2016-05-13 Parette Expert Report- Roxboro Site - 05-13-16

MATSON
ASSOCIATES
Opinions on the Appropriateness of Monitored Natural Attenuation
in Conjunction with Cap -in -Place at the Roxboro Steam Station,
Semora, NC
Prepared for:
Southern Environmental Law Center
22 South Pack Square, Suite 700
Asheville, NC 28801
Prepared by:
Robert Parette, Ph.D., P.E.
May 13, 2016

Table of Contents
I. Qualifications...................................................................................................................................1
11. Scope of Wo k.................................................................................................................................1
III. Basis of Opinion...........................................................................................................................1
IV. Methodology...............................................................................................................................1
V. Backg ound.....................................................................................................................................3
A. Coal Ash and the Roxbo o Site.....................................................................................................3
B. Monito ed Natu al Attenuation (MNA)........................................................................................4
VI. Opinion: Monito ed Natu al Attenuation is not an app op iate emedy at the Roxbo o site......... 7
A. A numbe of constituents of inte est at the Roxbo o site a e gene ally poo ly adso bed..............7
B. The bed ock at the Roxbo o site and the p oximity of the ash basins to Hyco Lake is p oblematic
with ega d to attenuation by adso ption uncle site-specific conditions at the Roxbo o site ............... 8
C. Kd values dete mined in the labo ato y cannot be eliably utilized as a means to demonst ate
site-specific attenuation fo the Roxbo o site.......................................................................................8
D. When MNA is based upon adso ption of COls onto i on and manganese oxides, the utilization of
MNA in combination with capping is p oblematic as conditions fo MNA a e less favo able following
capping................................................................................................................................................ 9
E. MNA based upon adso ption is not an app op iate emedy fo a senic at the Roxbo o site........ 10
VII. Refe ences.................................................................................................................................12

I. Qual f cat ons
I am an Envi onmental Enginee (P.E.) at Matson & Associates, Inc. (M&A). My academic
backg ound includes a B.S. in Chemical Enginee ing f om Wo ceste Polytechnic Institute and a
M.S. and a Ph.D. in Envi onmental Enginee ing f om The Pennsylvania State Unive sity. I have
ove 10 yea s of expe ience in the emoval of va ious contaminants f om envi onmental media.
I have co-autho ed seve al pee - eviewed pape s pe taining to the emoval of a senic f om
g oundwate utilizing i on o i on-tailo ed adso bents. I have also p eviously evaluated
emedies fo a senic contaminated g oundwate (including MNA) at coal ash disposal sites, as
well as the use of va ious technologies to t eat effluent impacted by metals and othe
constituents at a coal mining site. My CV is attached in Appendix A.
II. Scope of Work
I was etained by the Southe n Envi onmental Law Cente (SELC) to evaluate whethe
Monito ed Natu al Attenuation (MNA), in conjunction with Cap -in -Place, is an app op iate
emedy fo Duke's Roxbo o Steam Station (Roxbo o site) nea Semo a, NC.
III. Bas s of Op n on
My opinion is based on site specific documents, as well as scientific lite atu e, and my education
and expe ience. My opinion is based on a easonable deg ee of enginee ing and scientific
ce tainty. I ese ve the ight to supplement my opinions should additional info mation become
available.
IV. Methodology
My ole in this matte was to evaluate whethe MNA, in conjunction with Cap -in -Place, is an
app op iate emedy fo the Roxbo o site. An evaluation of the app op iateness of MNA at a site
equi es an in-depth uncle standing of the p ocesses that constitute natu al attenuation at the
site and how the site specific envi onmental conditions impact these p ocesses. Fo the
Roxbo o site, knowing the constituents of inte est (COls) and that the emedial st ategy
p oposed by Duke Ene gy was based in -pa t on co -p ecipitation and adso ption, I evaluated
whethe these p ocesses we a occu ing uncle the specific conditions found at the Roxbo o site.
My evaluation utilized the c ite is established by the U.S. EPA fo MNA of ino ganic
contaminants in g oundwate ; the same c ite is identified by Duke's consultants in thei
analyses.
My opinions a e based on conside ation of data and info mation p ovided in the Comp ehensive
Site Assessment Repo t, the Co ective Action Plan Pa t 1, and the Co ective Action Plan Pa t 2
fo the Roxbo o site; and on methods that I have used in othe scientific o enginee ing inqui ies
conce ning envi onmental p ocesses, which a e based upon methods and analysis widely
accepted in the published, pee - eviewed lite atu a in the fields of envi onmental science and
1

enginee ing. Fi st, I examine the data available to dete mine whethe a p ocess is wo king, and
if not I then conduct esea ch to dete mine why. This esea ch gene ally involves a eview of
the elevant pee - eviewed lite atu e followed by an analysis of the applicability of the lite atu e
to the p ocess being evaluated.
In this matte I examined all of the available data fo the coal ash, the g oundwate , and the
solid-wate pai analyses to gain an uncle standing as to whethe COls we e natu ally
attenuating and to gain an uncle standing of the site conditions influencing natu al attenuation.
I then conducted a lite atu a sea ch fo studies on simila MNA applications. Since the
application of MNA as a t eatment fo COls in g oundwate at coal ash sites is new, published
lite atu a on sites whe e it has been implemented fo such pu pose is not yet available fo
assessment of MNA's pe fo mance in this capacity. The efo e, I eviewed a numbe of studies in
the pee - eviewed lite atu a that had examined the effectiveness of t eating coal ash leachate
o individual COls in othe systems with i on o manganese oxides in active o enginee ed
t eatment systems. These studies included my own publications investigating how i on and
a senic behave in g oundwate t eatment ope ations in o de to design effective t eatment
systems fo a senic emoval. Based on my knowledge, t aining, education, and yea s of
expe ience esea ching emoval by adso ption onto i on oxides, I knew that a numbe of facto s
can influence the adso ption of COls to metal oxides so I conducted an investigation to
dete mine whethe the Roxbo o site conditions we a impacting MNA.
My ole also involved evaluating the effectiveness of MNA uncle Duke's p oposed plan of
capping the Roxbo o site. The efo e, I conducted a simila analysis of MNA uncle alte ed site
conditions based on this p oposed scena io.
My analysis focused on a numbe of COls as identified in my opinion below in Section VI. Fo the
COls not discussed, I am not making a dete mination on whethe MNA is an app op iate
emedy.
K

