NC0004987_Geochem Memo and Summary_20200320

TECHNICAL MEMO
To:
Scott Davies, PG, Duke Energy
526 South Church Street
Charlotte, North Carolina 28202
From:
Julie K Sueker, PhD, PH, PE (CO)
Margy Gentile, PhD, PE (CA)
Date:
March 20, 2020
Arcadis Project No.:
30012657
AARCAD IS Design &Consultancy
fornaturaland
built assets
Subject:
Summary of Geochemical Modeling Approach — Marshall Steam Station
Arcadis U.S., Inc.
11400 Parkside Drive
Suite 410
Knoxville
Tennessee 37934
Tel [phone]
Fax [phone]
Duke Energy was required to model potential geochemical effects related to ash basin decanting and ash
basin closure on the transport of constituents of interest (COls) in groundwater at the Marshall Steam
Station (Marshall or site; Figure 1). SynTerra, in collaboration with others, generated the geochemical
model for Marshall (SynTerra 2019a).
The objectives of the modeling' were to demonstrate an understanding of COI geochemical behavior,
describe source terms in the model, to simulate downgradient concentrations of COI at various stages of
closure, and to provide a basis for translating between detailed geochemical modeling and the sitewide
flow and transport model. Site -specific data incorporated into the modeling included COI concentrations
and trends in ash pore water and groundwater, solid phase mineralogy for estimates of sorption and ion
exchange sites, COI leaching behavior, and hydrogeologic information. Modeling analysis included
overviews of groundwater data, geochemical evaluations of ash leaching data2, batch PHREEQC3 models
and sorption coefficient derivations, and PHREEQC 1-D advection models.
KEY FINDINGS
The key findings of the geochemical modeling effort associated with the selected closure scenario
(closure -by -excavation for the ash basin) are listed below:
The framework was developed through collaboration with NCDEQ, William Deutsch (external reviewer for
NCDEQ), and the flow and transport (F&T) modeling team (CAP Update -Appendix G, Synterra 2019b)
over many meetings, presentations, and conference calls (Duke 2017a, Duke 2017b).
2 Via USEPA Method 3052 (1996) and USEPA LEAF Method 1313 (2012a) and 1316 (2012b).
3 PHREEQC- original acronym pH-REdox-EQuilibrium written in C programming language.
arcadis.com
Page:
1/2

