are similar to the upgradient locations.
Chromium
Chromium is similar to the upgradient adsorption plots; however, the maximum adsorption is
shifted to a pH of 7 SU and Cr (VI) is least sorbed at a pH of 7 SU. Cr (6) desorbs from the
matrix near a pH of 11 SU.
Manganese
The manganese curve is similar to the upgradient wells; however, manganese (6) sorption
occurs at a pH greater than 12 SU in the shallow wells. The deep and bedrock wells are similar
to the upgradient well.
GWA-21 D/BR/BRU
The adsorption of antimony, arsenic, beryllium, boron, cobalt, iron, manganese, selenium,
thallium and vanadium are similar to the upgradient locations.
Chromium
The chromium curve for the bedrock is similar to the upgradient MW-30S shallow well, and the
deep, downgradient well is similar to the MW-30D curve with the tighter adsorption curve.
The general shape of adsorption curves is similar from well to well for an individual COI, but can
differ substantially between COls. Based on titration results for chromium, the percentage of
chromium in a sorbed phase reaches a peak near mildly acidic to neutral pH, indicating the
greatest binding to HFO/HAO sites and/or precipitation to a solid phase. As pH increases,
chromium is expected to desorb and released to groundwater as Cr(VI). This is supported by
the Pourbaix diagrams presented in Section 2. In contrast, manganese indicates its maximum
29

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
adsorption between pH of 9 SU and 10 SU, whereas chromium has its greatest adsorption
between pH of 2 SU to 7 SU.
Titration results for the CSS site wells can be used to support evaluation of MNA or remediation
impacts. For example, titration results can be used to help determine the expected impact that
DO changes would have in response to addition of an engineered cap (leading to reduced
infiltration and lower recharge DO), or the introduction of oxygen creating a more oxic
environment, addition of acid or base to adjust the pH to conditions that prevent COls from
being solubilizing, or impact due to excavation and the release of TDS and other metals.
Changes in redox can occur also in response to DO increases or decreases as well as the
introduction of inorganic oxidants from anthropogenic contamination or changes in groundwater
flow vectors.
5. Sorption Capacity
A study of the potential soil sorption capacity of COls was completed as a precursor to MNA
Tier III site evaluations of acceptable groundwater concentrations. The sorption potential of soils
was evaluated using PHREEQC to numerically titrate a mixture of COls to evaluate how
aqueous and sorbed phase concentrations vary in response to COI additions. In these titrations,
the ratio of COls added was calculated as the geometric mean of COI concentrations measured
in source wells for the CSS. The eight COls in the mixture were antimony, arsenic, boron,
cobalt, chromium, iron, manganese, and vanadium. This mixture was incrementally added to the
mix of COls present in groundwater of each downgradient well at the CSS site.
As detailed in Section 3, the geochemical model accounts for surface complexation, ion
competition for binding sites, and the equilibrium that occurs between groundwater constituents
under different redox conditions as expressed by pH and electrochemical potential (eH).
Surface complexation in the model was again represented as a range described by minimum,
mean, and maximum distribution coefficient (Kd) values for CSS site soils.
With each incremental addition of Cols to a well, the model computes the distribution between
the aqueous and sorbed phases of the COI and accounts for changes to surface complexation
with other ions in solution. For downgradient wells and sorption capacity cases, titration results
are shown as a series of curves on graphs that express the relationship between aqueous and
sorbed COI concentrations for all eight COls in the titration mixture. On each graph, the
applicable 2L Standards or USEPA MCL is also shown for each COI. Under conditions where
sorption is favored, sorbed phase COls concentrations increase much more rapidly than
aqueous concentrations. When conditions unfavorable for sorption occur, aqueous COls
concentrations increase much more rapidly than sorbed phase concentrations.
In general, soil sorptive capacity for COls such as boron is typically small, yet even a small
addition of boron to groundwater is expected to result in increased aqueous concentrations of
boron. In contrast, soil sorptive capacity for COls such as arsenic or chromium is much higher
and relatively larger amount of arsenic or chromium could be added to soils in downgradient
areas of the CSS site without resulting in aqueous concentrations that exceed groundwater
standards.
30

