NC0004961_Evaluation of the Effect of Ash Disposal_19871207

1551 t .St ((
EVALUATION OF THE EFFECT OF ASH DISPOSAL
AT THE RIVERBEND PLANT OF
DUKE POWER COMPANY ON
GROUNDWATER AND SURFACE -WATER QUALITY
Prepared for
Duke Power Company
Prepared by
Kilkelly Environmental Associates
Ralph Heath
Miquel Medina
Harry LeGrand
Jonathan Butcher
Clayton Creager
Report No. I308 -12/7/87-01F

EXECUTIVE SUMMARY
In April 1985, Duke Power Company applied for a permit to construct a basin at its
Riverbend Plant in Gaston County, North Carolina, in which ash dredged from its
settling pond would be placed for drying. In response to the Company's request, the
North Carolina Division of Environmental Management (DEM) prepared a permit
requiring (1) an assessment to determine the existing groundwater quality in the
immediate vicinity of the dredge pond, (2) Duke Power to determine the effect of the
groundwater discharge from the ash ponds upon the Catawba River, and (3) a proposal
for a monitoring well network sufficient to detect any contaminants which could reach
the river.
On the basis of three independent investigations at the Allen Plant and studies
conducted by the EPA and other agencies at other coal-fired plants, Duke Power does
not believe a groundwater monitoring program is needed at the Riverbend Plant.
Kilkelly Environmental Associates conducted a comparative hydrogeologic study of the
Allen and Riverbend plants to estimate the input of ash trace elements to the Catawba
River through both surface runoff and groundwater, and to make a preliminary
assessment of the risk posed to health and the environment by groundwater
contaminants from the Riverbend ash pond reaching the Catawba River. The major
findings of this study are listed below:
��• The Allen and Riverbend coal-fired, electric -generating plants of Duke
Power Company are located 12 miles apart on the west bank of the Catawba
River in Gaston County, North Carolina.
�1 • Both plant sites are underlain by bedrock composed of granite and diorite of
the Charlotte Belt and by saprolite derived from the chemical and physical
breakdown of these rocks. Diorite, which forms an especially clay -rich
saprolite, appears to be somewhat more prevalent at the Allen Plant than at
the Riverbend Plant. However, the bedrock and the saprolite derived from it
are similar enough in mineral composition at both sites to permit hydrologic
and chemical data collected at one plant to be used in conjunction with
hydrogeologic data from the other plant.

V • Three intensive hydrogeologic and geochemical studies were conducted at
the Allen Plant in the early 1980's to determine the extent to which seepage
from its inactive and active ash ponds was affecting groundwater quality.
Relying on the similarity in the hydrogeologic conditions at the plants,
geochemical data from the Allen Plant was used in this report, in conjunction
with hydrogeologic data from the Riverbend Plant, to determine the effect
of ash -pond effluent on groundwater quality at the Riverbend Plant.
4J • Chemical analyses of water samples, both from ash and from saprolite at the
Allen Plant, show that metals, which are present in large concentrations in
the ash, are relatively insoluble with the result that ash pond effluent
contains only very small concentrations of most metals.
5) • The saprolite at both the Allen and Riverbend Plants has a large capacity,
through ion exchange and precipitation, to immobilize the metals contained
in the ash effluent. At Allen, calcium and strontium, which are among the
more mobile of the metal constituents in the effluent, move through the
upper, clay -rich part of the saprolite at a rate of only about 0.3 ft/yr. The
rate of movement at Riverbend may be somewhat faster, due to the smaller
percentage of clay in the saprolite at that site, but is believed to be
substantially less than 1 ft/yr. Thus, in the 30 years of operation of the
Riverbend ash pond, the more mobile metals are believed to have moved less
than 20 feet into the saprolite.
6) • Boring data at the Riverbend Plant indicate that the thickness of saprolite
beneath the ash pond ranges from about 60 to 100 ft. Thus the "advancing
fronts" of the most mobile metals derived from the ash are still well above
the bedrock surface.
�/• Relative to the effect of ash -pond effluent on the quality of the Catawba
River, analysis of streamflow records shows an average flow past the
Riverbend Plant of about 2866 cfs. Effluent from the pond reaches the river
both through surface outflow and through the groundwater system. The rate
of surface -outflow averages about 7.1 cfs and the rate of groundwater
outflow is estimated to be about 0.7 cfs, or one-tenth e rate of surface
outflow. Thtis,�t�le average flow of the river is about 400 times the rate of
surface outflow from the pond and about 370 times the combined surface and
groundwater rate.
`� • Although the velocity of the Catawba River is relatively slow past the
Riverbend site, modeling using conservative mixing coefficients suggests
that complete transverse mixing of the pond effluent occurs within about
three miles of the ash pond. Due to the small concentration of metals in the
surface outflow from the pond, and the large dilution factor, metals
contained in the effluent do not cause a detectable increase in concentration
once complete mixing has occurred.
%� • Relative to the effect of groundwater seepage on stream quality, modeling
of flow through the groundwater system using a retardation factor only 1/3
of the value estimated at the Allen Plant shows that no measurable
concentration of the metals that are subject to ion exchange and other
delaying reaction will reach the Catawba within the next 50 years (by 2037).

TABLE OF CONTENTS
SECTION DESCRIPTION PAGE
1. INTRODUCTION ...................................... 1
2. HYDROGEOLOGIC CONDITIONS AT ALLEN ................. 3
AND RIVERBEND PLANTS
2.1 General Hydrogeologic Conditions ..................... 3
2.2 Hydrogeologic Conditions at Riverbend Plant ............. 7
3. EFFECT OF RIVERBEND ASH POND ON GROUNDWATER ...... 17
3.1
Hydraulic Conductivity of Ash and Saprolite .............
18
3.2
Transmissivity of Saprolite and Bedrock ................
19
3.3
Hydraulic Gradients in the Saprolite and Bedrock ..........
23
3.4
Groundwater Outflow from the Ash Pond ...............
25
3.5
Groundwater Velocities and Time of Travel ..............
27
3.6
Effect of Pond Seepage on Groundwater Quality ..........
29
4. GENERAL SURFACE WATER QUANTITY AND QUALITY ....... 47
5. ANALYSIS OF POTENTIAL SURFACE WATER IMPACTS ........ 80
5.1 Vertical and Transverse Mixing ....................... 80
5.2
Analytical, Steady -State, Continuous Two -Dimensional
(Vertical Line Source) Model, SSCLS ..................
86
6. ANALYSIS OF POTENTIAL GROUNDWATER CONTRIBUTIONS ...
91
6.1
Potential Impact on Groundwater .....................
91
6.2
Potential Groundwater Contribution of Contaminants to
Catawba River/Mountain Island Lake ................
104
6.3
Comparison of Groundwater and Surface Contributions ....
105
6.4
Conclusions ....................................
106
ii

LIST OF FIGURES
FIGURE
NUMBER DESCRIPTION PAGE
1. Diagrammatic Sketch Showing the Relation Between Diorite and ... 4
Granite in the Charlotte Belt
Z. Location and Topography of the Riverbend Ash Pond Area ........ 8
3. Map of Riverbend Ash Pond Area Showing Segments Used to ...... 10
Calculate Ground -water Outflow from Pond, Ground -water
Divides, Locations of Hydrogeologic Cross Sections and
Boundary of Outflow Plume
4. Map of Riverbend Ash Pond Area Showing Land -Surface ......... 13
Topography Prior to Development of Ash Pond
5. Map of Riverbend Ash Pond Area Showing the Approximate ...... 14
Configuration of the Bedrock Surface
6. Map of Riverbend Ash Pond Area Showing Estimated ........... 15
Thickness of Saprolite Based on Boring Data
Supplied by Duke Power Co.
7. Map or Riverbend Ash Pond Area Showing the Approximate ...... 16
Altitude of the Water Table in the Saprolite and Dikes
8. Hydrogeologic Cross Section Along Line D -D' in Figure 3 ........ 20
9. Hydrogeologic Cross Sections Along Lines E -E" and F -F" in Figure 3 21
10. Estimated Times of Travel in Years Along Hypothetical ........ 30
Flowlines From the Riverbend Ash Pond to the
Catawba River
11. Map of the Ash Pond Area at the Allen Plant of Duke Power
Company Showing the Locations of Water -Quality Sampling
Wells and Other Features ............................... 38
12. Concentrations of Selected Metals Versus Length of Flowlines
in Saprolite at the Active Allen Powerplant Ash Pond of
Duke Power Company ................................. 39
13. Sections Showing Materials Penetrated and Positions of
Screens in Water -Quality Sampling Wells Drilled Through
the Ash in Both the Inactive and Active Ash Ponds at the
Allen Powerplant of Duke Power Company .................. 40
14. Geologic Section Extending from the Active Ash Pond at
the Allen Powerplant of Duke Power Company to the
CatawbaRiver ...................................... 41
iii

LIST OF FIGURES
(continued)
FIGURE
NUMBER
DESCRIPTION PAGE
15.
Concentration of Calcium and Strontium Versus Years
Per Foot for Wells Open to Ash and to Saprolite Beneath
the Ash Ponds at the Allen Powerplant of Duke
Power Company ......................................
45
16.
Mountain Island Lake Water Quality Monitoring Stations .........
50
17.
Map of Mountain Island Reservoir Study Area ................
51
18.
Profile of Iron Concentrations in Catawba River ..............
63
19.
Ash Basin Flow ......................................
64
20.
Time History of Effluent pH .............................
66
21.
Ash Basin Flow Versus pH ...............................
68
22.
Time History of pH Upstream and Downstream of Site ..........
69
23.
Time History of Effluent Temperature .....................
70
24.
Iron Concentrations at Hicks, NC .........................
72
25.
Iron Concentrations Upstream and Downstream of Site ..........
73
M.
Time History of Effluent Iron Concentrations ................
74
27.
Iron Concentration at Thrift, NC .........................
75
28.
Manganese Concentrations Upstream (Sta 278) and
Downstream (Sta 277) of Site ............................
76
29.
Time History of Effluent Zinc Concentrations ................
77
30.
Time Series of Effluent Selenium Concentrations ..............
78
31.
Time Series of Effluent Nickel Concentrations ................
79
32.
River Cross -Sections Through Mile 7 .......................
83
33.
River Cross -Sections at Miles 8 and 9 ................... • • •
84
34.
River Cross -Sections at Miles 10 and 11 ....................
85
35.
Concentration Ratio Profile Downstream from Site ............
89
1v

LIST OF FIGURES
(continued)
FIGURE
NUMBER DESCRIPTION PAGE
36. Computed Profile of Concentration Ratio, Downstream
of Ash Basin ........................................ 90
37. Three -Dimensional View of Percent Concentration ............. 97
38. State of North Carolina Groundwater Advisory
System, Output File ................................... 98
v

LIST OF TABLES
TABLE
NUMBER DESCRIPTION PAGE
1. Hydraulic Conductivity Data from the Allen Plant ............. 18
2. Thickness and Transmissivity of the Saprolite for the ........... 22
Outflow Segments Shown in Figure 3
3. Hydraulic Gradients for Saprolite and Bedrock for the .......... 24
Outflow Segments Shown in Figure 3
4. Groundwater Outflow from the Ash Pond .................... 26
5. Estimated Groundwater Velocities in Saprolite and Bedrock ...... 28
in the Riverbend Ash Pond Area
6. Concentration of Selected Metals in Leachate Extracted From ... 31
Ash Samples Through the Use of the EPA Extraction Procedure
and EPA Toxicity Criterion Limits for Solid Wastes Under the
Resource Conservation and Recovery Act.
7. Selected Chemical Analyses From the Allen Plant Related to .... 33
Ash -pond Seepage.
8. Basic Data Used in the Preparation of Figure 15 .............. 44
9. Statistical Summary of Flows Through Cowans Ford and
Mtn. Island Dams ..................................... 52
10.
Statistical Summary of Catawba River Water Quality at
Hicks, N.C., STORET Station No. 2142648 ...................
53
11.
Statistical Summary of Catawba River Water Quality at
Station 278, Upstream of Duke Power Riverbend Site ...........
54
12.
Statistical Summary of Ash Basin Effluent Water Quantity
and Quality, Duke Power Riverbend Site ....................
56
13.
Statistical Summary of Catawba River Water Quality at Station
277, Downstream from Duke Power Riverbend Site ............
59
14.
Statistical Summary of Catawba River Water Quality at
Thrift, N.C., STORET Station No. 2142808 ..................
61
15.
Statistical Summary of Catawba River Iron Concentration
in Downstream Order ..................................
62
vi

I. INTRODUCTION
At the time of issuance of permits for new ash -disposal facilities, or at the time of
renewal of permits, the North Carolina Division of Environmental Management (DEM)
may require electric utilities to institute monitoring programs to determine the effect
of ash -disposal operations on groundwater quality.
In April 1985, Duke Power Company applied for a permit to construct a basin at its
coal-fired Riverbend Plant in Gaston County, NC, in which ash dredged from its settling
pond would be placed for drying. The permit prepared by DEM in response to the
Company's request requires that:
"S. An assessment shall be made of the existing groundwater quality in the
immediate vicinity of the dredge pond. If contaminants are encountered at or
below a depth of 20 ft, the vertical and horizontal extent of those contaminants
should be established. This assessment should be made prior to use of the dredge
pond in order that background groundwater quality can be established."
T. A similar assessment shall be made to establish groundwater quality around
the periphery of the existing ash ponds. Since groundwater in the vicinity of the
ponds will ultimately discharge into the Catawba River, discovery of any
contaminants in the GA zone would lead to a determination of which areas, at
what depths, and in what concentrations those contaminants are entering the
river. The permittee shall, within 90 days of permit issuance, submit to the
Department for approval a proposed plan to assess groundwater quality at the
existing fly ash basins. The plan should include (a) a schedule for completion of
each phase of the investigation and (b) a proposed monitoring well network
sufficient to detect any contaminants which could reach the river."
Intensive studies on the effect of ash disposal have been conducted at the Allen Plant,
which is also located in Gaston County about 12 miles south of the Riverbend Plant.
These studies show that groundwater quality has not been significantly degraded by
seepage from the Allen plant ash ponds. In connection with this conclusion, it is
important to note the agencies that conducted the studies: Duke Power Company
(Roche, Gnilka and Harwood, Dec. 1984); Arthur D. Little, Inc., (June 1985) under

contract with the U.S. Environmental Protection Agency and Tetra Tech, Inc., under
contract with the Electric Power Research Institute (July 1985).
On the basis of these three independent investigations at the Allen Plant and studies
conducted by the EPA and other agencies at other coal-fired plants, Duke Power does
not believe a groundwater monitoring program is needed at the Riverbend Plant. In an
effort to determine if this is, in fact, the case, Duke Power requested Kilkelly
Environmental Associates (KEA) to conduct a comparative hydrogeologic study of the
Allen and Riverbend Plants, to estimate the input of ash trace elements to the Catawba
River both through surface runoff and groundwater, and to make a preliminary
assessment of the risk posed to health and the environment by groundwater
contaminants from the Riverbend ash pond reaching the Catawba River.
2

2. HYDROGEOLOGIC CONDMONS AT ALLEN AND RIVERBEND PLANTS
The intensive studies conducted at the Allen Plant clearly demonstrate the following:
1. The residual (saprolite) and alluvial materials underlying and adjacent to the
ash ponds have a very small hydraulic conductivity, resulting in a slow rate
of movement of water tlirougii"the materials."—__
2. The residual and alluvial materials have a very large capacity to immobilize
metals through ion exchange, which, together with the slow rate of
movement, results in negligible groundwater pollution) �q ��� .,� Ep �sg/
Ven m PJLLl.e.��
p®w -6 i
Both of these findings are consistent with the hydrogeologic conditions at the Allen
Plant. Before discussing the specific conditions at the Riverbend Plant, it will be useful
to review the general hydrogeologic conditions in the area.
2.1 GENERAL HYDROGEOLOGIC CONDITIONS
The rocks underlying both the Allen and Riverbend sites have considerable similarities.
Both sites are underlain by igneous rocks of the Charlotte Belt. They may be referred
to as "granite -diorite complex" because the light-colored granite is closely interspersed
with dark -colored diorite. The rocks are not bedded or layered but appear locally as
discrete cross -cutting dikes (see Figure 1). In some outcrops and well cores, the granite
appears to cut across, as dikes, the more predominant diorite, whereas in other places
the diorite appears to form dikes in granite.
There appears to be no pattern of orientation or directional trend along the discrete
granite and diorite boundaries. There is a slight tendency for the rock boundaries to be
more fractured than the large separate bodies of granite and diorite, at least in the
deeper fresh rock. In the saprolite, however, the spaces between the different rocks
3

Diagrammatic sketch showing the relation between
diorite and granite in the Charlotte Belt.
Conditions typical of the Allen Plant site are on
the left and those at Riverbend are on the right.
(Modified from LeGrand, 1952, f1g.3,)
FIGURE 1
4

appear to be closed by the swelling of the decomposed minerals, such as feldspars and
hornblende. The fracture system in the bedrock is difficult to determine. From present
knowledge, it can be deduced that some fractures are along the granite and diorite
boundaries but most are not. The overall yields of wells in the Charlotte Belt are about
average for the entire Piedmont region — certainly not above average.
The statements that follow apply to the Piedmont region generally and form a basis for
a conceptual model of specific sites. In these respects, both the Allen and Riverbend
sites are similar, and much of the knowledge gained from the Allen site can be applied
readily to the Riverbend site.
1. The gross groundwater system in the region is not an extensive continuum, as
is the case in most regions. Instead, the region is composed of countless
relatively small groundwater units, each unit almost confined to each small
surface drainage basin in which a perennial stream occurs.
Z. The region is underlain by igneous and metamorphic rocks; the rocks range in
chemical composition between that of granite (mainly silica and silicates of
aluminum and potassium) and that of diorite (chiefly silicates of aluminum,
iron, magnesium, and calcium).
3. A layer of saprolite lies on the fresh rock in most places; the thickness of the
saprolite ranges from a feather edge to slightly more than 100 feet.
4. Water occurs in two types of media: (a) clayey granular weathered material
and (b) underlying fractures and other linear openings in bedrock.
5. A close network of streams prevails and, in few places, the distance to a
perennial stream is more than one-half mile. A hill -and -dale topography
occurs, commonly with gentle slopes.
6. A continuous flow of groundwater occurs toward each stream. Some of the
overflowing groundwater is consumed as evapotranspiration in valleys; the
remainder discharges as small springs and as bank and channel seepage into
streams.
7. Because all the perennial streams receive groundwater from adjacent
interstream areas, streams are linear sinks in the water table. This part of
the water table is directly observable. The topography of the water table is
similar to that of the land surface, but its relief is less. Thus, it is easy to
construct synthetic water -table maps and to predetermine the general
direction of the natural movement of groundwater.