V. Background
A. Coal Ash and the Roxbo o Site
• Coal ash is the waste mate ial p oduced f om the bu ning of coal in coal-fi ed powe
plants. It is comp ised of a numbe of byp oducts. Fo example, fly ash is "a ve y fine
powde y mate ial comp ised mostly of silica made f om the bu ning of finely g ound
coal in a boile ." (U.S. EPA 2016). Bottom ash, is aptly named fo the ash pa ticles
that a e too la ge to be ca ied into the smoke stacks and thus collect in the bottom
of the boile (U.S. EPA 2016). Both p oducts a e commonly efe ed to as coal ash.
• Coal ash in gene al is comp ised mainly of oxides of silica, aluminum, i on and
calcium, and also contains nume ous othe elements such as a senic, bo on,
ch omium, cobalt, and vanadium. Exposu a of coal ash to wate is known to cause
the leaching and elease of coal ash constituents to g oundwate and su face wate
(Elect is Powe Resea ch Institute 2009, Liu et al. 2004, Lokeshappa et al. 2014,
Abe nethy et al. 1969).
• The Roxbo o Steam Station (Roxbo o site) is located along the easte n bank of the
Hyco Lake, nea Semo a, No th Ca olina. Beginning in 1966 coal was bu ned at the
site to p oduce elect icity. (Co ective Action Plan Pa t 1).
• Two ash basins have been utilized at the Roxbo o site. These a e efe ed to as the
"1966 semi -active east ash basin" and the "1973 active west ash basin", the yea s
eflecting the dates of initial const uction. An unlined ash landfill was const ucted
atop the east ash basin in the late 1980s and a lined landfill was const ucted ove
the west ash basin in 2004. In total, these ash basins occupy app oximately 495
ac es (Co ective Action Plan Pa t 1). Acco ding to Duke's website, the "total
volume of ash" at the Roxbo o site is 34.47 million tons (Duke Ene gy Ash Met ics).
The thickness of ash is app oximately 55 to 80 feet in the east basin and
app oximately 80 feet in the west basin. The ash basins a e impounded by ea then
dams (Comp ehensive Site Assessment Repo t).
• The ash basins we a fo med by impounding st eam valleys with ea then dams.
Duke indicated that g oundwate gene ally flows no th/no thwest ac oss the site to
the Hyco Rese voi (Comp ehensive Site Assessment Repo t).
• Coal ash disposal at the Roxbo o site has led to a numbe of constituents in site
g oundwate that exceed applicable standa ds. Based on the Co ective Action Plan
Pa t 1, these constituents a e antimony, ba ium, be yllium, bo on, chlo ide,
ch omium, cobalt, i on, manganese, nickel, selenium, sulfate, thallium, total
dissolved solids (TDS), and vanadium. A senic concent ations in wate sc eened
within the ash have also significantly exceeded its standa d. These g oundwate
constituents, he ein efe ed to as constituents of inte est (COls), will t avel with the
3

g oundwate unless they a e emoved (adso bed, p ecipitated, etc.) along the
g oundwate flow path.
• Duke has p oposed to utilize cap -in-place in combination with monito ed natu al
attenuation at the Roxbo o site. Duke indicates that cap -in-place at the Roxbo o
site will place "a low pe meable line ove the ash basin to contain the ash and
p event ainwate infilt ation" (Co ective Action Plan Pa t 1 Appendix Q.
B. Monito ed Natu al Attenuation (MNA)
• MNA elies solely on natu al envi onmental geochemical p ocesses to mitigate
concent ations of contaminants (no active emediation) with egula ly scheduled
monito ing to assess whethe MNA is wo king. Acco ding to the USEPA (2012):
"Monito ing typically involves collecting soil and g oundwate samples to
analyze them" fo the contaminants of conce n and othe site pa amete s. The
" ight conditions must exist uncle g ound to clean sites p ope ly and quickly
enough."
• COls at the Roxbo o site a e elements (ino ganics), and the efo a cannot be
dest oyed. Natu al p ocesses to immobilize ino ganics include adso ption,
p ecipitation/co-p ecipitation, and/o oxidation- eduction eactions (U.S. EPA
2007a). Duke indicated at the Roxbo o site that "attenuation is in the fo m of
dilution as constituents flow th ough the system; so ption whe eby aqueous ions
a e emoved by chemical inte actions with a su face [adso ption]; and chemical
p ecipitation which includes homogeneous mine al p ecipitation and co-
p ecipitation" (Co ective Action Plan Pa t 2). P ocesses such as adso ption and co-
p ecipitation can potentially immobilize a COI, and evaluating the effectiveness of
natu al attenuation via these p ocesses is the focus of this epo t. While
concent ations of COls can be educed via dilution, dilution does not p event the
movement o discha ge of a COI f om a site.
• Duke is elying in pa t upon adso ption/co-p ecipitation of COls onto hyd ous metal
oxides, such as i on and manganese in its selection of MNA as a emedy. In educed
fo ms (i on and manganese in the +2 oxidation state), i on and manganese (also
components of coal ash) have highe solubility and mobility. As these species
mig ate f om thei sou ce and encounte mo a oxidized conditions, p ecipitation
(pa ticula ly fo i on) as insoluble species (i on in the +3 oxidation state and
manganese in the +4 oxidation state) occu s c eating su faces onto which COls can
be adso bed. MNA based on these p ocesses is successful only uncle the ight
conditions, which will be discussed in mo a detail below.
• In gene al, many g oundwate constituents a e known to adso b onto i on oxide
su faces. In gene al, adso ption of g oundwate constituents onto manganese oxide
4