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
It is important to note that the relationship between aqueous and sorbed COI concentrations is
an equilibrium processes. As aqueous COI concentrations increase, there will be increases in
sorbed COI concentrations. Similarly, as aqueous COI concentrations decrease, there will be
decreases in sorbed concentrations. It is also important to note that the behavior of each COI
evaluated can differ. There may be no point in the spectrum of redox conditions were every
possible COI will sorb to soils. As noted above, some COls show little sorption potential while
others may be more readily sequestered in soils. Graphs displaying COI titration results for
downgradient wells are presented in Attachment F.
6. Summary
The modeling effort described above provides both qualitative and quantitative estimations of
the chemical speciation and adsorption behavior of several key COls. Relevant observations
from this modeling effort are as follows:
• Redox conditions vary widely at the CSS site, indicating that it has not reached
equilibrium, or data is not representative of the conditions sampled. Additional
groundwater results will assist in refining the model further or confirm these findings.
• The observed site condition of limited solubility of arsenic, chromium, and cobalt in the
CSS site groundwater is confirmed by the modeling.
• Each of the pH, Eh, and TDS figures can be further evaluated to potentially support MNA
or remediation. The addition of an engineered cap would reduce infiltration and
introduction of oxygen presumably creating a more anoxic environment. Alternately, the
addition of acid or base could adjust the pH to conditions that prevent COls from
solubilizing, or impacts due to excavation and the release of TDS and other metals.
• Methods such as capping could change groundwater flow, inhibit current microbial
reduction mechanisms, or cause secondary metal redox reductions that could be
modeled when options are chosen more specifically.
• Soil sorptive capacity for COls such as boron is typically small and relatively larger for
COls such as arsenic or chromium.
7. References
Adamson, A.W. 1976. Physical Chemistry of Surfaces. Wiley, New York.
Bethke, C. 2008. Geochemical and Biogeochemical Reaction Modeling. Cambridge University
Press.
Bethke, C. and Brady, P. 2000. How the Kd Approach Undermines Ground Water Cleanup.
Groundwater, 38(3): p. 435-443.
Davis, J. A, Coston, J. A., Kent, D. B., and Fuller, C. C. 1998. Application of the surface
complexation concept to complex mineral assemblages. Environmental Science and
Technology, 32: p. 2820-2828.
Davis, J. A., James, R.O., and Leckie, J. O. 1978. Surface ionization and complexation at the
oxide -water interface.1. Computation of electrical double layer properties in simple
electrolytes. Journal of Colloid and Interface Science, 63(3): p. 480-499.
31