8. The path of natural movement of groundwater is relatively short and is
almost invariably restricted to the zone underlying the gross topographic
slope extending from the land -surface divide to the stream.
9. From a point source of infiltration, water, or waste that might be with it,
extends as a narrow fan or expansive trail down -gradient toward the nearest
perennial stream; its dispersal depends on the kind and degree of
permeability, on the hydraulic gradient, and on the distance to the stream.
10. Almost all recharges and discharges are through porous granular material
(clayey soil or floodplain deposits), but much of the intermediate flow
between the recharge and discharge areas is through bedrock openings.
11. The saturated zone is not simple to define. Its top boundary is the water
table, which lies in the clayey weathered material more often than not, but
which becomes discontinuous where it lies in fractured bedrock. The lower
boundary is irregular and indistinct; it is represented by the base of the zone
in which interconnecting fractures exist. The saturated zone is absent where
unfractured rocks crop out, but it is commonly 50 to 300 ft thick. Water -
yielding capacity within the zone ranges through several orders of
magnitude; commonly it is less near the base of the saturated zone than near
the top.
12. The water table is near land surface in valleys and as much as 20 to 70 ft
below land surface beneath hills. The range of seasonal fluctuation of the
water table is as little as 3 ft in valleys and as much as 8 ft beneath hills.
13. Bedrock fractures tend to decrease in size and number with depth, and in
most places there is an insignificant storage and circulation of groundwater
below a depth of 400 ft.
14. Many fractures underlie "draws" or linear sags in the surface topography.
These draws, representing zones of relatively high permeability in the
bedrock, are sites for the best -producing wells.
15. The yields of wells range from less than 1 gallon per minute to as much as
200 gallons per minute; the sustained yield of most wells is less than 100 but
more than 6 gallons per minute. The cone of pumping depression of a
domestic well does not generally extend to the cone of another pumping well
a few hundred ft away.
16. Two distinctive chemical types of groundwater are present. The first
includes soft, slightly acidic water low in dissolved mineral constituents;
water of this type comes from light-colored rocks resembling granite in
composition, and includes granite, granite gneiss, mica schist, slate, and
rhyolite flows and tuffs. The second includes a hard, slightly alkaline water
relatively high in dissolved solids; water of this type comes from dark rocks,
such as diorite, gabbro, hornblende gneiss, and andesite flows and tuffs.

Without detailed field observations of the geologic conditions at the two sites, there is a
risk in pointing out specific similarities and differences. Some probable slight
differences might be inferred as follows:
A. Granite appears to be predominant at Riverbend,_whereas diorite is
predoYnutanf at'AlIen
B. The saprolite at Allen is probably thinner and more clay -rich than the
saprolite at Riverbend.
C. The groundwater in the bedrock at Allen may be more alkaline or slightly
less acid than the water in the bedrock at Riverbend.
2.2 HYDROGEOLOGIC CONDITIONS AT RIVERBEND PLANT
The Riverbend Plant, as the name implies, is located on a peninsula about 2.5 miles long
and generally less than a mile wide formed by a great eastward bend in the Catawba
River (Figure 2). The power plant is located near the narrow neck at the western end of
the peninsula, and the ash pond is to the east on a northward bulge in the peninsula.
A topographic divide runs in a generally east -west direction down the center of the
peninsula. It is consistent with our understanding of the hydrogeology of the Piedmont
to assume that the water -table divide is coincident with the topographic divide. The
southern edge of the ash pond is about 800 ft north of the topographic divide, and the
dredge pond (ash -basin) is centered about 700 ft north of the divide.
Test borings were constructed in the area occupied by the ash pond, both in connection
with the excavation of material used in dike construction and to determine the
foundation conditions at the dikes. These borings generally penetrated only 20 to 40 ft
in sandy, clayey, silty saprolite and thus ended well above the bedrock surface. A
relatively dense network of borings exists in the area immediately to the south and
southeast of the pond, which provides more complete geologic data from land surface to
7

1�
Cb
Vl 0
��I
_-; �oL� ?•„^, II ,(�,/•• ��` 'POwerplant !
10• ' � !'! ,� ''r, � /r )/ OtQQ � � � 750 ��s • �
C
all
00
- \�\` r• v' p �•) / / 649
I u �� ;, a Bridge
'�� �• / ;n� 6y J' �- // _ 676
a..•
SCALE 1.24 000
1 0
I MILE
1000 0 1000 2000 3000 4000 5000 6000 7000 FEET
1 .5 0 1 KILOMETER
CONTOUR INTERVAL 10 FEET
DATUM IS MEAN SEA LEVEL
Location and Topography of the Riverbend Ash Pond Area
Figure 2

the top of bedrock. Data from these borings can be used to infer the geology of the ash
pond site. Driller's logs of borings B-9 and R-11 are given below. The locations of these
borings in relation to the dredge pond and the ash pond are shown in Figure 3. Note that
Boring B-9 is located along the topographic (and water table) divide, and boring R-11 is
about 300 ft east of the ash pond and about midway between the topographic divide and
the Catawba River.
Driller's Log for Boring B-9, Duke Power Co. Riverbend Plant
Depth (ft) Material
0 - 5 Red to orange, fine to coarse sandy, gravelly clay
5 - 10 Red, slightly micaceous, fine to coarse sandy,
gravelly clay
10 - 15 Red to orange, fine to coarse sandy, gravelly, silty
clay
15-20 No sample
20 - 25 Pink to orange, slightly micaceous, fine to medium
sandy silt
25 -30 Tan to orange, fine to medium sandy silt
30-35 Micaceous, fine to medium sandy silt
35-40 Tan to orange, micaceous, fine to medium sandy silt
40-45 Micaceous, fine to medium, sandy silt
45-50 Tan to orange, slightly micaceous, fine to medium
slightly sandy silt
50-66 Fine to medium slightly sandy silt
Elevation of land surface 788 ft. Hole dry.
9

O .4,000 '10,00
t,
Scale, in feet
EXPLANATION
ath of outflow segmenta
2700 ft.
1600 ft.
1400 ft.
Location of borehole
referred to in text
.ne of hydrologic
cross section
Map of Riverbend Ash Pond Area Showing Segments Used to Calculate
Ground -water Outflow from Pond, Ground -water Divides, Locations of
Hydrogeologic Cross Sections and Boundary of Outflow Plume.
FIGURE 3
10

Driller's Log for Boring R-11, Duke Power Co. Riverbend Plant
Depth (ft) Material
0 - 5
No sample
5- 10
Orange, clayey silt
10- 15
Orange, slightly micaceous, fine to medium slightly
sandy, slightly clayey silt
15-20
Tan to red, micaceous, fine to medium sandy silt
20-30
Red to orange, micaceous, fine to medium slightly
sandy, silty clay
30-55
Brown to orange, fine to medium slightly sandy,
clayey silt
55-60
No sample
60-95
Brown to orange, slightly micaceous, fine to coarse
sandy silt
95- 120
Tan to brown, micaceous, fine to coarse sandy silt
(Groundwater freely entered the hole in this section)
Elevation of land surface 720 ft. Depth to water level about 55 ft
below land surface after 24 hours.
The logs of these borings, which are typical of many others in the area, show that the
upper part of the saprolite tends to contain significant clay because of the chemical
breakdown of feldspar present in the granitic rocks. Downward, toward the
unweathered rock, the material becomes silty with sand -size fragments of quartz.
Field logs of 17 test borings in the area east and south of the ash pond were analyzed to
determine the characteristics and thickness of the saprolite. The borings were
constructed both with a power auger and with a rotary drill, with water as the
circulating fluid. None of the borings penetrated bedrock, but an effort was made
during drilling, through the use of standard penetration resistance tests, to ensure that
the borings were near the bedrock surface.
11

Figure 4 shows the altitude of the land surface, based on a topographic map of the ash
pond area with a 2 -ft contour interval prepared by Duke Power Co. Figure 5 is a map of
the ash pond area showing the configuration of the bedrock surface based on the total
depth of the test borings. It is important to note that because none of the borings
actually penetrated bedrock, the altitudes of the bedrock surface are somewhat less
than those shown in Figure 5. The contours on the top of the bedrock were extended
into the site of the ash pond on the basis of the relations between land -surface
topography and depth to bedrock that exist in the area in which boring data were
obtained.
The approximate thickness of the saprolite can be determined from the difference in the
altitudes in Figures 4 and 5. The approximate thickness of the saprolite is shown in
Figure 6 in the area for which boring data were obtained. As shown on this figure, the
thickness of saprolite ranges from about 20 ft to more than 120 ft and averages between
60 and 80 ft in the area. Review of the penetration -resistance tests suggests that the
borings in the area where the saprolite appears to be relatively thin were terminated
well above the bedrock surface.
It is apparent from the preceding discussion that the specific data from the borings
provide key information about (1) the depth to the bedrock (and thickness of the
saprolite) and (2) depth to the water table (and thickness of the unsaturated zone).
Interpolation between data points (the placement of contours) that was necessary to
prepare Figures 4, 5 and 7 is based partly on hydrogeologic experience, especially on the
relation of land -surface topography to depth to bedrock and depth to the water table.
Thus, the contour maps and the cross-sections are believed to represent usefully close
approximations but are not necessarily exact representations of conditions at all points.
12

('.eww O /000 zoo*
Scale in feet
PLANATION
Contour on land
surface in feet
above mean sea level
Contour interval
20 feet
Map of Riverbend Ash Pond Area Showing Land -Surface
Topography Prior to Development of Ash Pond
FIGURE 4
13

COQ
t•
v
6°0- -
62°
bfCO
C60
660
w�
- /.
-- 6040 -
(�✓ Q
■Trri
1000 ZaOO
Scale in feet PLANATION
Contours based on
�--� boring data
Contours based on
topographic map
hour interval
20 ft
Map of Riverbend Ash Pond Area Showing the Approximate
Configuration of the Bedrock Surface
FIGURE 5
14

.ea
eo �Ir
'80 .Co
.c•
'Y.
v°
.a
i.
j
2 apo
e in feet
.0
s•
•8•
.8•
EXPLANATION
Thickness in ft
based on difference
in altitude between
land surface and top
of bedrock
Map of Riverbend Ash Pond Area Showing Estimated
Thickness of Saprolite Based on Boring
Data Supplied by Duke Power Co.
FIGURE 6
15

Cho. CV,1
1
Al
t n
�B. / •' 1111 I I I N 111 � ,'
_ _ �I� �•'`M Water table in rock in
this area at an altitude
less than 720 ft.
Contour based on
measurements in
bore holes
Contour estimated
from water levels
in pond and river
Contour interval 20 ft
Map of Riverbend Ash Pond Area Showing the Approximate
Altitude of the Water Table in the Saprolite and Dikes
FIGURE 7
16

3. EFFECT OF RIVERBEND ASH POND ON GROUNDWATER
The Riverbend ash pond is divided, by a dike, into two cells. The water surface is
maintained at a level of about 719 ft above sea level in the southern cell and at a level
of about 713 ft in the northern cell. During periods when ash slurry is being added to
the pond, excess water is discharged from the lower (northern) cell directly to the
Catawba River.
The Catawba River stage is maintained at an average level about 648 ft above sea level
by the Mountain Island Dam, which is 8 miles downstream. Because the river stage is
below that of the pond, there is a hydraulic gradient between the pond and the river
through the groundwater system. As a result, an important consideration related to the
operation of the ash pond is its effect on groundwater quality and, ultimately, the effect
of groundwater derived from the pond on the quality of the river.
The effect of the ash pond on groundwater depends primarily on
• the vertical hydraulic conductivity (K) of the ash and the saprolite,
• the transmissivity (T) of the saprolite and bedrock,
• the hydraulic gradients in the saprolite and bedrock, and
• the ion -exchange capacity of the saprolite.
The groundwater system at Riverbend, as at the Allen Plant and other places in the
Piedmont, consists of two distinctly different parts: (1) the granular surf icial layer of
saprolite that forms during the chemical and physical breakdown of bedrock, in the
process referred to as "weathering," and (2) the unweathered bedrock. Water occurs in
and moves through the saprolite in the pore spaces between rock particles and through
17

the bedrock in sheet-like openings that develop along fractures. Because of significant
differences in the hydraulic properties of saprolite and bedrock, it is necessary to treat
the groundwater system at Riverbend as a two-part system.
Water in the ash pond moves vertically downward across the ash layers into the saprolite
and also laterally through the dikes and surrounding saprolite. Most of the water that
enters the saprolite beneath the pond moves through the saprolite toward the Catawba
River; the remainder seeps downward into the bedrock and then laterally to the
Catawba River. Because the stages of the pond and river are maintained at relatively
constant levels, the movement of water from the ash pond to the river through the
groundwater system can be analyzed with steady-state equations that involve only the
water -transmitting characteristics of the ash, saprolite, and bedrock.
3.1 HYDRAULIC CONDUCTIVITY OF ASH AND SAPROLITE
No determination of hydraulic conductivity (K) of the ash and saprolite have been made
on samples from Riverbend. However, the report on the Allen Plant prepared by Arthur
D. Little, Inc. (ADL) contains results of laboratory determinations of the K of ash, and
the report prepared by Tetra Tech contains data on the saprolite. The data from these
reports are summarized in Table 1.
TABLE 1
Hydraulic Conductivity Data from the Allen Plant
Material Hydraulic Conductivity (K)
(ft/d)
Ash 2.8 X 10-4 - 5.6 x 10-1
(avg. of 4 samples = 1.6 X 10-1)
Saprolite 1.05 - 2.8
(avg. of 9 samples = 1.34)
18
Source
A.D. Little, Table 5-2
Tetra Tech, Table 4-18

The hydraulic conductivity of ash depends on the physical characteristics and thickness
of the ash and on the length of time it has compacted. As a result of compaction, the
bottom layers of ash in ponds are generally the least permeable, and it is these layers
that control the rate at which water seeps from ponds. However, for purposes of
computations at Riverbend that require the K of ash, the average value of 1.6 X 10-1
given in Table 1 will be used.
Relative to hydraulic conductivity of the saprolite, values for 9 samples from the Allen
Plant averaged 1.34 ft/d and ranged from 1.05 to 2.8 ft/d. Recognizing that the
saprolite at Riverbend may be somewhat more permeable than that at Allen, an average
value of 2 ft/d will be used in calculations for Riverbend.
3.2 TRANSMISS=Y OF THE SAPROLTTE AND BEDROCK
The quantity of water moving between the pond and the river depends on, among other
factors, the water -transmitting capacity of both the saprolite and the bedrock. This
capacity is referred to as transmissivity (T) ,which is equal to the hydraulic conductivity
(K) times the thickness (b) of the zone through which movement occurs (T = Kb).
The transmissivity of the saprolite at Riverbend can be estimated by multiplying the
hydraulic conductivity discussed in the previous section by the thickness of saprolite
through which water moves. Estimates of this thickness can be obtained from the
difference in the altitude of the bedrock surface (Figure 5) and the altitude of the water
table (Figure 7). However, for the purpose of this analysis, equally satisfactory values
for saprolite thickness can be obtained from the difference in altitude between the
water table and the bedrock surface on the hydrogeologic sections shown in Figures 8
and 9.
19