su faces is not as well studied, though it is known that manganese oxide is not an
effective adso bent fo a senic, but has been epo ted to have high affinity fo
cobaltuncle the ight conditions (Hoffman et al. 2006, Salminen 2005).
• In gene al, the e a e a numbe of facto s that influence the effectiveness of natu al
attenuation fo ino ganics at a site. These facto s include (but a e not limited to)
the oxidation- eduction potential of the g oundwate , the pH, competition fo
adso ption sites, complexation, the elative ates of oxidation between a COI and a
potential adso bent such as i on, the amount of COls in the coal ash sou ce, and
atios of COls to adso bents.
o The speciation of a COI is dependent upon pH and the oxidation- eduction
potential of the g oundwate . Potential emoval mechanisms fo a constituent
can change depending upon the speciation of the constituent.
o In addition to COls having to compete against each othe fo adso ption sites,
othe constituents p esent in the g oundwate can also significantly impact the
adso ption of COls onto hyd ous metal oxide su faces. Fo example, silica,
phosphate, bica bonate, and dissolved o ganic ca bon can p esent competition
p oblems (Meng et al. 2002, Meng et al. 2000, Holm et al. 2008, Su and Puls
2003, Holm 2002, Mille 2001, Swedlund and Webste 1999, Chen et al. 2007,
Sa ka et al. 2014, Ma iussen et al. 2015, U.S. EPA 2007b). Based on my eview
of site data, to my knowledge, concent ations of silica and phosphate in
Roxbo o site g ound g oundwate have not been monito ed.
■ G oundwate typically contains silica in the ange of 3.3 to 21 mg/L (as Si)
(Holm et al. 2008, Davis et al. 2001). Silica is known to limit the adso ption
of COls onto i on oxides (Meng et al. 2000, Davis et al. 2001, Swedlund and
Webste 1999). Dissolved silica at concent ations even below the typical
ange found in g oundwate has been shown to limit adso ption of the COls
(Mille 2001). Silica comp ises a majo f action of coal ash by weight
(Elect is Powe Resea ch Institute 2009), and has been found in the
g oundwate between 180 and 650 mg/L as Si at a coal ash site in the mid-
Atlantic egion.
■ Concent ations of phosphate below 1 mg/L have been shown to limit
adso ption of COls. The p esence of silica is known to magnify the impact of
phosphate. Phospho us pentaoxide (which fo ms phosphate in wate ) is a
known constituent of coal ash (Meng et al. 2002, Meng et al. 2000, Holm et
al. 2008, Su and Puls 2003, Holm 2002, Blackmo a et al. 1996, Liu et al.
2004).
■ With ega d to cobalt adso ption onto manganese oxide, othe metals such
as coppe , nickel, and zinc may outcompete it fo adso ption sites
L

(Mukhe jee et al. 2013). The p esence of bo ate (bo on is a site COI) can
also significantly educe the adso ption of cobalt onto manganese oxides
(Hasany and Qu eshi 1981).
o Othe constituents in g oundwate can fo m complexes with metals which can
dec ease COI adso ption/co-p ecipitation. Fo example, vanadium, a site COI,
can fo m complexes with sulfate (Salminen 2005, U.S. EPA 2007a).
o When utilizing i on to emove COls, such as a senic and vanadium, oxidation of
these ino ganics should occu simultaneously with the oxidation of i on to
achieve optimal emoval (Hoffman et al. 2006, Rocca o and Vagliasindi 2015).
• Due to the numbe of va iables that can impact MNA of ino ganics, the feasibility of
utilizing natu al attenuation fo ino ganics must be assessed on a site -by -site basis
because the effectiveness is highly site specific (Reisinge et al. 2005, Mille 2011).
• Acco ding to the U.S. EPA, dete mining the feasibility of using MNA at a site fo an
ino ganic constituent at a site is a fou -tie ed evaluation (U.S. EPA 2007a).
• The envi onmental geochemical p ocesses that constitute natu al attenuation fo
COls at the Roxbo o site a e ongoing, and have been ongoing since coal ash disposal
began; these p ocesses do not begin at the time MNA is app oved as a emedy fo a
site. The efo e, a Tie I evaluation establishes whethe natu al attenuation is
cu ently occu ing at the site, and if so is the fi st step in demonst ating the long
to m app op iateness of MNA as a site emedy.
• In the Tie I evaluation p ocess, concent ations of COls a e commonly measu ed in
site wate and soil (solid) f om the same location and depth. If attenuation is
occu ing via adso ption onto aquife solids, the concent ation of the COI on the
solid mate ial should inc ease with the concent ation of the COI in the g oundwate .
An example of this type of analysis is shown below in Figu e 1 (U.S. EPA 2007a). A
st ong positive co elation between the COI concent ation in wate and solid pai s
indicates attenuation. Conve sely, acco ding to the U.S. EPA, sites whe e COI
attenuation is not obse ved "would be eliminated f om fu the conside ation of
MNA as pa t of the cleanup emedy" (U.S. EPA 2007a). With ega d to the Roxbo o
site, a solid-wate pai analysis was attempted; howeve , Duke stated that "Roxbo o
is p ima ily uncle Iain by bed ock and fo this eason the e is insufficient data to
obse ve significant t ends." Duke also uncle stands that bed ock "is not likely to
so b la ge amounts of the key constituents" (Co ective Action Plan Pa t 2).
11