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Dulnee, S., Banerjee, D., Merkel, B. J., and Scheinost, A. C. 2013. Surface Complexation and
Oxidation of Snll by Nanomagnetite: Environmental Science & Technology, v. 47, no. 22,
p. 12852-12859.
Dzombak, D. A. and Morel, F. M. M. 1990. Surface complexation modeling: hydrous ferric
oxide. New York: Wiley. xvii, 393.
Goldberg, S. 1990. Use of Surface Complexation Models in Soil Chemical Systems, in
Advances in Agronomy, D.L. Sparks, Editor. Academic Press, Inc. p. 233-329.
HDR. 2015. Corrective Action Plan Part 1. Cliffside Steam Station Ash Basin, December 2015.
Hem, J. D. 1977. Reactions of metal ions at surfaces of hydrous iron oxide. Geochem.
Cosmochim. Acta., 41: p. 527-538.
Hem, J. D. 1985. Study and Interpretation of the Chemical Characteristics of Natural Water,
United States Geological Survey Water Supply Paper 2254.
Hyun, S. P., Fox, P. M., Davis, J. A., Campbell, K. M., Hayes, K. F., and Long, P. E. 2009.
Surface complexation modeling of U(VI) adsorption by aquifer sediments from a former
mill tailings site at Rifle, Colorado: Environ Sci Technol, v. 9368-9373.
Karamalidis, A. K. and Dzombak, D. A. 2010. Surface Complexation Modeling: Gibbsite.
Hoboken, NJ: John Wiley and Sons, Inc.
Kinniburgh, D. and Cooper, D. 2011. PhreePlot: Creating graphical output with PHREEQC.
Centre for Ecology and Hydrology, Deiniol Road, Bangor, Gwynedd, LL57 2UW, UK.
Available at: http://www.phreeplot.org/.
Martell, A.E. and R.K. Smith. 2001. Critical Stability Constants, Standard Reference Database
46, Version 6.30. National Institute of Standards, Gaithersburg, MD.
Merkel B. and Planer -Friedrich, B. 2008. Groundwater Geochemistry (Edited by Darrell Kirk
Nordstrom). Berlin: Oxford Press.
Parkhurst, D. L. and Appelo, C. A. J. 2013. Description of input and examples for PHREEQC
version 3—A computer program for speciation, batch -reaction, one-dimensional transport,
and inverse geochemical calculations: U.S. Geological Survey Techniques and Methods,
book 6, chap. A43, 497 p. Available only at http://pubs.usgs.gov/tm/06/a43.
Soltermann, D., Baeyens, B., Bradbury, M. H., and Fernandes, M. M. 2014. Fe(II) Uptake on
Natural Montmorillonites. Il. Surface Complexation Modeling: Environmental Science &
Technology, v. 48, no. 15, p. 8698-8705.
Sposito, G. 1989. The chemistry of soils. Oxford: Oxford University Press.
Stachowicz, M., Hiemstra, T., and van Riemsdijk, W. H. 2006. Surface speciation ofAs(111) and
As(V) in relation to charge distribution. Journal of Colloid and Interface Science, 302(1): p.
62-75.
32

Geochemical Modeling Report
Cliffside Steam Station Ash Basin
Stoliker, D. L., Kent, D. B., and Zachara, J. M. 2011. Quantifying Differences in the Impact of
Variable Chemistry on Equilibrium Uranium(VI) Adsorption Properties of Aquifer
Sediments: Environmental Science & Technology, v. 45, no. 20, p. 8733-8740.
Strawn, D., Doner, H., Zavarin, M., and McHugo, S. 2002. Microscale investigation into the
geochemistry of arsenic, selenium, and iron in soil developed in pyritic shale materials.
Geoderma, 108(3-4): p. 237-257.
Stumm, W. 1992. Chemistry of the Solid -Water Interface. New York Wiley.
Sverjensky, D. A. 1993. Genesis of stratabound ore deposits in the midcontinent basins of North
America. 1. The role of regional groundwater flow. American Journal of Science 293, 497-
497
USEPA (U.S. Environmental Protection Agency). 2007. Monitored Natural Attenuation of
Inorganic Contaminants in Ground Water. Volume 2. Assessment for Non -Radionuclides
Including Arsenic, Cadmium, Chromium, Copper, Lead, Nickel, Nitrate, Perchlorate, and
Selenium.
Westall, J. and Hohl, H. 1980. A comparison of electrostatic models for the oxide/solution
interface. Advances in Colloid Interface Science 12, no. 4: 265-294.
Xie, Y., Helvenston, E. M., Shuller-Nickles, L. C., and Powell, B. A. 2016. Surface Complexation
Modeling of Eu(III) and U(VI) Interactions with Graphene Oxide: Environmental Science &
Technology.
Zawislanski, P. T. and Zavarin, M. 1996. Nature and rates of selenium transformations: A
laboratory study of Kesterson Reservoir soils. Soil Science Society of America Journal,
60(3): p. 791- 800.
33

This page intentionally left blank