D
Land
Surface
�� F1
750
I I
700-----—,__
Saprolite
I i
Ash pond
,
p
— _
Ash pond
I
y\,\
Bedrock � potentiome rlcsurface? —?
Catawba
650
I _
o\ River
Bedrock
���� Saprolite
r ---
•�.?=
600..,
-?
IBedrock
560
0
50
1000 1500 2x00 2500 30po
3500 4000
Feet
Hydrogeologic Cross Section Along Line D -D' in Figure 3
Figure 8

E
> 750--
E I
50
E'
Land
700 ��' Ash pond water urface
Catawba ��°b�0 ? ? 0162
CO River __ tri — Catawba
650 — Saprolite Bedrock potentiomec�`ur -�� �` River
>
° 600
Bedrock
w 560
0
500 1000 1500 2000 2500
Feet
F
750
F
Q)
Land
n 700
Surfoc
��-� p
Ash
pond
h'°fey
Cba
bN'
Saprolite
Riow
�Woi i
650
Bedrock
potentiometric
Surface
oi^^------^-?
--------.
ca 600
;
Bedrock
`
f
0
500
1000 1500
2000
Feet
Catawba
River
3000 3500 4000
F"
Hydrogeologic Gross Sections Along Lines E -E" and F -F" in Figure 3.
Figure 9 -
e

The lines of these sections are shown on Figure 3, together with the segments that will
be used in calculating groundwater outflow from the pond to the river. The water -
transmitting thickness of saprolite for each of the outflow segments is shown in
Table Z.
TABLE 2
Thickness and Transmissivity of the Saprolite for the Outflow Segments
Shown in Figure 3
Outflow Hydrogeologic Saprolite Transmissivity
Segment Section Thickness) (ft) (ft2/d)
A E - E"
70 140
B F - F'
60 120
C F' - D'
90
F' - F"
110 200)
A� thickness midway between pond and river.
V Based on average thickness of 100 ft.
No data on hydraulic conductivity and transmissivity of the bedrock are available for
either the Allen or Mverbend Plant areas. Water -bearing fractures in the bedrock are
very irregular in occurrence, as discussed in Section 2.1. This fact, together with large
differences in water -transmitting capacity between different fractures and even at
different places along the same fracture, causes the hydraulic conductivity to range
from essentially zero in some parts of the bedrock to as much as 10 ft/d in narrow,
densely fractured zones.
However, to estimate the effect of the ash pond on groundwater, it is necessary to
estimate the transmissivity of the bedrock. The first consideration in this process is the
thickness (b) of the water -transmitting zone in the bedrock beneath the peninsula on
which the ash pond is located. The Catawba River is a groundwater discharge zone of
regional impact so that all groundwater in the bedrock beneath the peninsula is derived
from recharge on the peninsula. Because of the narrow width of the peninsula, the
22

longest possible groundwater flow lines are less than one-half mile in length. In view of
this, it is doubtful that there is any significant movement of groundwater below a depth
of 50 ft below the bedrock surface. Using this value for thickness and 1 ft/d for
hydraulic conductivity, the transmissivity of the bedrock is estimated to be 50 ftZ/d.
3.3 HYDRAULIC GRADIENTS IN THE SAPROLITE AND BEDROCK
As noted previously, the difference in the water levels between the pond and river
provides the condition necessary for the development of a hydraulic gradient between
the pond and river through the groundwater system.
The estimated altitude of the water table in the dikes and saprolite is shown in Figure 7.
The altitudes on this figure were used to draw the water -table profiles in Figures 8 and
9. It should be noted that the water -table profiles in Figures 8 and 9 are drawn to
connect the water surfaces of the pond and river. However, it is important to also note
that the hydraulic gradient represented by the water table is not the "effective"
hydraulic gradient in the saprolite between the pond and the river.
To understand this fact, note that the potentiometric surface of the bedrock is also
shown in Figures 8 and 9. This surface, which shows the height to which water would
stand in wells that are cased through the saprolite and finished as open holes in the
bedrock, is substantially below the pond levels and the water table. This is to be
expected because, for water to move downward across the ash in the pond and across
the saprolite to the bedrock, there must be a gradient between the pond surface and the
bedrock. The magnitude of this gradient (i.e., the difference in level between the pond
and the potentiometric surface) reflects the head loss as water moves across the ash and
saprolite.
23

The average effective hydraulic gradient that controls flow from the pond to the river
through the saprolite is much flatter than the water table but somewhat steeper than
the potentiometric surface of the bedrock. This conclusion is based on the boring logs,
which indicate that the lower part of the saprolite is the most permeable. As a result,
water from the pond will tend to move vertically downward across the upper part of the
saprolite and then laterally to the river through the lower part. Thus, the effective,
lateral hydraulic gradient in the saprolite, as stated above, will be much flatter than the
water table. However, in the absence of additional data, the average effective lateral
hydraulic gradient in the saprolite will be assumed, for purposes of outflow
computations, to be the average of the water table and bedrock gradients along each of
the outflow segments. These gradients are shown in Table 3.
TABLE 3
Hydraulic Gradients for Saprolite and Bedrock for the Outflow Segments
Shown in Figure 3
Hydraulic Average
Hydrogeologic Outflow Hydrogeologic Gradient Gradient for
Unit Segment Section (ft/ft) Segment (ft/ft)
Saprolite A
E - E'
.07
(and dikes)
E' - E"
.05
.06
B
F - F'
.05
.05
C
F' - D'
.09
F' - F"
.07
.08
Bedrock A
E - E'
.025
E' - E"
.02
.02
B
F - F'
.01
.01
C
F' - D'
.013
F' - F"
.012
.01
Relative to the potentiometric surface of the bedrock, it was noted previously that none
of the bore holes penetrated the bedrock. Therefore, no data are available on the
bedrock potentiometric surface. However, boring B-9, along the water -table divide, was
dry at a depth of 66 ft or at a bottom altitude of 722 ft (see Section 2.2). To estimate
24

the altitude of the water level in bedrock, the Jacob equation for the profile of the
water table was solved for a peninsula 1 mile wide, with a bedrock transmissivity of 50
ft2/d and a recharge rate of 100,000 gpd/mit. The solution shows a height of the water
table 31 ft above the controlling water level at the discharging boundary (the Catawba
River). This places the water table in the bedrock at the divide south of the Riverbend
ash pond at an altitude of about 679 ft (648 + 31), the position shown on Figure 8.
From this altitude on the divide, the bedrock potentiometric surface was sketched to
the river surface. The potentiometric surface, as sketched, probably results in a steeper
hydraulic gradient and, therefore, a larger outflow through the bedrock than actually
exists, because sediment on the bottom of the river hampers the discharge of water
from the bedrock, causing the potentiometric surface (static head) in the bedrock to be
above the river level beneath the river. The hydraulic gradient for the bedrock, based
on the profiles of the potentiometric surface shown in Figures 8 and 9, are also shown in
Table 3.
3.4 GROUNDWATER OUTFLOW FROM THE ASH POND
Sections 3.1 through 3.3 provide the data necessary to calculate the rate of flow of
water from the ash pond to the Catawba River through the groundwater system.
Darcy's law applies to the steady-state condition that exists in the area and is,
therefore, used to calculate the rate of flow. Darcy's law, expressed as an equation, is
Q = Tw dh/dl
where:
Q is rate of flow in ft3/d,
T is transmissivity in ft2/d,
w is width of the outflow segment in ft, and
dh/dl is the hydraulic gradient in ft/ft.
25

The values used for items in the above equation and the calculated outflow through each
of the segments shown in Figure 3 are given in Table 4.
TABLE 4
Groundwater Outflow from the Ash Pond
Outflow
Segment
Length of
Segment
Hydrogeologic
Unit
Transmissivity
(ft2/d)
Hydraulic
Gradient
(ft/ft)
Rate of
Outflow
(ft3/d)
A
2700
Saprolite
140
.06
22,680
Bedrock
50
.02
Z,700
B
1600
Saprolite
120
.05
9,600
Bedrock
50
.01
800
C
1400
Saprolite
200
.08
22,400
Bedrock
50
.01
700
TOTAL
58,900
As shown in Table 4, the estimated outflow from the ash pond to the Catawba River
through the groundwater system is 58,900 ft3/d or about 0.7 cfs or 440,000 gpd. For
comparative purposes, a water -budget analysis in the ADL report on the ash ponds at
the Allen Plant shows an outflow from the ponds through the dikes and the groundwater
system of about 71,000 ft3/d (533,000 gpd).
It may also be of interest to compare the total outflow shown in Table 4, of 0.7 cfs, with
the flow of the Catawba River. The average flow of the river at the Riverbend Plant is
estimated to be 2600 cfs, so the inflow from the ash pond added through groundwater to
the river represents only about 1/3700 of the average flow of the river.
The computations in Table 4 show a rate of outflow through the bedrock of 4200 ft3/d
(31,000 gpd). This quantity includes both water originating in the pond and water that
reaches the bedrock in the area between the water -table divide and the south side of the
W.

pond (see Figure 3). This area is estimated to include about 19 acres or about .03 square
mile. If the recharge to the bedrock in this area is about 100,000 gpd/mit, as estimated
in Section 3.3, the rate of movement of water from this area toward the pond is about
400 ft3/d (3,000 gpd). This water moves through the bedrock beneath the pond and is,
therefore, a part of the bedrock outflow calculated in Table 4. Subtracting the
400 ft3/d from the 4200 ft3/d moving into the river from the bedrock leaves 3800 ft3/d
reaching the bedrock from the ash pond.
The boundaries of the zone through which water originating in the ash pond and dredge
pond may be reaching the river are shown in Figure 3. It is important to note that the
maximum extent of the plume, as predicted in Figure 3, is entirely on property owned by
Duke Power Company.
3.5 GROUNDWATER VELOCITIES AND TIME OF TRAVEL
The rate at which water from the pond moves through the groundwater system to the
river is also of interest in connection with the effect of the pond effluent on
groundwater and river quality. It is well known, at least among hydrogeologists, that
groundwater moves slowly - compared, say, to the rate of movement of water in
streams. The rate of movement of groundwater in the vicinity of the Riverbend ash
pond was calculated for both the saprolite and the bedrock.
The equation used in these calculations is Darcy's Law which, in terms of velocity, is:
27

KA
V =
n dl
where:
v is velocity in ft/d
K is hydraulic conductivity in ft/d
n is "effective" porosity and is dimensionless, and
dh/dl is hydraulic gradient in ft/ft.
The values used in solving the equation and the results are shown in Table 5.
TABLE 5
Estimated Groundwater Velocities in Saprolite and Bedrock in the Riverbend Ash
Pond Area
Estimated Velocity
Relative to Table 5, it should be noted that because the upper part of the saprolite tends
to be less permeable than the lower part, the value of hydraulic conductivity used in
calculating vertical velocity is 1/10 the value used in calculating horizontal velocities.
28
Hydraulic
Hydraulic
Conductivity
Effective
Gradient
Material
(ft/d)
Porosity
(ft/ft)
(ft/d)
(ft/yr)
Saprolite
Vertical
0.2
0.2
0.5
0.5
180
Horizontal
2
0.2
E - E'
0.06
0.6
220
E' - E"
0.06
0.6
220
F - F
0.05
0.5
180
F' - F"
0.08
0.8
290
E' - D'
0.02
02
70
Bedrock
1
0.001
E' - D'
0.007
7
2500
F' - D'
0.01
10
3600
Relative to Table 5, it should be noted that because the upper part of the saprolite tends
to be less permeable than the lower part, the value of hydraulic conductivity used in
calculating vertical velocity is 1/10 the value used in calculating horizontal velocities.
28

The calculated velocity of groundwater movement through the bedrock seems too large.
If this is, in fact, the case, it suggests that either the porosity is too small or the
hydraulic conductivity or hydraulic gradient are too large.
In any case, it is clear from the rounded values of velocity shown in the last column of
Table 5 that sufficient time has elapsed since the Riverbend ash pond was placed in
operation in 1957 for water to move from the pond to the river. This is also apparent
from Figure 10 which shows the estimated times of travel for water originating in the
ash pond to reach the Catawba River. The effect of "apparently" faster velocity in the
bedrock is clearly evident in the figure.
3.6 EFFECT OF POND SEEPAGE ON GROUNDWATER QUALITY
One of the primary concerns of the North Carolina Division of Environmental
Management is the effect of seepage from the Riverbend ash pond on groundwater
quality. This was also the major concern of the three studies conducted at the Allen
Plant that are referred to in the introduction to this report.
The materials disposed of in the Riverbend ash pond include fly ash, bottom ash, and
boiler cleaning wastes. These are the same types of materials disposed of in the Allen
Plant ash ponds so that the results of the intensive chemical analyses and studies
conducted at the Allen Plant are also generally applicable to Riverbend.
The similarity in the chemical composition of ponded ash at the Allen and Riverbend
Plants is shown in Table 6 which contains the results of leachate extraction studies
conducted by Duke Power Company on samples collected at both plants in 1980. The
leachate was generated through the application of the Environmental Protection Agency
Extraction Procedure which is described by Roche, Gnilka, and Harwood (1984, p.2).
29

U
750
Land
face
700 - -" - - Ash pond Ash pond D�
- -
Saprolite - - - - - - - - - - �2 Catawba
650 �� y°0raRiver
20 years 15 years
600Bedrock -----5 years
10 years —
w 560
0 500 1000 1500 2000 2500 3000 3500 4000
Feet
Estimated Times of Travel in Years Along Hypothetical Flowlines
From the Riverbend Ash Pond to the Catawba River
Figure 10

TABLE 6
Concentration of Selected Metals in Leachate Extracted From
Ash Samples Through the Use of the EPA Extraction Procedure
and EPA Toxicity Criterion Limits for Solid Wastes Under the
Resource Conservation and Recovery Act. (From Roche, Gnilka,
and Harwood, 1984, Table 1 and p. 3). (All concentrations in
Constituent
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
parts per billion.) �t
EPA Toxicity
Criterion s
Allen
51
1200
<25
10
<500
0.11
<6
150
Riverbend
Cell 1 Cell 2
82
1100
<25
20
<500
<0.1
<6
30
75
1300
<25
60
<500
<0.1
<6
40
6-
5000 G-3
1000
100,000
1000
5000
5000
200
5000
Coal, and ash derived from its combustion, contains a large number of metals in trace
concentrations. However, the results in Table 6 suggest that most of the metals in the
ash are not readily soluble. In this regard, it should be noted that the EPA extraction
procedure involves pH adjustment to 5.0 in the leachate -extraction process.
The studies conducted at the Allen Plant by the Arthur D. Little Co. included trace
metal analyses of soils, ash solids, ponded ash, and of samples of groundwater upgradient
from the ash pond, beneath the inactive and active ponds, and down gradient from the
ponds. Unfortunately, the metals in the ADL analyses included only arsenic from among
the metals listed in Table 6. The analyses made during the ADL study were also
31

subjected to an intensive analysis in the Tetra Tech report, including an averaging of the
analyses from groups of wells reflecting conditions in different parts of the effluent
plume. (See Tetra Tech, Inc., 1985, Tables 4-18 and 4-19.)
Table 7 contains analyses from the ADL report, that show the concentration of selected
constituents in the soil and ponded ash, in the native groundwater (upgradient), and the
well in saprolite beneath the active ash pond that should contain the largest
concentration of metals derived from the ash. The locations of wells and other features
of the Allen ash ponds are shown in Figure 11. Two values are shown for several of the
constituents in some of the analyses in Table 7. The significance of these values are not
explained in the ADL report but are assumed to represent the minimum and maximum
values of a series of samples obtained from the wells that were sampled.
The data in Table 7 support the conclusion that the metals in ash are not readily soluble.
The most obvious exceptions to this are iron and manganese which occur in much larger
concentrations in the well in the plume than in the up gradient well and also in larger
concentrations than those allowed by EPA and North Carolina standards. Fortunately,
the concentrations of iron and manganese allowed by the standards are based on taste
and the discoloration of laundry, bathroom and other fixtures rather than public-health
considerations. In view of the fact that the maximum possible area that could be
occupied by the plume from the Riverbend ash pond is on property owned by Duke Power
Company (Figure 3) the concentrations of iron and manganese in excess of the standards
are of relatively little significance. In fact, it is possible that some or all of the
additional iron and manganese in ground water from the plume is due to the ion
exchange of metals in the pond effluent for iron and manganese that occurs naturally in
the saprolite.
32
SAN. /