F gure 1. Example of a sol d -water pa r analys s to
demonstrate attenuat on (adapted from U.S. EPA 2007a)
Site -Specific
Sorption Trend
(A)
AttenuationAttenuation
(B)
NoNo Attenuation
Arpuaos (Ppm)
• Simila to the elationship analysis shown in Figu e 1, a p opo tionality constant (Kd)
can be dete mined in the labo ato y by allowing site soil samples to equilib ate with
wate . Kd is defined as the concent ation of the COI in the solid divided by the
concent ation of the COI dissolved in the wate . In a solid-wate pai analysis with a
positive linea t end, Kd would be the slope of the t end line. Kd values can give an
indication of the potential fo dissolved COIs in g oundwate to adso b onto soil,
which then can be elated to the ate at which a COI moves in elation to the ate of
g oundwate movement.
• With ega d to the Roxbo o site, Duke also pe fo med geochemical modeling
(Co ective Action Plan Pa t 2). The U.S. EPA ecommended that modeling esults
be "used as seconda y lines of evidence in suppo t of site-specific measu ements
that demonst ate active so ption of the contaminant onto aquife solids" and
cannot be elied upon independently (U.S. EPA 2007a).
VI. Op n on: Mon tored Natural Attenuat on s not an appropr ate remedy at the Roxboro
s te.
A. A numbe of constituents of inte est at the Roxbo o site a e gene ally poo Iy adso bed.
• In gene al, it is uncle stood that bo on is poo ly adso bed.
o Results of the labo ato y soil so ption expe iments indicate that "bo on so ption
is ve y low in Roxbo o site soil samples" (Co ective Action Plan Pa t 1 Appendix
D).
7

o With ega d to coal combustion p oduct leachate, MNA pathways do not exist
fo bo on (Mille 2011).
o Bo on was not effectively emoved f om coal ash leachate by adso ption/co-
p ecipitation with i on in a settling basin at a site in Pennsylvania (Rightnou and
Hoove 1998)
• F om my expe ience with othe coal ash sites (including othe Duke sites),
evaluation of wastewate t eatment p ocesses, and esea ch expe ience, the COls
sulfate, TDS, and antimony a e also gene ally poo ly adso bed.
B. The bed ock at the Roxbo o site and the p oximity of the ash basins to Hyco Lake is p oblematic
with ega d to attenuation by adso ption uncle site-specific conditions at the Roxbo o site.
• Duke uncle stands that "Roxbo o is p ima ily uncle Iain by bed ock" and that
bed ock "is not likely to so b la ge amounts of the key constituents" (Co ective
Action Plan Pa t 2). As long as the coal ash emains in place, this facto will emain
unchanged.
• Duke cites the close p oximity of the ash basins to Hyco Lake as a facto that actually
" educe[s] the potential fo impacted g oundwate to mig ate" (Co ective Action
Plan Pa t 2). This is clea ly inco ect. While g oundwate that ente s Hyco Lake may
be then to med "su face wate ", constituents ente ing the su face wate f om the
g oundwate may become diluted upon ente ing the su face wate , but will
continue to mig ate f om the site.
• Additionally, the e a e limited mate ials between the ash basins and Hyco Lake (due
to the close p oximity) onto which adso ption of Cols could potentially occu . These
limited mate ials have a finite capacity fo adso ption of COls. These mate ials have
al eady been in contact with COls leached f om coal ash fo seve al decades.
Adso bents utilized in the wate t eatment indust y a e changed on a egula basis
when thei capacity is exhausted.
C. Kd values dete mined in the labo ato v cannot be eliably utilized as a means to demonst ate
site-specific attenuation fo the Roxbo o site.
• Site-specific adso ption was not demonst ated in the soil-wate pai analysis
attempted by Duke (Co ective Action Plan Pa t 2).
• Kd values dete mined in the labo ato y a e not ep esentative of site-specific Kd
values fo the Roxbo o site. As stated above, Kd is a pa amete to estimate the
potential fo the adso ption of the COls in the g oundwate onto soil. Fo the
Roxbo o site, Kd values we a dete mined expe imentally in the labo ato y using
N