MEMO
1. Closure activities are anticipated to minimize groundwater flow through the ash basin and maximize
the input of upgradient unaffected groundwater, resulting in decreased downgradient COI
concentrations.
2. The pH and redox potential (EH) remain stable to maintain sorption as a dominant attenuation
mechanism for most nonconservative COls.
3. Closure activities that generate extreme pH values (generally less than 4 and greater than 10) may
cause increased mobility locally of certain COls. In particular, the oxidation of pyrite, which can
produce low pH conditions, has the potential to mobilize COls.
4. Increased EH values that may be generated from oxygen infiltration during decanting or other closure
activities, such as closure -by -excavation, will not cause enhanced mobility of most COls. An increase
in EH would make ferrihydrite more stable, resulting in more sorption sites. Potential exceptions might
be enhanced mobility of hexavalent chromium or pentavalent arsenic if EH conditions allow such
species to persist, although these species were not identified as COls for the site.
COI Evaluation
At the Marshall site, 17 COls exhibit mean concentrations greater than background threshold
values (BTVs), 02L standards, or interim allowable maximum concentrations (IMACs) at or beyond the
ash basin geographic limitation: antimony (Sb), barium, boron (B), beryllium (Be), chloride, cobalt (Co),
iron (Fe), lithium (Li), manganese (Mn), molybdenum (Mo), selenium, strontium, sulfate (SO4 2),
thallium (TI), total dissolved solids, total radium (Ra), and vanadium (V). Results from site -specific partition
coefficient (Kd) values evaluations are as follows:
Nonconservative, reactive COls: Kd values for Be, Sr, V and other nonconservative, reactive COI
remained high in most cases, and are unlikely to be affected geochemically by remedial actions
based on Kd evaluation (values remained high for tested scenarios in most cases).
• Conservative, nonreactive COI: Kd values for Sb, B, Li, and SO4 2 were low (less than 1 liter per
kilogram) for all modeled scenarios and will not change significantly due to changes related to
closure.
• Variably reactive COls: Kd values for Co, Fe, Mn, Mo, Ra, and TI were greatly variable in relation to
geochemical changes and dependent on the pH and EH.
References
Duke Energy. 2017a. DWR-ARO Meeting to Discuss Asheville Models. Asheville, North Carolina: NCDEQ.
Duke Energy. 2017b. NCDEQ Meeting to Review Cliffside Models and CSAs. Asheville, North Carolina:
NCDEQ. October 11.
SynTerra. 2019a. CAP Update- Appendix H. Geochemical Model Report. In Corrective Action Plan Update.
Terrell, North Carolina.
SynTerra. 2019b. CAP Update -Appendix G. In Corrective Action Plan Update. Terrell, North Carolina.
USEPA. 1996. Method 3052: Microwave assisted acid digestion of siliceous and organically based matrices -
Revision 0. SW-846. USEPA. December.
USEPA. 2012a. Method 1313 - Liquid -solid partitioning as a function of extract pH using parallel batch
extraction procedure. Test methods for evaluating solid waste: Physical/chemical methods. SW-846, 3rd.
USEPA. October.
USEPA. 2012b. Method 1316 - Liquid -solid partitioning as a function of liquid -to -solid ratio in solid materials
using a parallel batch procedure. Test methods for evaluating solid waste: Physical/chemical methods.
SW-846, 3rd. USEPA. October.
arcadis.com Page:
2/2

GRAPHIC SCALE1 �• } I
CATAWBA Laao o Loco 2000 - 1 �✓/
COUNTY
WNsrarvsALEM ON FEET)
ASHEVILLE.
IL
1, illi r INDUSTRIAL'/�
LANDFILL #1---
` 151AND POW WD 1 ■ LANDFILL COMPLIANCES
�1LF STRUCTi.1YAL r BOUNDARY �-
lfj� pI FILL ■ {
Ili •—}' _
f11AR`: HALLS3EAM STATION , j'., C&/DLANCC-ILL ' n ` =J •.�
' PARCEL LINE
_I _ DRY ASH LAN'i�-ILL
ASBESTOS rt1r� y.a �' r [PHASE 14 { }
I _ LANDFILL r
° 11 MASH BASIN GEOGRAPHIC'�_J
[? i I
LIMITATION
_,
AaH PA91M 2� z
FF`JS UCTURAL 'OVASTE BOUNDARY,
t. FILL
LAILJFY-L COMPL'aANI =
r ^ •� BOUNDARY
BCQROWAREA;
(` �.. DRY ASH LANDFILL ✓ - {
` ACCESS RCI AD
STRUCTURAL FILLASH BASIN
LANDFILL COMPLIANCE '- r+'4,
BOUNDARY +oo
-'. POD RESIDUE
LANDFILL BOUNDARY . t a' . 0 •
riAlq�HOLDING BASIN t
`- LRB
GYPSUM , MARSHALLC
STEAM STATION
MAW _ _
1V
J�40o
SOURCE: ,T
2416 USGn TOPOGRAPHrMAP, TROUTMAN & LAKE NORMAN
NORTH QUADRANGLE, OBTAINED FROM THE USGS
STCRE AT https:/Istore.usgs.govtmap-locator. ,n v
SUMMARY OF
GEOCHEMICAL MODELING /A�'AJ FIGURE
APPROACH - MARSHALL SITE MAP VAARG%V IJ buitas5etsntl
h u I It assets
STEAM STATION