TABLE 7
Selected Chemical Analyses From the Allen Plant Related to Ash -pond Seepage. (From Arthur
D. Little, Inc., 1985, Table 5.3.) (All constituents in mg/l except arsenic which is in pg/l.)
q Ash solids Groundwater Groundwater EPA Drinking
Saprdiite (Wells 3-2, up Gradient in Plume i' j Water
Constituent (Well 3-4) 3-3) (Well 3-4B) (Well 3-2) i Standards)
Calcium
471-4056
2251-4578
9.95-10.9
15.8-17 �- I
Sulfate
2.1
1.4 3s -7 250
Arsenic
0.6-1.41 ,If
16.2-57.
< 0.2-7.0
0.057-0.76 `-Iyl
50
Boron
< 0.005-0.016
< 0.15-1.6 pia' , 0.75)
Copper
952-17.6
20.8-45.1
< 0.008
< 0.008 1
Iron
11,164-16,558 11,700-29,491
< 0.01
25.9 0.3
Manganese
155-303
83-171
< 0.01-0.07
6.44-14 0.05
Nickel
4.48-10.8
5.3-26.0
< 0.05
< 0.05
Strontium
8.85-33.1
112-239
0.141-0.166
0.241-0.274
Vanadium
28.1-49.1
22.2-41.5
<0.005-0.016
<0.005
Zinc
22.8-36-2
18.5-45.7
<0.05
<0.05 5
1 v, V"�
North aro ina underground water
quality standards
are the same as those of EPA for
arsenic, iron, and manganese.
No standards
are specified for
the other constituents listed.
2/ EPA criterion for protection of sensitive crops.
33

The very low concentration of metals derived from the ash in the water samples
obtained from the well in the plume is not surprising. This is due, in the first place, and
as mentioned above, to the fact that the metals in the ash are not readily soluble. It is
due in the second place, and even more importantly, to the well-known capacity of
saprolite to immobilize metals through the process of ion exchange - that is, by the
process whereby metals dissolved from the ash and carried in solution in the effluent
from the pond are exchanged, upon reaching the saprolite, for other metals which are
natural constituents of the saprolite.
The ADL study included analyses of the ion -exchange capacity of alluvial materials and
of saprolite at the Allen Plant. The results of these studies are contained in Table 5.4
of the ADL report and show a wide range in the capacity of the different materials that
were tested to tie up metals contained in the pond effluent. It is not possible, without a
detailed analysis of the basic data (which are not included in the report), to identify the
reason for the wide range.
However, it is suspected that the samples which show the smallest ion -exchange
capacity were from parts of the plume where much of the ion -exchange capacity had
already been depleted. If this is the case, the largest exchange capacities are the ones
of greatest importance to the context of this discussion. These show that the saprolite
and other materials have the capacity to remove metals, such as arsenic, copper, and
selenium, equal to 1,000 to nearly 10,000 times the concentration present in the pond
effluent. These numbers mean, in effect, that unit volumes of the saprolite and other
materials have the capacity to remove the metals from 1,000 to 10,000 unit volumes of
the effluent.
34

The large ion -exchange capacity of the residual soils (saprolite) in the Piedmont area is
obviously an important factor in the lack of significant groundwater pollution problems
in the vicinity of ash -disposal ponds. With this in mind, it is important to call attention
again to the thickness of saprolite in the Riverbend ash pond area which, as shown in
Figure 6, ranges from about 20 to 120-f t. The exact thickness beneath the -pond is not
known but is believed to range from about 60 to possibly more than 100 ft.
The Tetra Tech report contains an analysis of the vertical attenuation of chemical
constituents based on analyses of water samples from wells in the ash pond at the Allen
Plant. This analysis confirms the results of the ADL analyses and also shows a large
ion -exchange capacity still remaining at a depth of only 8 ft. below the ash in 1981 and
1982, about nine years after use of the "active" ash pond began. (See Table 4-20 in the
Tetra Tech report.) In other words, the rate of seepage and the concentration of metals
in the effluent are such that the ion -exchange capacity of the saprolite is being depleted
("used up") at a rate considerably less than 1 ft/yr. Assuming that the ion -exchange
capacity of the saprolite at Riverbend is similar to that at Allen, as appears to be the
case based on the similarity of the hydrogeology of the two sites, it appears likely that
metals derived from the Riverbend ash pond have not only not yet reached the Catawba
but have not yet reached the bedrock.
The effect of ash pond seepage, both on groundwater and, ultimately, on the Catawba
River involves both hydrogeologic and geochemical factors. The preceding discussion in
this section has dealt with some of the results of previous studies, including the effect
of ion exchange on the concentration of metals contained in the seepage. With the
background of that discussion, it will be useful at this point to deal somewhat more
comprehensively with the hydrogeologic and geochemical factors. These factors
include:
35

1. Differences in the mineral composition of the saproliteJ from place to place
resulting from differences in mineral composition of the granite -diorite
complex that forms the bedrock at both the Allen and Riverbend sites.
These differences affect both the "background" groundwater quality and
cation -exchange capacity of the saprolite.
2. Lateral and vertical differences in mineral composition of the ash resulting
from the discharge to the pond at different times of bottom ash, fly ash, and
boiler -cleaning wastes. These differences affect both the rate of movement
of water through the ash and its chemical composition as it enters the
saprolite.
3. Differences in the "attractive force" between the saprolite and the different
metals dissolved in the pond effluent. It is this "force" that determines the
ion -exchange capacity.
Relative to the first factor, all of the materials through which the effluent moves are
capable of ion exchange; the amount depending on mineral composition and grain size of
the materials (Davis and De Wiest, 1966, p. 88-92). Because the saprolite tends to be
finer grained and more clay -rich with increasing distance above the bedrock, the upper
part of the saprolite has the largest ion -exchange capacity.
Relative to the third factor, the attractive force between ions and the surface of the
rock particles is different for different metals with the result that those that are most
strongly attracted will displace those that are weakly held. Freeze and Cherry (1979, p.
133) show that the normal order of preference for some monovalent and divalent cations
for most clays is
Affinity For Adsorption
Monovalent ions Cs > Rb > K > Na > Li
Stronger -* Weaker
Divalent ions Ba > Sr > Ca > Mg
For convenience in this discussion, all of the earth materials between land surface
and the top of bedrock will be referred to as saprolite. However, it should be noted
that they include not only saprolite but also the materials used to construct the
enclosing dikes and, near the river, thin alluvial deposits.
36

The divalent ions are normally held more strongly than the monovalent ions.
Differences in the attractive force not only result in replacement of ions that are
natural constituents of the rocks with ions in solution in the effluent but also result in a
"partitioning" of the metals in the effluent along flowlines. Thus, referring to the above
diagram, calcium should advance through the saprolite more rapidly than strontium and
strontium should advance more rapidly than barium.
Other factors that affect the rate of movement of metals from the ash pond include pH
and temperature. However, the preceding discussion is sufficient to show that chemical
analyses of water samples from wells at different depths in the saprolite and at
different distances from the pond should not be expected to show a perfectly consistent
pattern radially away from the pond.
In an effort to evaluate the effect of the factors discussed above on the movement of
metals from the pond to the river, the concentration of selected constituents for wells
screened in the saprolite at different depths beneath the ash in the active Allen pond
and at different distances from the pond were plotted on Figure 12. The locations of
the wells are shown on Figure 11 and their positions with respect to the pond are shown
on Figures 13 and 14. It should be noted that the large range in chemical concentrations
required the use of a logarithmic scale in the preparation of Figure 12. The thickness of
non -ash material between the wells and the pond was also plotted on a logarithmic scale
in order to expand the separation of the wells in the ash pond.
The concentrations of the different constituents plotted on Figure 12 represent samples
collected from the wells on two different dates, one in late February and one in late
37

NN
Oil -
120
1ALLEN DISPOSAL SITEGASTON COUNTYNORTH CAROLINAADL WELLS
O UTILITY WELLS13pA�'
G SURFACE WATER SAMPLING STATIONS �• Pond
•"_• Dili i+ti
SCALE
T
0 275 550
FEET
• 3-41A
• 3-48
01
09
3-2 •
A•1• 3-3
Coal
Pile
Runoff
LAKE WYLIE
•
3-1
nds,,
Im
Map of the Ash Pond Area at the Allen Plant of Duke Power Company
Showing the Locations of Water -Quality Sampling Wells and Other
Features (from Tetratech, Inc., 1985, Fig. 4-48).
Figure 11
38
aaaaZ ■
08
L__�
1
r %.
x
!
ASH
.�
DELTA
r,
0
2
ACTIVE 1 -
�'
�•
ASH POND I Ash
(1973 to Present) 1 Discharge
\
00•.
(
4 5,
To Power
RECLAIMED
Plant
\
ASH POND
`
(1957-1973)
3-2 •
A•1• 3-3
Coal
Pile
Runoff
LAKE WYLIE
•
3-1
nds,,
Im
Map of the Ash Pond Area at the Allen Plant of Duke Power Company
Showing the Locations of Water -Quality Sampling Wells and Other
Features (from Tetratech, Inc., 1985, Fig. 4-48).
Figure 11
38
aaaaZ ■

In a.h
3-2A
o Oj I
E
O
•F
IM
Calowba River-
NAr
3
3 -TA
9ackpr°und
3-7
Groundwater
3-4e•Actud
3'7
_�l
vahm
1 °t_•n w.
men w►,.
NO °A.
°A.
e
e.F.
,
ce
0 O F.
.
F F
e.
e.
e.
Nn• OW
►N
•
Mn MA
$.I
s<j
IS,
s<
e 1 1
F.
c°
0°
C.
c<•
Explanation
• _ 3 -2 -Well number
•8a -Chemical symbol a'•
• -Actual concentration
F` I -Range In concentration
o -Concentration less than
G�
. Thickness, in feet, of non -ash material along flowine.
Concentrations of Selected Metals Versus Length of Flowlines in
Saprolite at the Active Allen Powerplant Ash Pond of Duke Power
Company. The Dashed Lines Show the Decrease in Concentration
With Distance Beneath the Ash for Selected Metals.
Figure 12
39

INACTIVE POND
Well 3-1
5.5 ft
560ft
Fly
ash
Bottom
ash
Alluvial
deposits
Well 3-2
5ft
570ft
Fly
ash
Soil
zone
Soprolite
ACTIVE POND
Well 3-3
I
Well 3-2A
Fly
ash
Fly
ash
9 ft Saprolite
586 ft 586 ft
0
Vertical
scale
(ft)
10
Sections Showing Materials Penetrated and Positions of Screens in
Water -Quality Sampling Wells Drilled Through the Ash in Both the
Inactive and Active Ash Ponds at the Allen Powerplant of Duke Power Company.
Figure 13
40

E
Catawba River
H100ft
250
W
offset in line of Dike
I cross section _ - - 620ft
Water
-7A -7 ---------
-- -_�= 600ft
'
--- FILL -compacted sandy,
clayey sflt. '
3-6A
Saprolite _ . -'
Saprolif 0
Bedrock
1 I 1 1
200 150 100 50 0
Distance in feet from a§h pond
Length of flowlines- from -ash pond to well screens.
(Based on the shortest straightliine- distance)
Well no. Distance (ft)
3-6 175'
3-6A 200
3-7 220
3- 7A 25d
500ft
Geologic Section Extending from=the Active Ash Pond at the
Allen Powerplant of Duke Power Company to the Catawba River
(Lake Wylie). Relative Locations and Screened Intervals of Wells
Used in Figure 12 are Shown on the Section.
Figure 14
41

March 198 1. Differences in concentration on the different dates are shown as vertical
bars. It is not possible, from the available data, to determine the cause of the
differences in concentration between the sampling dates. However, it is suspected that
much, if not all, of the differences are due to inaccuracies in the analytical procedures
and to differences in sampling procedures. (This statement is not intended as a
criticism of either the analysts or the sample collectors, but is made to call attention to
the unavoidable difficulties in determining the concentration of substances that are
present in only fractions of a part per million.)
In reviewing Figure 12, it is important to note that the values plotted along the left-
hand side of the graph are of analyses of water samples obtained from the ash and which
thus show the concentrations in the effluent that reaches the saprolite. Referring back
to the second factor listed above, it probably should not be assumed that these analyses
show the composition of the effluent at all points in the pond.
Note also in Figure 12 that the values plotted along the right-hand side of the graph are
of analyses of water samples from a well upgradient from the pond which were collected
to show the "natural" quality of the groundwater. Again, referring back to the first
factor listed above, it probably should not be assumed that these analyses show the
composition of unpolluted groundwater at all points in the area.
It is evident from the left-hand side of Figure 12 that metals from the active ash pond
at the Allen plant have moved into the upper several feet of the saprolite - that is,
much of the cation -exchange capacity has been utilized to this depth. It is also evident
from the right-hand side of the figure that none of the metals had reached the nearest
wells downgradient from the pond by 1981. In this regard, note that the concentrations
of the different metals in wells 3-6, 3-6A, 3-7, and 3-7A coincide closely to the

concentrations in well 3-4B, the "background" well. The scatter in the data are believed
to reflect the effect of the factors discussed above plus analytical and sampling
problems. The effect of the second factor seems to be especially evident from a
comparison of the values for barium, iron, and manganese in well 3-2A with the higher
values in wells 3-2 and 3-3. It is believed that the higher values in wells 3-2 and 3-3
show the effect of differences in effluent quality from time to time and from place to
place in the pond.
No wells were constructed that would permit samples to be obtained from the zone
between 9 ft and 175 ft. Nevertheless, the trend lines on Figure 12 are believed to show
the change in concentration with depth of several of the metals. The slope of the lines
indicate that the different "metal fronts" had advanced into the saprolite in the eight
years prior to 1981 to a distance of only a few feet. This distance gives a "rate of
advance" of less than 1 ft/yr. This number, if in error, is too large because (1) the
center of the screens were used in determining the thickness of non -ash material at
wells 3-2 and 3-3, rather than the depth to the upper part of the screen and sand pack,
and (2) "pumping" of the wells to obtain samples may have resulted in downward
movement of water from zones above the screened interval. Relative to the rate of
advance, studies by Gibb and Cartwright (1982) of the migration of zinc, cadmium,
copper, and lead through geologic materials at zinc smelter sites in Illinois show rates of
advance of a few tenths of a foot per year.
In an effort to confirm the rate of advance discussed above and, if possible, to obtain a
more precise answer, an analysis was made of only the data obtained from the wells in
the ash and in the saprolite beneath the ash ponds at the Allen plant. It was desirable in
this analysis to be able to combine the data from both the active and inactive ponds.
43

Figure 15 is a graph showing the concentrations of calcium and strontium versus years
per foot in samples obtained in late February and late March 1981 from the wells listed
in Table 8. The factor of years per foot was obtained for each well by dividing the years
TABLE 8
Basic Data Used in the Preparation of Figure 15
since each pond began receiving ash by the distance between the bottom of each pond
and the center of the well screens. Data used in preparing the figure are shown in
Table 8.
The range in each constituent for the two analyses for each well are shown as a bar.
The trend lines drawn through the bars are believed to show the decrease in
concentration of calcium and strontium as the effluent moves from the pond through the
saprolite. The tendency of these lines to approach the horizontal at the higher
concentrations suggests that three to four years is required to essentially "use up" the
cation -exchange capacity of each foot of saprolite. Or, in other words, the "metal
front" moves at a rate of a few tenths of a foot per year.
It is important to note here that this rate applies only to the upper, clay -rich, part of
the saprolite. Because the clay content of the lower part is less than that of the upper
part, the ion -exchange capacity of the lower part is smaller and the metal front, once it
EV
Depth
Time of
Well
Below
Travel
Calcium
Strontium
No
Ash (ft)
(years)
Years/ft
(ppm)
(ppm)
3-2A
In ash
—
—
126-129
3.7-4.13
3-1
5.5
24
4
59.5-60.7
3.6-4.7
3-2
5
8
1.6
15.8-17
0.25-0.274
3-3
9
8
0.9
2.5-377
0.007-0.008
since each pond began receiving ash by the distance between the bottom of each pond
and the center of the well screens. Data used in preparing the figure are shown in
Table 8.
The range in each constituent for the two analyses for each well are shown as a bar.
The trend lines drawn through the bars are believed to show the decrease in
concentration of calcium and strontium as the effluent moves from the pond through the
saprolite. The tendency of these lines to approach the horizontal at the higher
concentrations suggests that three to four years is required to essentially "use up" the
cation -exchange capacity of each foot of saprolite. Or, in other words, the "metal
front" moves at a rate of a few tenths of a foot per year.
It is important to note here that this rate applies only to the upper, clay -rich, part of
the saprolite. Because the clay content of the lower part is less than that of the upper
part, the ion -exchange capacity of the lower part is smaller and the metal front, once it
EV