"synthetic g oundwate ", which was ve y limited in to ms of the numbe of
constituents and in many cases the concent ation of these constituents, and
the efo a was not ep esentative of g oundwate impacted by coal ash at the
Roxbo o site.
o Duke, in the Co ective Action Plan Pa t 1 (Appendix D) states that "a solution
intended to ep esent ae� ne is g oundwate is used as the backg ound to
which [contaminants of conce n] we a added" and that "this solution diffe s
f om the actual solution in g oundwate f om which the soil sample was
sampled".
• The Kd expe iments we a conducted uncle oxic (high oxygen) conditions. In
Co ective Action Plan Pa t 2 epo is fo othe sites (Allen, Belews C eek, Buck,
Cliffside, Ma shall), Duke stated that "typically Kd expe iments a e also conducted
uncle oxic [high oxygen] conditions, which may not eflect actual site conditions.
Consequently, Kd values f om labo ato y expe iments should be inte p eted and
used with caution because they may not account fo the full ange of so ption
conditions that occu in hete ogeneous, natu al soils."
• Additionally, the "synthetic g oundwate " did not include a numbe of competing
constituents that a e known to limit adso ption of COls. These include silica and
phosphate, which can significantly impact adso ption. Silica, in pa ticula , can be
p esent at ve y high concent ations at coal ash disposal sites.
• The efo e, the Kd values dete mined in the labo ato y we e likely biased high
because the "synthetic g oundwate " did not have othe constituents competing fo
the same adso ption sites.
D. When MNA is based upon adso otion of COls onto i on and manLyanese oxides. the utilization of
MNA in combination with capping is p oblematic as conditions fo MNA a e less favo able
following capping.
• Reducing infilt ation will educe the echa ge of oxygen into site g oundwate and
likely c eate mo a anoxic (less oxygen) conditions in the site g oundwate , which
can impact the mobility of constituents in the g oundwate (Schwa tz et al. 2016).
• The solubility of i on and manganese is st ongly influenced by oxidation- eduction
potential conditions, with solubility inc easing as the oxidation- eduction potential
is dec eased. Both a e eadily mobilized o esolubilized uncle anoxic conditions.
COls that had p eviously adso bed to i on o manganese will be emobilized and
potentially lead to a slug of COls ente ing the g oundwate . When i on and
manganese a e in the dissolved phase, COls cannot adso b to them. (Salminen
2005, Masscheleyn et al. 1991, Ha to et al. 2012, Schwa tz et al. 2016, U.S. EPA
2007b, Ta utis and Unz 1995).
O7

• Though i on is eadily oxidized when it comes into contact with oxygen, ae ation
(oxygen) is not effective fo oxidizing dissolved manganese (manganese in the +2
oxidation state) to the elatively insoluble +4 oxidation state. The efo e d inking
wate t eatment plants utilize a st ong oxidant such as ozone, chlo ine dioxide, o
potassium pe manganate when manganese emoval is equi ed. Additionally, the
p esence of dissolved i on in the g oundwate inhibits the fo mation of manganese
oxides. Only 14% manganese emoval was obse ved in a settling basin following
ae ation used to t eat coal ash leachate at a site in Pennsylvania (Hoffman et al.
2006, Ta utis and Unz 1995, Robinson -Lo a and B ennan 2011, Rightnou and
Hoove 1998). This is impo tant because the fo mation of manganese oxides has
been identified as a COI at the Roxbo o site.
• Infilt ation (and thus oxygen echa ge) would be educed ove the a ea of the
Roxbo o site uncle the cap, and thus i on and manganese mobility will be enhanced
(less i on and manganese oxide p ecipitation/COI adso ption occu ing within the
a ea that is capped). This inhe ently sho tens the distance to the site bounda y ove
which these MNA p ocesses can potentially occu , and has Duke has stated, the ash
basins a e in close p oximity to Hyco Lake.
E. MNA based upon adso ption is not an app op iate emedy fo a senic at the Roxbo o site.
• A senic in wate is typically p esent in the +3 oxidation state [As(III)] o the +5
oxidation state depending the oxidation- eduction potential (oxygen concent ation)
of the g oundwate . As(III) is favo ed uncle educed (low oxygen) conditions and
As(V) is favo ed uncle oxidized conditions. Compa ed to As(V), As(III) is mo a toxic,
mo a soluble, and mo e difficult to adso b (e.g. mo a mobile). The efo e the
leaching and mobility of a senic gene ally inc eases as edox potential dec eases
(Schwa tz et al. 2016, Ha to et al. 2012, Masscheleyn et al. 1991). Duke
uncle stands that a senic is gene ally less mobile uncle oxidizing conditions
(Co ective Action Plan Pa t 2 Appendix C).
• Fo a senic in g oundwate at the Roxbo o site, As(III) is the p edominant species of
a senic measu ed in site g oundwate samples (Co ective Action Plan Pa t 2
Appendix C).
• Capping the site will make conditions less favo able fo MNA of a senic. Capping of
the site will lead to mo a anoxic conditions in the site g oundwate (Schwa tz et al.
2016), which would likely lead to an inc ease in the atio of As(III) to As(V); thus the
addition of a cap would likely inc ease the mobility of a senic at the Roxbo o site.
• Conce n ove a eduction in site g oundwate oxidation- eduction potential and
enhancement of a senic solubility as a esult of closu a plans fo coal ash
impoundments that include capping, led a g oup of scientists to ecently state that
10

"even if no p evious g oundwate contamination issues have been epo ted,
capping methods that might induce anae obic conditions should be avoided in the
closu a of unlined impoundments" (Schwa tz et al. 2016).
11