U
0
Yoam per foot
S.I
Concentrations of Calcium and Strontium Versus Years Per Foot for
Wells Open to Ash and to Saprolite Beneath the Ash Ponds at the
Allen Powerplant of Duke Power Company
Figure 15
45

moves into the lower part of the saprolite, will move faster. How much faster is not
known but it might be as much as a foot per year.
It has been noted previously that the saprolite at Riverbend appears to contain a smaller
content of clay than the saprolite at Allen. This would result in a somewhat faster rate
of movement of the metal front at Riverbend. However, it is highly doubtful that the
metal front at Riverbend has advanced as much as 15 to 20 ft into the saprolite in the
30 years the pond has been in use. Thus, it is believed that the front is still well above
the bedrock surface and, as far as lateral movement in the saprolite is concerned, is still
a considerable distance from the river.
In concluding this discussion, it is important to note that the locations of both the Allen
and Riverbend ash ponds closely conform to the criteria listed by Heath and Lehr (1987)
for solid -waste disposal sites. The ash ponds at both sites are underlain by thick
sections of clay -rich saprolite having sufficient cation -exchange capacity to immobilize
the metals in the ash effluent from many decades of operation. The ash ponds are also
adjacent to a major stream that has a very large dilution capacity. Based on Figure 15,
it appears unlikely that the Riverbend ash pond will remain active long enough for
metals derived from the ash to reach the river. However, one of the objectives of this
study was to determine what effect unmodified ash effluent reaching the Catawba
through the groundwater system would have on the quality of the river. This topic is
discussed in Section 6.
46

4. General Surface Water Quantity and Quality
The location of flow and water quality monitoring
stations along the reaches of the Catawba River/Mountain
Island Lake is shown in Figures 16 and 17 (the latter
indicating river miles and the location of the Charlotte
water intake). The Riverbend ash basin discharges near
Mile 7 (see Figure 17).
A statistical summary of the flows through Cowans
Ford and Mountain Island dams is presented in Table 9.
Such data were independently verified by direct communication
with U.S.G.S. staff and direct access to computerized HISARS
records. The 12 -year Duke Power Company record (1973-1985)
yields a mean flow of about 2866 cfs. Earlier estimates were
from 2600 cfs (41 -year HISARS record) to 2800 cfs, depending
upon mean flow values reported for Rock Hill, South Carolina
(4251 cfs or 4495 cfs). The lower figure incorporates flows
of pre -impoundment years (which would tend to lower the mean.
STORET water quality files were accessed for : (1) Station
2142648 (just downstream from Lake Norman, Catawba River at
N.C. Hwy 73 near Hicks, Lat: 35 25 38 and Long: 80 57 24,
record length 1968 to 1987) ; and Station 2142808 (same
latitude as northern Charlotte city limits, Catawba River at
N.C. Hwy 27 near Thrift, Lat: 35 17 53 and Long: 81 00 14,
record length 1968 to 1987). Duke Power supplied water
quality records at : Station 288 (about 1 mile upstream of
the Riverbend ash basin), Station 277 (about 3 miles downstream
from the ash basin discharge), and at the Riverbend ash basin
outfall. The data provide a reasonable profile of water quality
along the Catawba. Statistical summaries are presented in
Tables 10 through 15.
In the figures that follow, some distortion of the actual
profiles is built-in due to unequal time intervals in the
47

recorded data (but most have been properly adjusted). It is
nevertheless useful to view the magnitude of the observed
iron concentrations along the river, as well as the values of
other parameters at the ash basin outfall. The iron limit is
1.0 mg/l (N.C. Administrative Code, 1985): iron concentrations
at the Riverbend ash basin effluent are below this number and
would be further diluted by a 1,:400 ratio (from a mean effluent
discharge of 7.1 cfs and mean river flow of 2866 cfs). A
dilution ratio of about 394 was calculated 5 miles downstream
for a continuous discharge at the source (later section). Iron
concentrations at the STORET station near Hicks (just downstream
from Lake Norman) reached 2.4 mg/l once (probably during anoxic
conditions in Lake Norman from late August to November),
and 1.5 rag/l at Thrift (near Charlotte latitude). The
mean iron concentrations in downstream order are : 0.32,
0.76, 0.23 (ash basin effluent partially mixed with Catawba
River), 0.35 and 0.41 mg/l (see Table 15). The profile
is illustrated in Figure 18. It should be noted that the
mean iron concentrations are higher upstream of the ash
basin. The mean effluent flow of 6.63 cfs (Table 12) is
the result of sampling at equal time intervals : the 7.1
cfs value represents the average of all reported
observations, and it was used in the transport model because
it yields a more conservative prediction (less dilution).
Figures 19 to 31 depict all other known flow and water
quality data related to the Riverbend Site. Most of the
variation in dissolved oxygen at both Duke Power stations 278
and 277 (river miles 6 and 10.5 approximately) seems to be
explained largely by the temperature variation (see regression
analysis below).
48

Regression Output:
Constant 12.45
Std Err of Y Est 0.68
R Squared 0.89
No. of Observations 150.00
Degrees of Freedom 148.00
X Coefficient(s) -0.20
Std Err of Coef. 0.01
Tables 11 and 13 and Figure 22 reveal a slight increase
in pH downstream from the ash basin (from 6.8 upstream to
7.04 downstream). Water quality is well within established
State standards throughout these reaches.
i. n

50
;E

Figure 117 . Map of Mountain Island Reservoir study area_
S1

J xii W�
TABLE 9. Statistical Summary of Flows �
Through Cowans Ford and Mtn.
Island Dams
Statistic Cowans Ford Mtn. Island
Dam (cfs) Dam (cfs)
Sample size 4639 4636
Average 2845.98 2885.74 _
Median 2092 2267.5
Mode 80 80
Geometric mean
Variance 7.80655E6 6.62877E6
Standard deviation 2794.02 2574.64
Standard error 41.022 37.8133
Minimum 0 0
Maximum 28954 11981
Range 28954 11981
Lower quartile 419 417.5
Upper quartile 4688 4810
Interquartile range 4269 4392.5
Skewness 1.35048 0.766127
Standardized skewness 37.5512 21.296
Kurtosis 3.95845 -0.325744
Standardized kurtosis 55.0341 -4.52733
52

TABLE 10. Statistical Summary of Catawba River
Water Quality at Hicks, N. C.,
STORET Station No. 2142648
Statistic
Temperature
pH
Arsenic
(deg. C)
(std units)
(µg/1)
Sample size
210
185
68
Average
17.3895
6.85189
10.1471
Median
18.5
6.8
10
Mode
26
6.9
10
Geometric mean
15.7503
6.83951
10.1025
Variance
47.2315
0.170988
1.47059
Standard deviation
6.87252
0.413507
1.21268
Standard error
0.474249
0.0304017
0.147059
Minimum
3
6
10
Maximum
29
7.9
20
Range
26
1.9
10
Lower quartile
11
6.5
10
Upper quartile
24
7.2
10
Interquartile range
13
0.7
0
Skewness
-0.193008
0.139583
8.24621
Standardized skewness
-1.14185
0.775076
27.7609
Kurtosis
-1.22982
-0.703723
68
Standardized kurtosis
-3.63785
-1.95381
114.461
Statistic
Cadmium
Lead
Mercury
(Ag/1)
(gg/1)
(µg/1)
Sample size
78
78
64
Average
37.7692
86.4103
0.89375
Median
50
100
0.5
Mode
50
100
0.5
Geometric mean
27.7473
73.3362
0.464913
Variance
387.842
922.011
6.84091
Standard deviation
19.6937
30.3646
2.61551
Standard error
2.22987
3.43812
0.326939
Minimum
2
10
0.2
Maximum
100•
100
21
Range
98
90
20.8
Lower quartile
20
100
0.2
Upper quartile
50
100
0.5
Interquartile range
30
0
0.3
Skewness
-0.382606
-1.94997
7.45485
Standardized skewness
-1.3795
-7.03072
24.3474
Kurtosis
0.0137749
2.12744
57.8397
Standardized kurtosis
0.024833
3.8353
94.4519
53

TABLE 11. Statistical Summary of Catawba River
Water Quality at Station 278, Upstream
of Duke Power Riverbend Site
Statistic Temperature
Dissolved
pH
(mg/1)
Oxygen
(gg/1)
Sample size
(deg C)
(mg/1)
(std units)
Sample size
19
19
19
Average
16.7316
8.98421
6.82105
Median
17.6
8.9
6.9
Mode
25.8
6.5
6.7
Geometric mean
13.5942
8.73557
6.81463
Variance
83.4001
4.63029
0.0895322
Standard deviation
9.13236
2.15181
0.299219
Standard error
2.09511
0.493659
0.0686456
Minimum
3
5.4
6.1
Maximum
28.5
12.4
7.2
Range
25.5
7
1.1
Lower quartile
6.8
7.3
6.7
Upper quartile
26
11.5
7.1
Interquartile range
19.2
4.2
0.4
Skewness
-0.190007
0.133655
-1.15878
Standardized skewness
-0.33812
0.237841
-2.06206
Kurtosis
-1.53364
-1.1858
1.05092
Standardized kurtosis
-1.36457
-1.05507
0.935061
Statistic Manganese
Chromium
Copper
(mg/1)
(N•5/1)
(gg/1)
Sample size
12
4
4
Average
0.0483333
28.8
52.7
Median
0.05
2.5
4
Mode
0.05
0.2
2.8
Geometric mean
0.0420671
3.41194
9.63041
Variance
8.87879E-4
2931.65
9643.99
Standard deviation
0.0297973
54.1447
98.2038
Standard error
8.60174E-3
27.0724
49.1019
Minimum
0.02
0.2
2.8
Maximum
13�
110
2 0 0
Range
0.11
109.8
197.2
Lower quartile
0.03
1.2
3
Upper quartile
0.05
56.4
102.4
Interquartile range
0.02
55.2
99.4
Skewness
2.02414
1.99746
1.99954
Standardized skewness
2.86257
1.63092
1.63261
Kurtosis
5.22732
3.99167
3.99844
Standardized kurtosis
3.69628
1.62959
1.63236
54

TABLE 11. Statistical Summary of Catawba River
Water Quality at Station 278, Upstream
of Duke Power Riverbend Site (Cont.)
Statistic
Lead
Zinc
(gg/1)
(fig/1)
Sample size
4
4
Average
15.15
50.15
Median
2.3
30.3
Mode
1
10
Geometric mean
4.12221
32.9004
Variance
706.19
2936.57
Standard deviation
26.5742
54.1901
Standard error
13.2871
27.0951
Minimum
1
10
Maximum
(-`b5-)
130
Range
54
120
Lower quartile
1.55
18.1
Upper quartile
28.75
82.2
Interquartile range
27.2
64.1
Skewness
1.99657
1.78784
Standardized skewness
1.63019
1.45977
Kurtosis
3.98875
3.36984
Standardized kurtosis
1.6284
1.37573
55

TABLE 12. Statistical Summary of Ash Basin
Effluent Water Quantity and Quality,
Duke Power Riverbend Site
k b 15-17 b '?-
Statistic
Z
Statistic
Selenium
(ug/1)
Chromium
(gg/1)
Mercury
(µg/1)
Statistic
Flow
Iron
Arsenic
Average
(cfs)
(mg/1)
(mg/1)
Sample size
97
155
121
Average
6.62875
0.235935
0.12119
Median
7.5814
0.19
0.057
Mode
9.4384
0.1
0.045
Geometric mean
5.183
0.190631
0.0585635
Variance
12.9126
0.031397
0.29332
Standard deviation
3.59342
0.177192
0.54159
Standard error
0.364856
0.0142324
0.0492355
Minimum
0.15472
0.01
4E-3
Maximum
16.555
1.15
6�
Range
16.4003
1.14
5.996
Lower quartile
3.4038
0.11
0.035
Upper quartile
9.4384
0.29
0.,094
Interquartile range
6.0346
0.18
0.059
Skewness
-0.0463643
2.3268
10.8355
Standardized skewness
-0.186421
11.8263
48.6592
Kurtosis
-0.871952
6.95531
118.546
Standardized kurtosis
-1.75296
17.6757
266.178
k b 15-17 b '?-
Statistic
Z
Statistic
Selenium
(ug/1)
Chromium
(gg/1)
Mercury
(µg/1)
Sample size
121
88
88
Average
4.62066
9.34886
1.19773
Median
5
2.3
0.1
Mode
5
50
0.1
Geometric mean
3.92827
2.98314
0.383448
Variance
7.55599
291.757
4.19976
Standard deviation
2.74882
17.0809
2.04933
Standard error
0.249892
1.82083
0.21846
Minimum
0.2
0.5
0.1
Maximum
<20�,
50
13 4
Range
19.8
49.5
13.3
Lower quartile
2
1.2
0.1
Upper quartile
5
4.1
2
Interquartile range
3
2.9
1.9
Skewness
1.99863
1.99099
3.86023
Standardized skewness
8.97532
7.6249
14.7836
Kurtosis
7.5056
2.06318
18.4327
Standardized kurtosis
16.8528
3.9507
35.2959
56

TABLE 12. Statistical Summary of Ash Basin
Effluent Water Quality,
Duke Power Riverbend Site (Cont.)
,o 1 5 Al
Statistic pH Temperature Copper
(std units) (deg C) (mg/1)
Sample size
155
91
155
Average
7.67935
17.8352
0.0445419
Median
7.5
18.3
0.05
Mode
7.5
20
0.05
Geometric mean
7.65732
15.745
0.0295483
Variance
0.348532
61.4376
1.14893E-3
Standard deviation
0.590366
7.83822
0.0338958
Standard error
0.0474193
0.821668
2.72258E-3
Minimum
6.2
2.8
IE -3 -
Maximum
9.5
30
0.125
Range
3.3
27.2
0.124
Lower quartile
7.3
11
0.012
Upper quartile
8
25
0.05
Interquartile range
0.7
14
0.038
Skewness
0.59465
-0.121861
0.583951
Standardized skewness
3.0224
-0.474579
2.96802
Kurtosis
0.172883
-1.27452
-0.844369
Standardized kurtosis
0.439353
-2.48178
-2.14582
Z
sv
2s
Statistic
Cadmium
Nickel
Lead
(µg/1)
(µg/1)
(µg/1)
Sample size
88
88
88
Average
3.53295
22.8443
25.0557
Median
0.3
8.15
1
Mode
0.2
5
1
Geometric mean
0.545998
11.344
2.57906
Variance
53.5491
1092.74
3620.05
Standard deviation
7.31773
33.0566
60.1668
Standard error
0.780072
3.52385
6.41381
Minimum
0.1
1
1
Maximum
Cj
d00
(_20 6)
Range
22.9
99
199
Lower quartile
0.2
5.25
1
Upper quartile
0,7
14.8
2.35
Interquartile range
0.5
9.55
1.35
Skewness
1.8904
1.86027
2.37291
Standardized skewness
7.23968
7.1243
9.08756
Kurtosis
1.64984
1.69603
4.10615
Standardized kurtosis
3.15921
3.24765
7.86269
57

TABLE 12. Statistical Summary of Ash Basin
Effluent Water Quality,
Duke Power Riverbend Site (Cont.)
so A I
Statistic Zinc
(µg/1)
Sample size
88
Average
12.8932
Median
10.1
Mode
10
Geometric mean
9.16569
Variance
94.4503
Standard deviation
9.71855
Standard error
1.036
Minimum
1
Maximum
57
Range
56
Lower quartile
5.05
Upper quartile
18.35
Interquartile range
13.3
Skewness
1.45091
Standardized skewness
5.55655
Kurtosis
3.90366
Standardized kurtosis
7.47494
m

TABLE 13. Statistical Summary of Catawba River
Water Quality at Station 277, Downstream
from Duke Power Riverbend Site
Statistic Temperature
Dissolved
pH
(mg/1)
Oxygen
(Ag/1)
Sample size
(deg. C)
(mg/1)
(std units)
Sample size
29
29
29
Average
18.1448
9.37931
7.04483
Median
17.9
8.8
7.1
Mode
6.9
11.5
6.9
Geometric mean
14.5869
9.14392
7.04014
Variance
109.605
4.8067
0.0675616
Standard deviation
10.4693
2.19242
0.259926
Standard error
1.94409
0.407122
0.0482671
Minimum
3.5
6.6
6.3
Maximum
31.5
15
7.5
Range
28
8.4
1.2
Lower quartile
7.3
7.6
6.9
Upper quartile
28.4
11.5
7.2
Interquartile range
21.1
3.9
0.3
Skewness
-0.0479907
0.616404
-0.428173
Standardized skewness
-0.105507
1.35515
-0.941332
Kurtosis
-1.76087
-0.384441
1.14615
Standardized kurtosis
-1.93562
-0.422593
1.2599
Statistic Manganese
Chromium
Copper
(mg/1)
(gg/1)
(Ag/1)
Sample size
22
4
4
Average
0.0354545
6.975
3.375
Median
0.03
3.3
4.15
Mode
0.02
0.3
0.4
Geometric mean
0.0310165
2.7945
2.39722
Variance
3.78355E-4
90.2292
4.0425
Standard deviation
0.0194513
9.4989
2.0106
Standard error
4.14704E-3
4.74945
1.0053
Minimum
0.01
0.3
0.4
Maximum
0.09
21
4.8
Range
0.08
20.7
4.4
Lower quartile
0.02
1.25
2.2
Upper quartile
0.05
12.7
4.55
Interquartile range
0.03
11.45
2.35
Skewness
1.22714
1.81654
-1.83742
Standardized skewness
2.3498
1.4832
-1.50025
Kurtosis
1.55931
3.38517
3.49969
Standardized kurtosis
1.49292
1.38199
1.42874
59