VII. References
Abe nethy, RF; MJ Pete son and FH Gibson (1969). Spect ochemical analyses of coal ash fo t ace
elements. United States Depa tment of the Inte io , Repo t of Investigations 7281.
Blackmo e, DPT; J Ellis and PJ Riley (1996). T eatment of a vanadium -containing effluent by
adso ption/cop ecipitation with i on oxyhyd oxide. Wate Resea ch 30(10): 2512-6.
Chen, W; R Pa ette, J Zou, FS Cannon and BA Dempsey (2007). A senic emoval by i on modified
activated ca bon. Wate Resea ch 41: 1851-8.
Comp ehensive Site Assessment Repo t. Roxbo o Steam Elect is Plant. Fo Duke Ene gy P og ess,
p epa ed by SynTe a. Septembe 2, 2015.
Co ective Action Plan Pa t 2. Roxbo o Steam Elect is Plant. Fo Duke Ene gy P og ess, p epa ed by
SynTe a. Feb ua y 29, 2016.
Co ective Action Plan Pa t 1. Roxbo o Steam Elect is Plant. Fo Duke Ene gy P og ess, p epa ed by
SynTe a. Decembe 1, 2015.
Davis, CC; WR Knocke and M Edwa ds (2001). Implications of aqueous silica so ption to i on
hyd oxide: mobilization of i on colloids and inte fe ence with so ption of a senate and humic
substances. Envi onmental Science & Technology 35: 3158-62.
Duke Ene gy (2015). Duke Ene gy Ash Met ics. https://www.duke-ene gy.com/pdfs/duke-ene gy-
ash-met ics.pdf
Elect is Powe Resea ch Institute (2009). Coal Ash: Cha acte istics, Management, and Envi onmental
Issues. www.whitehouse.gov/sites/default/files/omb/assets/oi a_2050/2050_meeting_101609-
2.pdf
Ha te, PT; JD Ayotte, A Hoffman, KM Revesz, S Lamb and JK Bohlke (2012). Hete ogeneous edox
conditions, a senic mobility, and g oundwate flow in a f actu ed- ock aquife nea a waste
eposito y site in New Hampshi e, USA. Hyd ogeology Jou nal 20(6): 1189-1201.
Hasany, QM and MA Qu eshi (1981). Adso ption studies of cobalt (II) on manganese dioxide f om
aqueous solutions. Inte national Jou nal of Applied Radiation and Isotopes 32: 747-52.
Hoffman, GL; DA Lytle, TJ So g, ASC Chen and L Wang (2006). Design Manual: Removal of A senic
f om D inking Wate Supplies by I on Removal P ocess. Ap it 2006. EPA/600/R-06/030.
Holm, TR; WR Kelly, SD Wilson and JL Talbott (2008). A senic emoval at Illinois i on emoval plants.
Jou nal AWWA 100(9): 139-150.
12

Holm, TR (2002). Effects of C032-/bica bonate, Si, and PO43-on a senic so ption to HFO. Jou nal
AWWA 94(4):174-81.
Liu, G; H Zhang, L Gao, L Zheng and Z Peng (2004). Pet ological and mine alogical cha acte izations
and chemical compositions of coal ashes f om powe plants in Yanzhou mining dist ict, China. Fuel
P ocessing Technology 85: 1635-46.
Lokeshappa, B; AK Dikshit, Y Luo, TJ Hutchinson and DE Giamma (2014). Assessing bioaccessible
f actions of a senic, ch omium, lead, selenium, and zinc in coal fly ashes. Inte national Jou nal of
Envi onmental Science and Technology 11: 1601-10.
Ma iussen, E: IV Johnsen and AE St omseng (2015). Selective adso ption of lead, coppe , and
antimony in unoff wate f om a small a ms shooting nage with a combination of cha coal and i on
hyd oxide. Jou nal of Envi onmental Management 150: 281-7.
Masscheleyn, PH; RD Delaune and WH Pat ick J . (1991). A senic and selenium chemist y as affected
by sediment edox potential and pH. Jou nal of Envi onmental Quality 20: 522-7.
Meng, X; GP Ko fiatis, S Bang and KW Bang (2002). Combined effects of anions on a senic emoval
by i on hyd oxides. Toxicology Lette s 133: 103-11.
Meng, X; S Bang and GP Ko fiatis (2000). Effects of silicate, sulfate, and ca bonate on a senic
emoval by fe is chlo ide. Wate Resea ch 34(4): 1255-61.
Mille , GP (2011). Monito ed natu al attenuation: a emediation st ategy fo g oundwate impacted
by coal combustion p oduct leachate. 2011 Wo Id of Coal Ash Confe ence. May 9-12, 2011. Denve ,
CO.
Mille , GP (2001). Su face complexation modeling of a senic in natu al wate and sediment systems.
Ph.D. disse tation, New Mexico Institute of Mining and Technology.
Mukhe jee, J; J Ramkuma , S Chand amoukeeswa an, R Shukla and AK Tyagi (2013). So ption
cha acte istics of nano manganese oxide: efficient so bent fo emoval of metal ions f om aqueous
st earns. Jou nal of Radioanalytical and Nuclea Chemist y 297: 49-57.
Rightnou , TA and KL Hoove (1998). The Sp ingdale P oject: Applying Const ucted Wetland
T eatment to Coal Combustion By -P oduct Leachate. EPRI epo t TR -111473.
Robinson -Lo a, MA and RA B ennan (2011). Anae obic p ecipitation of manganese and co -existing
metals in mine impacted wate t eated with c ab shell -associated mine als. Applied Geochemist y
26: 853-62.
13