TABLE 13. Statistical Summary of Catawba River
Water Quality at Station 277, Downstream
from Duke Power Riverbend Site (Cont.)
Statistic
Lead
Zinc
(ug/1)
(µg/1)
Sample size
4
4
Average
3.225
19.475
Median
2.45
10.95
Mode
1
10
Geometric mean
2.5329
15.2959
Variance
6.8825
313.502
Standard deviation
2.62345
17.706
Standard error
1.31173
8.853
Minimum
1
10
Maximum
7
46
Range
6
36
Lower quartile
1.55
10
Upper quartile
4.9
28.95
Interquartile range
3.35
18.95
Skewness
1.52417
1.98484
Standardized skewness
1.24448
1.62062
Kurtosis
2.63041
3.94711
Standardized kurtosis
1.07386
1.6114

TABLE 14. Statistical Summary of Catawba River
Water Quality at Thrift, N. C.,
STORET Station No. 2142808
Statistic Temperature
pH
Arsenic
Mercury
(deg. C)
(std units)
(µg/1)
Sample size
236
208
56
Average
18.7657
6.7899
10.7143
Median
20
6.8
10
Mode
27
6.8
10
Geometric mean
16.7718
6.77513
10.4537
Variance
56.1885
0.195695
10.3896
Standard deviation
7.4959
0.442374
3.22329
Standard error
0.487941
0.0306731
0.43073
Minimum
2
5.1
10
Maximum
30
7.8
30
Range
28
2.7
20
Lower quartile
12
6.5
10
Upper quartile
25
7.1
10
Interquartile range
13
0.6
0
Skewness
-0.371736
-0.517578
4.9294
Standardized skewness
-2.33139
-3.04742
15.0596
Kurtosis
-1.0807
0.909028
25.6132
Standardized kurtosis
-3.38886
2.67611
39.1248
Statistic
Cadmium
Lead
Mercury
(µg/.l)
(lig/1)
(gg/1)
Sample size
66
71
63
Average34.9394
87.0423
0.601587
Median
0
00
0.5
Mode
50
100
0.5
Geometric mean
24.9299
74.1933
0.407864
Variance
359.412
898.27
0.804352
,
Standard deviation
18.9582
29.9711
0.896857
Standard error
2.33359
3.55692
0.112993
Minimum
2
10
0.2
Maximum
'50
100
6.3
Range
48
90
6.1
Lower quartile
20
100
0.2
Upper quartile
50
100
0.5
Interquartile range
30
0
0.3
Skewness
-0.616447
-2.03838
4.84899
Standardized skewness
-2.04452
-7.01195
15.7125
Kurtosis
-1.35167
2.48096
27.3782
Standardized kurtosis
-2.2415
4.2672
44.3578
61

TABLE 15. Statistical Summary of Catawba River
Iron Concentrations In Downstream Order
Statistic @ Hicks @ Sta. 278 Ash Basin Effluent
(µg/1) (mg/1) (mg/1)
Sample size
38
12
155
Average
320.813
0.766667
0.235935
Median
220
0.3
0.19
Mode
100
0.3
0.1
Geometric mean
195.836
0.36504
0.190631
Variance
161203
1.60424
0.031397
Standard deviation
401.501
1.26659
0.177192
Standard error
65.132
0.365632
0.0142324
Minimum
0.9
0.1
0.01
Maximum
2400
4.5
1.15
Range
2399.1
4.4
1.14
Lower quartile
100
0.2
0.11
Upper quartile
400
0.6
0.29
Interquartile range
300
0.4
0.18
Skewness
4.05073
2.77276
2.3268
Standardized skewness
10.1941
3.92128
11.8263
Kurtosis
19.946
7.98366
6.95531
Standardized kurtosis
25.0982
5.6453
17.6757
Statistic @
Sta. 277
@ Thrift
(mg/1)
(µg/1)
Sample size
23
38
Average
0.356522
405.789
Median
0.2
300
Mode
0.1
300
Geometric mean
316.519
Variance
0.240751
110717
Standard deviation
0.490664
332.742
Standard error
0.10231
53.9778
Minimum
0
100
Maximum
1.9
1500
Range
1.9
1400
Lower quartile
0.1
200
Upper quartile
0.3
400
Interquartile range
0.2
200
Skewness
2.29041
1.90865
Standardized skewness
4.48436
4.80333
Kurtosis
4.49326
3.2099
Standardized kurtosis
4.39865
4.03903
62

W
C7
z
0
F
Q
U
z
0
z
0
x
2.2
2
L8
L6
L4
1.2
1
0.8
0.6
0.4
0.2
0
DUKE POWER CO. RIVERBEND SITE
CATAWBA RIVER, HICKS TO THRIFT
STA 278
SITE
A 277
0 2 4 6 8 10 12
DISTANCE IN RIVER MILES
❑ PROFILE OF MEAN + (MEAN plus STD DEMI
Figure 18. Profile of Iron Concentrations In Catawba River
14 16

14
13
12
n
to
9
8
7
6
s
4
3
2
0
70[77
14
M
12
n
to
9
s
7
6
5
4
3
2
l
D
RIVERBEND ASH BASIN
DUKE POWER OOK4PANV
92877 122277 32278 61478 90678 MM8
DATE
Figure 19. Ash Basin Flow
RIVERBEND ASH BASIN
DUKE POWER GC*.W kW
t0279 50279 91779 t[780 51280 904130 t2
DATE
64

14
13
12
tt
10
9
8
s
s
4
3
2
1
0
ml
14
13
12
if
to
9
8
7
s
s
4
3
2
t
0
- RIVERBEND ASH BASIN
BIKE POWER CC*.WANY
42781 81781 12(781 4582
DATE
Figure 19. Ash Basin Flow (Cont.)
RIVERBEND ASH BASIN
DUKE POWER COWANY
10383 42783 81783 EW983 33084 72084 11684
DATE
65

S
S
CL
RIVERBEND ASH BASIN
DUKE POWER COKWAM'
Is
u
n
Q
n
w
9
8
7
6
6
4
3
2
I
0
7OU7 92877 E2= 32278 61478 90678 U3078
DATE
Figure 20. Time History of Effluent pH
RIVERBEND ASH BASIN
DUKE POWER CC&V k Y
D
V
n
�o
9
8
7
6
5
4
3
2
l
10279 50279 9i779 Mo SUM 90480 t231£3[l
DATE
rM

14
93
12
U
a
9
8
7
6-
54 5-
4
3-
2
1
1038(
14
0
12
R
a
9
e
7
s
s
4
3
2
1
RIVERBEND ASH BASIN
DUKE POWER CC* -PAM'
42781 8fM 121781 41582 80582
DATE
Figure 20. Time History of Effluent pH (Cont.)
RIVERBEND ASH BASIN
13UKE POWER C:OWAW
10383 42783 81783 120983 33084 72084 10584
DATE
67

20
16
co
12
0
Ef
Flow ( MGD) ; PH ( std units)
77 79 81 83 85 87 89
Time a calendar years
Figure 21. Ash Basin Flow Versus pH

7,6
M
PN Upstream and Downstream of Site
76 78 80 82 84 86 88
Time (calendar years)
Figure 22. Time History of pH Upstream and
Downstream of Site
6,7
16,4
t
s
6,1
PN Upstream and Downstream of Site
76 78 80 82 84 86 88
Time (calendar years)
Figure 22. Time History of pH Upstream and
Downstream of Site

RIVERBEND ASH BASIN
DUM POWER CO WAVY
35
30
25
J
0
70[77 92877 @?277 32278 61478 90578
DATE
Figure 23. Time History of Effluent Temperature
RIVERBEND ASH BASIN
DUKE POWER COWANY
5
10279
50279 9[779 U780
DATE
70
51260 90480 123L80

RIVERBEND ASH BASIN
DUKE PO%VM C XWANY
0
10181 42781 SUM 121781 41582 80582 112682
DATE
Figure 23. Time History of Effluent Temperature (Cont.)
RIVERBEND ASH BASIN
DUKE POV/M COi1WANY
35
r=
s
0--
10383
42783 817133 120983
DATE
71
33084 72084 111584

V
N
(X i0o) STORET ; Iron @ Hicks, N, C,
24
1 20
0 16
n
I
40 80 120. 160 200 240
Time Sequence
Figure 24. Iron Concentrations At Hicks, N.C.

W
Iron Upstream and Downstream of Site
78 80 82 84 86 88
Time ( cal endar years)
Figure 25. Iron Concentrations Upstream and Downstream
of Site

1,2
I
r
�
V
0,6
m 0,4
2
Ri verbend Ash Basin Iron concentrations
77 79 81 83 85 87 89
Time (calendar years)
Figure 26. Time History of Effluent Iron Concentrations

V
In
� � i0o) STORET ; Iron � Thr i ft a M
73 75
77 79 81
Time ( calendar years)
Figure 27. Iron Concentration At Thrift, N.C.

V
Q�
I
m
f
1
0, 2
0,12
0,09
0,06
Manganese Upstream and Downstream
76 78 80 82 84 86 88
Time ( oal endar years)
Figure 28. Manganese Concentrations Upstream (Sta 278)
and Downstream ('Sta 277) of Site

I
Z 50
n 40
v
V
a 30
0
R 1 YERDI D Ash Basin Zinc
77 79 81 83 85
Time ( calendar years)
Figure 29. Time History of Effluent Zinc Concentrations

suoz;Pa::Ueou00 uintuaTaS ;uan-tJJa Jo szzTjaS auiTj • pS aanSTJ
(5jee npua tso) aw tZ
68
L8
98
£8
S8
6L
ZL
b j
n
a l ua l as u t SSI qq gggjj� I a
3Y
oz
w
t
8

w
N
'
so
60V
1
M Of
R 1 VERDIND Ash Basin Nickel
X
77 79 K 83 85
Time ( oa1 endar years)
Figure 31. Time Series of Effluent Nickel Concentrations

B
5. Analysis of Potential Surface Water Impacts
As noted earlier, the mean effluent flow rate of
the Riverbend Ash Basin is 7.1 cfs. The mean river
flow was estimated at 2600 cfs for a 41 -year record
including pre -impoundment flows, and 2866 cfs for a
shorter (12 -year) more recent record. These values
immediately suggest a dilution ratio from 1/366 to
1/404. A detailed examination of mixing phenomena
in the Catawba River, including a two-dimensional
transport model, is presented in the next sections.
5.1 Vertical and Transverse Mixing
From the river cross-sections supplied by Duke Power,
an average cross-sectional area of flow between reaches
7 through it of 1388 meters squared (14,940 square feet) is
obtained (see Figures 32 through 34) :
u = Q / A = 0.19 ft/sec (very sluggish)
avg avg avg
Using only the average flow area between reaches 7 and 8,
u = 2800 cfs / 6566 sq ft = 0.43 ft/sec (still slow)
avg
The vertical mixing (diffusion) coefficient may be
estimated by :
E = 0.067 d u
v
where d = stream depth
u = shear velocity
= (g d S)"
91
(about 1 meter roughly
at Mile 7)

g = gravitational constant
S = river bed slope (approx. 0.001 from
the cross-sections)
With u* = 0.324 fps,
2
E = 0.071 ft /sec
v
The transverse mixing (diffusion) coefficient may be
estimated by :
2
E = 1.5 d u = 4.5 ft /sec
t
(1.5 for sharp bends, irregular channels)
( d about 2 meters average for reaches)
( shear velocity = 0.46 fps for d = 2 m)
and the distance downstream from the contaminant source
for complete mixing by :
2
L = 0.4 u W / e
avg t
where W = stream width average (approx. 297 meters, or 974
feet for reaches 7 through 11). Therefore,
2
L = 0.4 (0.19) (974) / 4.5 (5280)
= 3.03 miles
Because of the tremendous bend between Mile 9 and 10, it
is very likely that the ash basin discharge has mixed in the
cross-section by the time the plume reaches the Charlotte
water intake (slightly over 4 miles downstream from the
ash basin outfall) as predicted above.

A longitudinal dispersion coefficient can be estimated
(in the absence of dye experiments) by :
2 2
E L avg = 0.011 (u ) W / (d u*)
z
E 125 ft /sec
L
which is used later in applying the advective-dispersive
equations to predict pollutant transport. The mixing time is
proportional to the square of the average length traversed
divided by the mixing coefficient. For our case, the transverse
mixing time is :
: 2
(W/dt / ev) _ (297/2) / (4.5/0.071)
= 348
times the vertical mixing time. In other words, complete
vertical mixing would occur within 50 feet of the outfall.

145 m
0
:::A:r:ea= 595 m2
2 m -
4 m -
Mile 0
115 m
0
Area = 390 m2
2 m -
4 m
Mile 2
115 m
0
Area = 340 m2
2 m
4 m
Mile 4
120 m
0
Area = �3112
2 m -
4 m
Mile 6
C
2 m
115 m
0
Area 500 m2
2 m -
4 m -
Mi le 1
120 m
0
Area = 475 m2
2 m -
4 m -
Mile 3
120 m
0
Area 365 m2
2 m -
4 m -
Mile 5
Figure 32. River Cross -Sections
Through Mile 7
355 m
Mile 7
Note: Cross sections looking, downstream.
83

195 m
0 Area a525 mz
2 m
4 m Mile 8
205 m
C
2 m - \ Area = 1275 m2
4 m
6 m -
Mile 9
Figure 33. River Cross -Sections
at Miles 8 and 9
Note: Cross sections looking downstream.