Rocca o, P and FGA Vagliasindi (2015). Cop ecipitation of vanadium with i on(III) in d inking wate : a
pilot -scale study. Desalination and Wate T eatment 55: 799-809.
Salminen, R (2005). Geochemical Atlas of Eu ope, Pa t 1— Backg ound Info mation, Methodology,
and Maps. weppi.gtk.fi/publ/fo egsatlas/text.
Sa ka , D; DK De, R Das and B Mandal (2014). Removal of o ganic matte and oxides of i on and
manganese f om soil influences bo on adso ption in soil. Geode ma 214-5: 213-6.
Schwa tz, GE; N Rive a, SW Lee, JM Ha ington, JC Howe , KE Levine, A Vengosh and H Hsu -Kim
(2016). Leaching potential and edox t ansfo mations of a senic and selenium in sediment
mic ocosms with fly ash. Applied Geochemist y 67: 177-85.
Swedlund, PJ and JG Webste (1999). Adso ption and polyme ization of silicic acid on fe ihyd ide,
and its effect on a senic adso ption. Wate Resea ch 33(16): 3413-22.
Ta utis, WJ and RF Unz (1995). 1 on and manganese elease in coal mine d ainage wetland
mic ocosms. Wate Science & Technology 32(3): 187-92.
U.S. EPA (2007a). Monito ed Natu al Attenuation of Ino ganic Contaminants in G ound Wate ,
Volume 1, Technical Basis of Assessment. EPA/600/R-03/139.
U.S. EPA (2007b). Monito ed Natu al Attenuation of Ino ganic Contaminants in G ound Wate ,
Volume 2, Assessment fo Non -Radionuclides Including A senic, Cadmium, Ch omium, Coppe , Lead,
Nickel, Nit ate, Pe chlo ate, and Selenium. EPA/600/R-03/140.
USEPA (2012, Septembe ). A Citizens Guide to Monito ed Natu al Attenuation. Ret ieved f om
https://clu-in.o g/
U.S. EPA (2016). Coal Ash Basics. Janua y 15, 2016. https://www.epa.gov/coalash/coal-ash-
basics#02
14

APPENDIX A
D . Robe t Pa ette CV
15

ROBERT PARETTE
Matson & Assoc ates, Inc.
331 East Foster Avenue, State College, PA 16801
Phone: [phone] Ema I: bparette@matson-assoc ates.com
EDUCATIO
Doctor of Philosophy August 2005
Env ronmental Eng neer ng
The Pennsylvan a State Un vers ty, Un vers ty Park, PA
D ssertat on: The Removal of Perchlorate from Groundwater v a Quaternary
Ammon um Cat on c Surfactant Pre -Loaded GAC
Master of Science May 2004
Env ronmental Eng neer ng
The Pennsylvan a State Un vers ty, Un vers ty Park, PA
Thes s: GAC Ta lored w th Organ c Cat ons for the Removal of Perchlorate from
Groundwater and Base/Ac d Chem cal Regenerat on of the GAC
Bachelor of Science February 1999
Chem cal Eng neer ng
Worcester Polytechn c Inst tute, Worcester, MA
WORK EXPERIE CE
Matson and Associates, Inc., State College, PA
Env ronmental Eng neer (P.E.)/Env ronmental Chem st (2009-2014)
V ce Pres dent of Eng neer ng/Project Manager for Matson B ofuels (2009-2011)
The Pennsylvania State University, Dept. of Civil and Environmental Engineering
Post -Doctoral Research Assoc ate (2005-2009)
Graduate Research Ass stant (2000-2005)
Environmental Consultant
Ass stant to Penn State professor sery ng as an expert w tness (2007)
Hercules Incorporated Research Center, Wilmington, DE
Research Ass stant (1998)
REPRESE TATIVE PROJECT EXPERIE CE
• Rev ewed closure plan alternat ves for a coal ash mpoundment located n Conway, South
Carol na. Prepared wr tten comments address ng the lack of groundwater treatment (as part of
these closure plan alternat ves) and the feas b I ty of remed at ng the arsen c contam nated
groundwater.
A-1

• Evaluated the d sposal pract ces of waste salt conta n ng arsen c at a fac I ty along the
Menom nee R ver n W scons n n relat on to waste d sposal standards and the alternat ve
d sposal opt ons ava fable. Invest gated the spec at on of arsen c n the waste salts, the
propert es of the var ous arsen c spec es, and the transformat ons these arsen c spec es would
undergo n the env ronment.
• At Penn State, 9% years of R&D on technolog es to remed ate ground water. Th s R&D resulted
n numerous publ cat ons (I sted below) on the removal of arsen c and perchlorate from water.
Des gned and produced custom adsorbents for the removal of arsen c from groundwater.
Des gned and fabr cated a p lot plant w th a capac ty n excess of one m II on gallons per year to
study the removal of arsen c by a comb nat on of act vated carbon and ron from a var ety of
d fferent sources. Respons ble for develop ng exper mental and sampl ng plans for a number of
p lot and laboratory stud es.
• Evaluated treatment technolog es, and the feas b I ty of employ ng such technolog es, n
connect on w th remed at ng a large lake contam nated w th selen um. Prepared a report to be
ut I zed for publ c comments and negot at ons w th the power company.
• Evaluated the potent al mplementat on of var ous technolog es for a number of outfalls at a
coal m n ng operat on n western Pennsylvan a. Prepared wr tten comments address ng
shortcom ngs of a consent decree perta n ng to the s te.
• Evaluated wastewater treatment pract ces of a paper m II along the Fox R ver n W scons n n
relat on to the treatment opt ons ava fable (chronolog cally) to the pulp and paper ndustry and
the removal eff c enc es ach eved by var ous treatment technolog es across the ndustry.
• Des gned and conducted laboratory exper ments to test the mpact of MTBE on the removal of
other organ c contam nants from groundwater for a case nvolv ng MTBE n Suffolk County, NY.
Ut I zed mult ple f eld data sets to assess the ab I ty of act vated carbon to remove MTBE,
ass sted n the preparat on of expert w tness reports, and performed an extens ve I terature
rev ew pert nent to the case.
• Correlated s to act v t es at an abandoned chem cal manufactur ng plant n NJ w th
contam nants found n the near offshore areas of the s te; determ ned the or g n and chem cal
react ons beh nd the presence of 2,4,6,8-tetrachlorod benzoth ophene (2,4,6,8-TCDT), a
contam nant whose source was prev ously unknown.
• Evaluated the s ngle phase eff c ency for the str pp ng of CO2 n a cool ng tower model for an
nternat onal company that spec al zes n gases. Mod f ed the model's source code to reflect the
determ ned s ngle stage eff c ency, evaluated the I qu d to gas rat o ut I zed by the model, and
compared results from the model aga nst the exper mental data.
• Invest gated groundwater qual ty and consulted w th a forester at a res dent al s to n Centre
County, PA, mpacted by a nearby hous ng development.
• D rected the R&D efforts for Matson B ofuels from 2009 through 2011. Respons ble for the
management of three b od esel projects, w th grant fund ng n excess of $250K. Developed sol d
A-2

catalysts capable of convert ng low qual ty o I feedstocks to b od esel w th no feedstock
pretreatment and no soap format on. Des gned a p lot plant, w th a capac ty to produce more
than 1 m II on gallons of b od esel annually.
PUBLICATIO 5
R Parette, R McCr ndle, KS McMahon, M Pena-Abaurrea, E Re ner, B Ch tt m, N R ddell, G Voss, FL
Dorman, WN Pearson and M Robson (2016). Response to the comment on "Halogenated nd go dyes
I kely source of 1,3,6,8-tetrabromocarbazole and some other halogenated carbazoles n the
env ronment". Chemosphere 150: 414-5.
N R ddell, UH J n, S Safe, Y Cheng, B Ch tt m, A Konstant nov, R Parette, M Pena-Abaurrea, EJ Re ner, D
Po r er, T Stefanac, AJ McAlees and R McCr ndle (2015). Character zat on and b olog cal potency of
mono- to tetra -halogenated carbazoles. Env ronmental Sc ence & Technology 49 (17): 10658-66.
R Parette, R McCr ndle, KS McMahon, M Pena-Abaurrea, E Re ner, B Ch tt m, N R ddell, G Voss, FL
Dorman and WN Pearson (2015). Halogenated nd go dyes: a I kely source of 1,3,6,8-
tetrabromocarbazole and some other halogenated carbazoles n the env ronment. Chemosphere 127:
18-26.
R Parette and WN Pearson (2014). 2,4,6,8-Tetrachlorod benzoth ophene n the Newark Bay Estuary: the
I kely source and react on pathways. Chemosphere 111: 157-63.
W Chen, R Parette and FS Cannon (2012). P lot -scale stud es of arsen c removal w th granular act vated
carbon and zero -valent ron. Env ronmental Eng neer ng Sc ence 29(9): 897-901.
K Gombotz, R Parette, G Aust c, D Kannan and J Matson (2012). MnO and T O sol d catalysts w th low-
grade feedstocks for b od esel product on. Fuel 92(1): 9-15.
MK Sel em, S Komarnen, R Parette, H Katsuk, FS Cannon, MG Shah en, AA Khal I and IM Abdel -Ga d
(2011). Perchlorate uptake by organos I cas, organo -clay m nerals and compos tes of r ce husk w th
MCM-48. Appl ed Clay Sc ence 53: 621-6.
JP Patterson, R Parette and FS Cannon (2011). Compet t on of an ons w th perchlorate for exchange s tes
on cat on c surfactant-ta lored GAC. Env ronmental Eng neer ng Sc ence 28(4): 249-56.
JY K m, S Komarnen , R Parette and FS Cannon (2011). Perchlorate uptake by synthet c layered double
hydrox des and organoclays. Appl ed Clay Sc ence 51(1-2): 158-64.
JP Patterson, R Parette and FS Cannon (2010). Ox dat on of ntermed ate sulfur spec es (th osulfate) by
free chlor ne to ncrease the bed I fe of to lored GAC to remove perchlorate. Env ronmental Eng neer ng
Sc ence 27(10): 835-43.
MK Sel em, S Komarnen, R Parette, H Katsuk, FS Cannon, MG Shah en, AA Khal I and IM Abdel -Ga d
(2010). Compos tes of MCM-41 s I ca w th r ce husk: Hydrothermal Synthes s, character zat on and
appl cat on for perchlorate separat on. Mater als Research Innovat ons 14(5): 351-4.
A-3

S Komarnen , J Y K m, R Parette and FS Cannon (2010). As -Synthes zed MCM-41 s I ca: New adsorbent for
perchlorate. Journal of Porous Mater als 17: 651-6.
MJ Jang, FS Cannon, R Parette, S Yoon and W Chen (2009). Comb ned hydrous ferr c ox de and
quaternary ammon um surfactant to for ng of granular act vated carbon for concurrent arsen c and
perchlorate removal. Water Research 43(12): 3133-43.
W Chen, R Parette and FS Cannon FS (2008). Arsen c adsorpt on v a ron preloaded act vated carbon and
zero -valent ron. Journal of Amer can Water Works Assoc at on 100(8): 96-105.
W Chen, R Parette, J Zou, FS Cannon and BA Dempsey (2007). Arsen c removal by ron-mod f ed
act vated carbon. Water Research 41(9): 1851-8.
R Parette and FS Cannon (2006). Perchlorate removal by mod f ed act vated carbon. In Perchlorate
Env ronmental Occurrence, Chem stry, Tox cology, and Remed at on Technolog es. Spr nger Publ sh ng
Company, New York.
R Parette and FS Cannon (2005). The removal of perchlorate from groundwater by act vated carbon
to lored w th cat on c surfactants. Water Research 39(16): 4020-28.
R Parette, FS Cannon and K Weeks (2005). Remov ng low ppb level perchlorate, RDX, and HMX from
groundwater w th cetyltr methylammon um chlor de (CTAC) pre -loaded act vated carbon. Water
Research 39(19): 4683-92.
A-4