0
2
m -
4
m '
6
m `
00
0
2 m
4 m -
6 m -
400 m
Area = 2140 m2
Mile 10
330 m
Area - 230$ m2
Note: Cross sections looking downstream.
Mile 11
Figure 34. River Cross -Sections at
Miles 10 and 11

5.2 ANALYTICAL, STEADY-STATE, CONTINUOUS TWO-DIMENSIONAL
(VERTICAL LINE SOURCE) MODEL, SSCLS
The two dimensional advective-dispersive equation for
a continuous vertical line source is a parabolic partial
differential equation given by :
'2X
2
EX6X + Eyy2 - USX - KC = St
The time derivative goes to zero at steady-state, and the
above equation reduces to a second order ordinary differential
equation :
EXdxz2 + Eydzy2 - UX - KC = 0
The solution, in terms of the modified Bessel function of
the second kind of order zero, for a side discharge is :
q exp (XU l
i2EX u l2 02)
where
�(EYX2 + EXY2) (U2
2EY + 4KEXEy)
02 = 4 EX EY
and
q=PCIQ1

and K0(2B2) is computed using polynomial approximations
given by Abramowitz and Stegun (NBS Handbook of Mathematical
Functions, 1964).
A dilution ratio may be computed at any point across
and along the stream (y,x) by normalizing the predicted
concentration with the input concentration : C/C i . A
computer program was developed (SSCLS) and results were
checked with a hand-held electronic calculator. If 28 is
2
much greater than 1, the Bessel function can be approximated
by an exponential function. Therefore,
Ko(202) = 4� exp( -202)
2
and
q exp NUE
X
C(x,y) = exp( -202)
np da„gy n 402
or
L-2—Ex J � exp( -2 02)
C, n �Vg XEY �vl 4.02
Thus, for a depth averaged across the stream cross-section
at the source (Mile 7, d = 9.432 feet), for y = 487
av
feet (stream width average over 5 -mile stretch), zero
decay, and for the transverse and longitudinal mixing
values calculated earlier :
uavg = 0.19 ft/sec
Ex = 125 ft2/sec z EL
Ey = 4.5 ft2/sec z et
WN

x = 26,400 feet = 5.0 miles
Qi = 7.1 cfs
the concentration ratio is
C(x,y=487)/C. = 0.002534 = 1 / 394.568
1
or about a 1:400 dilution ratio near the Charlotte water
intake. The dimensionless concentration profile at the
stream centerline is presented in Figure 35. Although
the pollutants reach the stream centerline about 0.75
miles downstream, complete mixing across the entire river
width occurs at 3.03 miles and beyond. The computed
profile is presented in Figure 36. The concentration of
any pollutant at the source (in any units) would simply
be multiplied by the concentration ratio at the
corresponding distance downstream. For example, on
October 19, 1987 samples taken from the Riverbend effluent
weir were analyzed for copper, iron, arsenic and selenium
content, their concentrations 5 miles downstream would be
as follows :
Concentration Concentration Concentration
At Source Ratio 5 Miles
(Ash Basin) Downstream
Cu, < 0.1 mg/l 0.002534 < 0.0002534 mg/1
Fe, < 0.1 mg/1 0.002534 < 0.0002534 mg/l
As, 131 µg/l 0.002534 0.331954 µg/l
Se, < 2 µg/l 0.002534 < 0.005068 gg/1
These constitute conservative estimates since
sinks such as precipitation due to oxidation (which could be
approximated in the decay coefficient K) have been set to
zero.
::

I
!1
3
2
i
.l,
DUKE POWER COe RIVERBEND SITE
Concentration X 1000
0 1 2 2 4 5
Distance Downstream , Miles
Figure 35. Concentration Ratio Profile Downstream from Site

Figure 36. Computed Profile of Concentration
Ratio, Dawmtream of Ash Basin
DILUTION RATIO BELOW SOURCE
FOR PRISTINE RIVER = 404.662000
Concentration Ratio At 5.0 MILES = .002534
Dilution Ratio At Same Distance = 1 TO 394.568100
DISTANCE =
.0000 MILES
CONC. =
.000000
DISTANCE =
.1000 MILES
CONC. =
.001700
DISTANCE =
.2000 MILES
CONC. =
.002205
DISTANCE =
.3000 MILES
CONC. =
.002646
DISTANCE =
.4000 MILES
CONC. =
.002996
DISTANCE =
.5000 MILES
CONC. =
.003255
DISTANCE =
.6000 MILES
CONC. =
.003437
DISTANCE =
.7000 MILES
CONC. =
.003558
DISTANCE =
.8000 MILES
CONC. =
.003634
DISTANCE =
.9000 MILES
CONC. =
.003676
DISTANCE =
1.0000 MILES
CONC. =
.003694
DISTANCE =
1.1000 MILES
CONC. =
.003694
DISTANCE =
1.2000 MILES
CONC. =
.003682
DISTANCE =
1.3000 MILES
CONC. =
.003661
DISTANCE =
1.4000 MILES
CONC. =
.003633
DISTANCE =
1.5000 MILES
CONC. =
.003601
DISTANCE =
1.6000 MILES
CONC. =
.003566
DISTANCE =
1.7000 MILES
CONC. =
.003529
DISTANCE =
1.8000 MILES
CONC. =
.003491
DISTANCE =
1.9000 MILES
CONC. =
.003452
DISTANCE =
2.0000 MILES
CONC. =
.003412
DISTANCE =
2.1000 MILES
CONC. =
.003373
DISTANCE =
2.2000 MILES
CONC. =
.003334
DISTANCE =
2.3000 MILES
CONC. =
.003296
DISTANCE =
2.4000 MILES
CONC. =
.003258
DISTANCE =
2.5000 MILES
CONC. =
.003221
DISTANCE =
2.6000 MILES
CONC. =
.003185
DISTANCE =
2.7000 MILES
CONC. =
.003149
DISTANCE =
2.8000 MILES
CONC. =
.003115
DISTANCE =
2.9000 MILES
CONC. =
.003081
DISTANCE =
3.0000 MILES
CONC. =
.003048
DISTANCE =
3.1000 MILES
CONC. =
.003016
DISTANCE =
3.2000 MILES
CONC. =
.002984
DISTANCE =
3.3000 MILES
CONC. =
.002953
DISTANCE =
3.4000 MILES
CONC. =
.002924
DISTANCE =
3.5000 MILES
CONC. =
.002894
DISTANCE =
3.6000 MILES
CONC. =
.002866
DISTANCE =
3.7000 MILES
CONC. =
.002838
DISTANCE =
3.8000 MILES
CONC. =
.002811
DISTANCE =
3.9000 MILES
CONC. =
.002785
DISTANCE =
4.0000 MILES
CONC. =
.002759
DISTANCE =
4.1000 MILES
CONC. =
.002734
DISTANCE =
4.2000 MILES
CONC. =
.002710
DISTANCE =
4.3000 MILES
CONC. =
.002686
DISTANCE =
4.4000 MILES
CONC. =
.002663
DISTANCE =
4.5000 MILES
CONC. =
.002640
DISTANCE =
4.6000 MILES
CONC. =
.002618
DISTANCE =
4.7000 MILES
CONC. =
.002596
DISTANCE =
4.8000 MILES
CONC. =
.002575
DISTANCE =
4.9000 MILES
CONC. =
.002555
DISTANCE =
5.0000 MILES
CONC. =
.002534
r•1

6. Analysis of Potential Groundwater Contributions
Scarce data are available on actual groundwater quality
conditions at the Riverbend Plant. However, the relevant hydro -
geologic conditions are likely to be very similar to those at
Plant Allen, which have been extensively studied. The ash
introduced into the settling basins contains many metal
constituents, some of them in fairly high concentrations.
However, the majority of this material appears to remain in
insoluble form, and the dewatered ash will eventually be removed
from the settling basin and landfilled. The important questions
are whether leaching from this ash will adversely affect
groundwater downgradient from the basin, and, if this occurs,
whether surface water conditions will be adversely impacted in
the Catawba River.
The analysis thus naturally divides into three parts:
1) Potential impact on groundwater near the basin.
2) Potential groundwater contribution of contaminants to
Catawba River/Mountain Island Lake.
3) Comparison of potential groundwater contribution to
permitted surface discharges from the ash basin.
6.1 Potential Impact on Groundwater
The potential impact on groundwater must be considered in
two sections. We first consider the trace metals, which include
toxic contaminants. The naturally dominant metals iron and
manganese must be considered separately.
Trace Metals
The trace metals examined in the ash include barium,
cadmium, chromium, copper, arsenic, selenium, mercury and lead.
Arsenic is apparently of particular concern. It should first be
noted that monitoring at Plant Allen shows no instances of
violation of groundwater standards for these constituents either
underneath the ash basins or downgradient, despite the fact that
several of these metals (arsenic, selenium, barium, cadmium,
chromium) are found in relatively high concentrations in dry
91

fly ash. Based on the Plant Allen experience, it is thus expected
that these trace metals will not pose any significant hazard to
groundwater.
The concentrations of any trace metals which do
reach groundwater are expected to be reduced by two
processes: precipitation reactions, resulting in insoluble
compounds (which may be of particular importance for arsenic in
the presence of iron), and reversible adsorption to the clay
particles of the saprolite. It appears that the adsorptive
capacity of the saprolite is by itself sufficient to prevent
contamination problems from these trace metals, which seems to be
borne out by the experience at Plant Allen.
Where adsorption can be described by a linear, reversible
isotherm, adsorption can be readily accounted for in a
groundwater transport model by use of a retardation coefficient
(R), which describes the reduced apparent velocity of the adsorbed
contaminant relative to the movement of the groundwater:
R = V/V
c
where V is the velocity of the groundwater, and V the apparent
c
velocity of the contaminant. For fast reversible adsorption
described by a linear isotherm, Freeze and Cherry (1979) provide
an approximation for R in terms of the soil -water distribution
coefficient, K
d
(1 + 4Kd) <= R <= (1 + 10Kd)
where Kd, expressed in units of ml/g, is the rate of change of
the adsorbed mass per unit mass of soil (S) divided by the rate
of change of solute concentration (C):
K = dS/dC
d
Unfortunately, the behavior of most metal ions is not
M

accurately described by a linear isotherm. Instead, adsorptive
behavior for these constituents is more accurately described by
the nonlinear Langmuir isotherm (EPRI, "Chemical Attenuation
Rates..."):
S = (KLA mC)/(1+KLC)
where:
S = moles adsorbed at equilibrium per gram of solid
A = maximum adsorptive capacity of the soil
,„
= Langmuir adsorption constant
C = total concentration in solution at equilibrium.
From this relationship,
dS/dC = K L A /(1+K C)Z
m L
This value will be approximately linear for concentrations which
are small relative to the adsorptive capacity of the soil, which
appears to hold for all the trace metals under consideration.
Further, the rate of adsorption will be greater at lower
concentrations.
The EPA report on Plant Allen gives estimates of the
adsorptive capacity of several types of native soil found at the
plant for various trace metals (as 1/g), analyzed at varying solution
concentrations (Table 5.4, EPA, 1985). These allow computation of an
approximate linear soil -water distribution coefficient over the
concentration range of the metals found beneath the plant. Taking
the minimum reported values from these tests, and applying the
relationship given by Freeze and Cherry we can estimate minimum
approximate values for the retardation coefficients, as follows:
93
Minimum
Minimum
K (ml/9)
R
----------------------------------------------
Arsenic
13600
54401
Selenium
510
2041
Cadmium
240
961
Copper
830
3321
Nickel
140
561
Vanadium
140
561
93

These values would indicate that these metals
would be nearly immobile in the groundwater. It is possible
that the estimates may be high and the observed reductions may
include precipitation as well as adsorption. However, coefficients
for Langmuir type adsorption for arsenic given in EPRI, "Chemical
Attenuation Rates..." (1986), suggest that the value of K for
arsenic in the range of observed concentrations should be somewhat
greater than 2000 ml/g, which is still quite high.
From the data presented in Figure 15 and Table 8, Section
3.6, a rate of contaminant migration (V ) of 0.3 feet per year
c
for the trace metals was estimated. From Table 5, Section 3.6,
a groundwater velocity of 180 feet per year (z 0.5 ft/day) is
obtained. Thus, the resulting retardation coefficient is 600.
It is interesting to note that this value is slightly above the
minimum values reported above for nickel and vanadium. This
R value was then used in a two-dimensional analytical contaminant
transport model (TDAST, Javandel et al., 1984) within a
groundwater quality advisory system (Medina, et al., 1987).
The Riverbend Plant has currently been in operation for 30 years,
and the minimum distance from the ash basin to the Catawba River
is about 500 feet. Since the clay content of the saprolite at
Riverbend appears to be slightly smaller than at Allen, the R
value was also reduced by a factor of 3 (very conservative) to
R = 200. At the reduced retardation coefficient, breakthrough
of contaminants to the river along the shortest flowline would
still require over 1600 years !
For a conservative, worst case analysis the
following values were used :
V = 0.493 ft/day (180 ft/yr)
EL = 9.3 x 10-2 ft2/day
94
Table 5, Section 3.5
Figure 9.4, Freeze
and Cherry (1979)

ET
= 9.3
x 10-3
ft2/day
1/10
of
EL value
R =
200
--->
(600/3)
Table
5,
Table 8
HL = 1250 ft. (half-length of the basin profile
parallel to the river, see Figures 3
and 9, Sections 2.2 and 3.2 respectively)
Output of the transport model, given these parameters,
follows, giving predictions of percent of under -site
concentrations at years 1987, 2007 and 2037. These results
indicate that no measurable concentration of the retarded
contaminants will reach the river along the shortest flowlines
even in the next 50 years. It is expected that the ash basin
will have been removed from service long before this date, and
the residual ash landfilled, so that the input concentrations
will be decreasing well before this date is reached.
ANALYTICAL TWO-DIMENSIONAL CONTAMINANT TRANSPORT MODEL, TDAST
The model TDAST provides an analytical solution to steady state
flow in a two-dimensional Cartesian coordinate system, and is docu-
mented in Javandel et al. (1984). If the flow is coincident with the
x axis, and the longitudinal and transverse components of the disper-
sion tensor are assumed independent of position and designated by D
L
and DT, the general governing equation for a confined, homogeneous,
isotropic aquifer can be written as:
2 z
DI5X + DT
5y2 - v8X - UC = R St
where lambda represents decay and R is the retardation coefficient.
For a particular solution we first assume that the medium is
initially free of the solute, and that at a certain time a strip type
source of length 2a, orthogonal to the flow direction, is introduced
along the y axis. If the source concentration diminishes exponen-
tially with time, the initial and boundary conditions are:
95

C(O,y,t) = Coe -at -a - y < a
C(O,y,t) = 0 lyl > a
Where the source "strip" is arranged orthogonal to the direction
of flow, an analytical (but not closed form) solution is presented by
Cleary and Ungs (1978) as:
C(x,y,t) = 4COX
exp l D - at]
t/R
z
exp AR - aR + 4D )T 4D z T ] T-3/2
L L
0
I 4a+ y
erf1
a y)+ erf�JJST
TT
The results of applying TDAST to the Riverbend Site are presented
in Figures 37 and 38. Even for the year 2037, no measurable pollutant
concentrations are reaching the river (x = 500 feet, 152 meters).
Figure 37 represents a three-dimensional view of the contaminant
plume for the year 2037 (80 years in operation) for R = 200 : the
pollutant concentrations (in percent, 100 percent at the source) are
measurable at 21.3 meters (70 feet), with another 430 feet to go to
reach the river, across a 1260 foot front (y-axis is parallel to the
river). The computed output file is presented in Figure 38 for the
years 1987, 2007 and 2037. The contaminants are immobile through
the year 2007, showing some movement by 2037.

--
Figure 37. Three - Dimensional View
of Percent Concentration,

Figure 38. State of North Carolina Groundwater
Advisory System, Output File
SITE NAME: Riverbend
SITID = RIVRBN FILES= A:RIVRBN.OUT A:RIVRBN.SIT
Application of analytical model TDAST
Run # 1
YEAR 1987
Time = 10958. days ( 30 years)
HALF LENGTH OF SOURCE: 381.00
SOURCE CONC.: 100.00 %
LONG. DISPERSIVITY: .86400E-02 (m2/d)
TRANS. DISPERSIVITY: .86400E-03 (m2/d)
X VELOCITY: .15033 (m/day)
RETARDATION COEF: 600.00
DECAY FACTOR OF SOURCE: .00000 (1/day)
DECAY FACTOR OF SOLUTE: .00000 (1/day)
Aquifer concentration at source input directly,
not calculated from leaching
CONCENTRATION RESULTS (percent)
Y AXIS (PERPENDICULAR TO FLOW) (meters)
.000
27.4
54.9
82.3
110.
X---------------------------------------
.000
100.
100.
100.
100.
100.
10.7
.140E-42
.140E-42
.140E-42
.140E-42
.140E-42
21.3
.000
.000
.000
.000
.000
32.0
.000
.000
.000
.000
.000
42.7
( .000
.000
.000
.000
.000
53.3
.000
.000
.000
.000
.000
64.0
.000
.000
.000
.000
.000
74.7
.000
.000
.000
.000
.000
85.3
.000
.000
.000
.000
.000
96.0
.000
.000
.000
.000
.000
107.
.000
.000
.000
.000
.000
117.
.000
.000
.000
.000
.000
128.
.000
.000
.000
.000
.000
139.
.000
.000
.000
.000
.000
149.
.000
.000
.000
.000,
.000

1-A I-• N N N fD OJ N O Cn oP W N N
�P W NI-�OO�Cn�P�P WNNI�00
• 10 O W N O W N O W N O'JC
N N
OOOOOO6006666h'0 N
OOOOOOo0000004O 1 ,P
00000000000000. 1
tri I
I I
.P I
N I
I
O O O O O O O O O O
6,0 O N O 1 0
0000000000000,P-01"
00000000000000-1.
t+i I
I I
.P I
N I
00000000000oo1-�0
O O O O O O O O O O O O O $P O
00000000000000.
t=i
I
�P
N
N
OOOOOOOOOOOO) ID --O
0000000000000,010
0 0 0 0 0 0 0 0 0 0 0 0 0 0
C�1
I
�P
N
O O O O O O O O O O O O O 6:D
000000000000000
000000000000000
NI-+NNNc0W0 40 04 W Nh+•
�P W NI�0Q10410• W NNF-+00
• 10 00 O W N O W N O W N O ,'x•
O O O O O O O O O O 66 6;1 O
00000000000000.0
00000000000000.
t3J
I
�P
N
00000000000001-•0
0000000000000,010
0 0 0 0 0 0 0 0 0 0 0 0 0 0
tz1
I
.P
N
wO O O �
N O O O O O O O O O O h- 0
cD 000000000000040
0 0 0 0 0 0 0 0 0 0 0 0 0 0
txl
I
N
wf..
Cn O O O O O O O O O O O O O N O
N 00000000000000.0
• 00000000000000
C4
I
.P
N
W 000000000000010
lA 0000000000000010
•
00000000000000-
N 0000000000000
W
O
> W
p. rt
ca
�• rt
M CD
O
1 O
`c M
cn z
'C O
rt rt
(D tr
CJ
a
0111
G O
V W
G G
rt a
� O
µ Ij
N O
cD G
-- a
Ci E
O a
0 rt
rt M
I

Figure 38. State of North Carolina Groundwater
Advisory System, Output File (Cont.)
YEAR 2007
Time = 18263. days
RETARDATION COEF: 600.00
CONCENTRATION RESULTS
Y AXIS (PERPENDICULAR TO FLOW) (meters)
100
.000
27.4
54.9
82.3
110.
X---------------------------------------
.000
100.
100.
100.
100.
100.
10.7
.298E-14
.298E-14
.298E-14
.298E-14
.298E-14
21.3
.000
.000
.000
.000
.000
32.0
.000
.000
.000
.000
.000
42.7
.000
.000
.000
.000
.000
53.3
.000
.000
.000
.000
.000
64.0
.000
.000
.000
.000
.000
74.7
.000
.000
.000
.000
.000
85.3
.000
.000
.000
.000
.000
96.0
.000
.000
.000
.000
.000
107.
.000
.000
.000
.000
.000
117.
.000
.000
.000
.000
.000
128.
.000
.000
.000
.000
.000
139.
.000
.000
.000
.000
.000
149.
.000
.000
.000
.000
.000
137.
165.
192.
219.
247.
X---------------------------------------
.000
x.100.
100.
100.
100.
100.
10.7
.298E-14
.298E-14
.298E-14
.298E-14
.298E-14
21.3
.000
.000
.000
.000
.000
32.0
J .000
.000
.000
.000
.000
42.7
.000
.000
.000
.000
.000
53.3
.000
.000
.000
.000
.000
64.0
.000
.000
.000
.000
.000
74.7
.000
.000
.000
.000
.000
85.3
.000
.000
.000
.000
.000
96.0
.000
.000
.000
.000
.000
107.
.000
.000
.000
.000
.000
117.
.000
.000
.000
.000
.000
128.
.000
.000
.000
.000
.000
139.
.000
.000
.000
.000
.000
149.
.000
.000
.000
.000
.000
100

I-+ ►-+ N N N W C>D N 01 0 jr, W N N
�P W N�-OOI-+00',T1.�CHr(n
tD lD co N N p
• 0 CA3 b4 W 1.40
00
zC: H
Cb+
z
�JrCn• 0
o
tl0• til (D
tea
> 0 0
---------------
H G
k
H
H H O0
00oo
0►CCnroz
000000000o►
off•
z.. ►bpgn
0000000000► -+N• 00 1
OAC
y
ms u- N
0000000000(7)()PI.- I
OH
H
0 :0 W
CzJ [TJ I
W
O
O CD H N
z
ti HC O
W O
..
hil►-'CH O
ON I
b
xu
•• CnHHl-`
i
ro
CA) p
rn l-� ►-A
0000000606•�AMOO1N4m
N
c
r
N 0
00000000001-"• 00 1•
z
H
O O
0000000000MO4. • 1
4U
CD
0 p,
L=] L=] 1
H
pp (y
I I I
n
O co 0) ae "C
WO I
C
O^014 N
rn N I
r
�oo
1
�
I
t1 O m
00 00000000►-� I�M00 1
SPH
&41CA 0
00000000001-+N• Oo I
• O
X -ON
Ooo00000000)(0�P• • 1
to
W
0 N I
$
N\
N
W
0
w
• rn N N I
00
•--
•.�
0 0 0 0 0 O O O O O hA 0) 0 0 1
N-
00000000001 -+N• 00 1
• 5
00000000000)(D&+• • I
WCD
� C�1 _I
rf•
I I I
CD
W O 1
K
0) N I
I
N
1
0) N N I
N
O O O O O O O O O O N N 01 O O
N
00000000001 -AN• 00
O
000000000omwol •
I I
W O
0 N
NNNNNCD W N0 0gjhP W NN•
ork. w" + 0 0 0 0. 006 W N N M+ O O
to CD 00 0.4 �4 . . . .p
0w0.40 O W N0>4
I.A N
O O O O O O O O O O O O O N O a
0000000000000c00 1 1P
000000000000OC1D• I
m I
�P I
O O O O O O O O O O O O O N O I O
0000000000000001"
000000000000000• 1
tzl I
I-` 1
� I
I
• IJ 1 w
O O O O O O O O O O O O O N O I N
00000000000000010
0000000000000co. I
I� I
I -A 1
iP 1
O O O O O O O O O O O O O N O
000000000000000
0000000000000CD•
I
N
�P
1
000000000000000 co
000000000000000 4%.
000000000000000
W
co
CA
Crt
w r+
CC1 fD
O
CA z
0 K
Zr
0
I✓ O
rt N
ro w
>r �
rt p►
N Li
wti
i� O
cD �
CL
O rF
rt. IY

is 4 +
a) r.
++ O
roU
3 --
b
s~
� m
or -I
is "
CD w
ro +J
0 0
.rl a
H 4-J
O 0
;4 O
ro
U
6
W
++ ++
ii (D
O 'Jr
z (n
w >4
O I~
O N
a-+ •r(
ro D
4-Jro
(1)4
CD
c0
O
P
b1
r1
W
rn
N
N
0)
ri
1n
(D
ri
r
M
ri
h (D
O CO
I I
W W
•dl0)(D0000000000
00 .c.4H0000000000
O O W H" 0 0 0 0 0 0 0 0 0 0
H ri 0 . . . . . . . . . . . .
N (D
0 cr)
i I
ww
• OMW0000000000
OO •N'r'10000000000
OO(DT-i.-,0000000000
ri"W . . . . . . . . . . . .
N (D
O co
I I
W W
•'qmw 0000000000
00 •N.-10000000000
oowv-i.-10000000000
ri .-1 W . . . . . . . . . . . .
h (D
O c0
I I
W W
110(00000000000
00-04-10000000000
0ow""0000000000
r'1 ri (D . . . . . . . . . . . .
h w
O c0
I I
W W
• V O W 0 0 0 0 0 0 0 0 0 0
00-04-10000000000
OOWHT-10000000000
ri ri (D . . . . . .
xot-MONMONMO
•rrco
o rnrn
O O ri N NCO d' d' (C) (D O ri N CO cr
• .-1 N M 10 0 (D h 0 0) r1 .-i ri ri ri
r
LO
M
c0 h h
.-i O ri
I I I
WWw
0 10 1O D 0 0 0 0 0 0 0 O O 0 0
Od'�d'00000000000
OCOHW 0000000OO o0
. . . . . . . . . . . . . . .
h (D
O co
I 1
Ww
•"00W0000000000
00 •N ri 0000000 000
OO(DHH0000000000
ri 'i (D . . . . . . . . . . . .
I r to
I O CO
I I I
i W W
• I •d+0)(D0000000000
0) 1 00 •N ri000000O000
N 100(D.-1.-10000000000
CO I H ri (D . . . . . . . . . . . .
I
I
I h (D
I O c0
I I I
I W W
• I • 00)(D 0000000000
N 1 00 Nv-10000000000
O 1 0 0 w r -I .-1 0 0 0 0 0 0 0 0 0 0
M I •-i ri (D . . . . . . . . . . . .
I
I
I h (D
I O C7
I I I
I W W
• 1 10)(D0000000000
10 1 00 •N x-10000000000
r OO(DH"0000000000
N .-i ri (D . . .
---------------
O O .-1 N N CO (D O •-I N CO NP
• ri N 0 V tp (D h CO 0 .1 .-( . ri ri
N
O
.-1

Iron and Manganese
Analysis of the impact on iron and manganese
concentrations requires somewhat different assumptions. This is
primarily due to the fact that these are among the dominant metal
ions in the native saprolite. Further, these two metals are found
in large quantities in the ash, and these are the only
constituents for which violations of the groundwater standards
were noted beneath Plant Allen.
Because the saprolite may contain large concentrations
of iron and manganese (depending on the parent material), it is
to be expected that the residual capacity to adsorb these ions
may be reduced. The data available on adsorption rates are not
conclusive, largely dealing with artificial substrates, but
measured rate constants suggest that the marginal rate of
adsorption for these metals may be rather low at the higher
observed concentrations beneath Plant Allen (EPRI, 1986).
Instead, the dominant factors controlling concentrations of these
metals are likely precipitation/ dissolution reactions. For
instance, metastable ferrosic hydroxide [Fe 3(OH)8] will control
maximum iron concentrations under most redox conditions.
Similarly, MnCO 3 (rhodochrosite) and various Mn oxides are
suspected to control solubility of manganese, although the data
are less conclusive here. It is suspected that a substantial
proportion of both the Mn and Fe which enters the groundwater
system will be immobilized by precipitation reactions, thus
reducing the concentrations away from the site.
For a conservative analysis, it may be assumed that
a substantial proportion of the manganese and iron which reaches
the groundwater will be transported with the groundwater, with
little retardation of apparent velocity.
The Riverbend site has been in operation for 30 years. If
103

little retardation occurs, and precipitation of the metals is not
considered, flow rates are such that it must then be assumed that
these metals will be present along the groundwater flowlines
between the site and the river at essentially the same
concentrations as beneath the site. (However, we should emphasize
that we suspect that concentrations will actually be reduced by
precipitation reactions).
6.2 Potential Groundwater Contribution of Contaminants to
Catawba River/Mountain Island Lake
The analysis presented in Section 6.1 suggests that the
only metal contaminants likely to be of concern downgradient from
the site will be iron and manganese. This contention is supported by
surface water quality monitoring downstream from the site (Duke
Power station 277, USGS/EPA station Thrift, No. 02142808), which
does not show violation of standards for the trace metals. A few
violations of iron standards have been noted downstream. However,
iron is ubiquitous in the region, and it is of interest to note that
iron concentrations are on average higher upstream from the site.
In any case, it is instructive to consider the potential
contribution of these metals due to groundwater flow from the
ash basin to the river, given no attenuation. The rate of flow
from the ash basin to the river is estimated at about 5.85 x 104
3
ft /day. In comparison, the river flow is estimated to average
8 3
2.476 x 10 ft /day : a 1/4200 dilution factor.
The highest concentrations observed in the
groundwater beneath the Plant Allen basin are:
Fe 25.9 mg/l
Mn 14.0 mg/l
From the analysis presented earlier, dilution on reaching the river
will be about 1/400. For both metals, this dilution would be
104

sufficient to reduce the metal concentrations well below the
surface water standards. It should further be noted that the bottom
waters in the river adjacent to the site are generally not
anoxic, and thus further precipitation of these metals should be
expected when groundwater enters the river.
6.3 Comparison of Groundwater and Surface Contributions
5 3
The average surface discharge is 6.13 x 10 ft /day,
an order of magnitude greater than the estimated groundwater
flow from the basin. The relative contribution of the groundwater
will depend on the degree to which contaminants are attenuated
before reaching the -river. For the worst case scenario with iron,
no attenuation, the relative contributions would be:
6
Surface: 4.10 x 10 7 mg/day
Groundwater: 4.29 x 10 mg/day
Thus the groundwater contribution of iron could range up
to 10 times the surface contribution, but this is likely an
overestimation. The 1:400 dilution upon reaching the river and
flowing down to the Charlotte water intake would still result in
concentrations below the permitted surface discharge.
By comparison, the worst case scenario for arsenic, with
no adsorption or precipitation, gives the following
contributions:
6
Surface: 2.06 x 10 gg/day
6
Groundwater: 2.65 x 10 µg/day
Thus for arsenic the maximum potential contribution of
the groundwater would be close to the average surface contribution.
Further, the sum of these contributions would still be
considerablv less than the permitted surface discharge. It is
105

even more likely that the groundwater contribution of arsenic will be
undetectable.
6.4 CONCLUSION
It appears highly unlikely that any subsurface or surface -water
quality problems exist as a result of the ash pond as demonstrated by
the absence of a detectable increase in anv metals downstream from
the Riverbend Plant. Relative to the effect of ash -pond effluent on
the quality of the Catawba River, analysis of streamflow records past
the Riverbend Plant indicates that the average flow of the river is
about 400 times the rate of surface outflow from the pond and about
370 times the combined surface and groundwater rate. Although the
velocity of the Catawba River is relatively show past the Riverbend
site, modeling using conservative mixing coefficients suggests that
complete transverse mixing of the pond effluent occurs within about
three miles of the ash pond. Due to the small concentration of
metals in the surface outflow from the pond, and the large dilution
factor, metals contained in the effluent do not cause a detectable
increase in concentration once complete mixing has occurred.
The effect of groundwater seepage on stream quality was determined by
modeling the flow through the groundwater system using a retardation
factor only 1/3 of the value estimated at the Allen Plant. Results
indicate that no measurable concentration of the metals that are
subject to ion exchange and other delaying reaction will reach the
Catawba within the next 50 years (by 2037).
106

REFERENCES
Arthur D. Little, Inc.. 1985. Full-scale field evaluation of waste disposal from coal-
fired electric generating plants. EPA -600/7-85-028a, June 1985, Volume I, Section 5.
Davis, S.N. and R.J.M. Dewiest. 1966. Hydrology: New York, John Wiley & Sons, Inc.
Freeze, R.A. and J.A. Cherry. 1979. Groundwater: Englewood Cliffs (N.J.), Prentice -
Hall, Inc.
Gibb, J.P. and W.K. Cartwright. 1982. Retention of zinc, cadmium, copper, and lead by
geologic materials. Illinois Dept. of Energy and Natural Resources Cooperative
Groundwater Report. 9.
Heath, R.C. and J.H. Lehr. 1987. A new approach to the disposal of solid waste on
land. Ground Water, V. 25, No. 3, May -June 1987.
LeGrand, H.E. 1952. Geology and groundwater in the Charlotte area, North Carolina.
North Carolina Department of Conservation and Development Bulletin No. 63.
Roche, D.P., A. Gnilka, and J.E. Harwood. 1984. Investigations of coal ash disposal and
its impact upon groundwater. Duke Power Company.
Tetra Tech, Inc. 1985. Groundwater data analyses at utility waste disposal sites.
Electric Power Research Institute. EPRI EA -4165.
107

Appendix I. Source Code Listing for Model SSCLS,
Steady -State Continuous Line Source

PROGRAM SSCLS
C ..... Steady State Continuous Line Source
C... Boundary Source (Soln X 2):
C...
C... q' exp (XU/2Ex)
C... C =------------------ KO (2 s)
C... n 6 ✓ ExEy
C........................................
C... Dr. M. A. Medina , Duke University
C........................................
REAL*8 KO,IO
REAL*4 K,XAXIS(51),YAXIS(51)
REAL XC,YC
CHARACTER*7 FMT
CHARACTER*l TITLE(75)
INTEGER*4 SIZE, INTARY(4000)
CHARACTER*30 LABL,LABLY,LABLX
CHARACTER*6 NAMEL
CHARACTER*l NUL,ESC,BEL
COMMON LABL,LABLY,LABLX,NAMEL,XC,YC
COMMON /GRACOM/ SIZE, INTARY
CALL FILES(IN)
CALL FILES(IOUT)
FMT = '(/1X,A)'
RHO = 1.94
PI = 4. * ATAN(1.)
SIZE = 4000
WRITE(*,FMT) 'TITLE
READ(IN,1) TITLE
1 FORMAT(75A1)
WRITE(IOUT,3) TITLE
3 FORMAT(//5X,75A1)
WRITE(*,3) TITLE
WRITE(*,FMT) 'U,X,Y,K,Ex,Ey
WRITE(*,FMT)
.'1.........2.........3.........4.........5.........6........x'
READ(IN,5) U,X,Y,K,Ex,Ey
5 FORMAT(3F10.O,E10.0,2F10.0)
WRITE(*,111) U,X,Y,K,Ex,Ey
111 F0RMAT(3F10.2,E10.2,2F10.2)
WRITE(*,FMT) 'Havg,Qavg,Cavg,Qriv :'
WRITE(*,FMT)
.'1......... 2 ......... 3 ......... 4 ........ if
READ(IN,7) Havg,Qavg,Cavg,Qriv
7 FORMAT(4F10.0)
WRITE(*,113) Havg,Qavg,Cavg,Qriv
113 FORMAT(4F10.2)

XX=X/5280.
C.... Dilution Ratio at Source
DRO = Cavg / ((Qavg * Cavg)/(Qavg + Qriv))
WRITE(IOUT,11) DRO
it FORMAT(/5X,'DILUTION RATIO BELOW SOURCE '/
5X,'FOR PRISTINE RIVER=1,F15.6)
BETA2 = SQRT((Ey*X*X + Ex*Y*Y)*(U*U*Ey + 4.*Ex*Ey*K))
BETA2 = BETA2/(4.*Ex*Ey)
C... Dilution Ratio Downstream End ... C/Cavg
q = RHO * Cavg * Qavg
CCO = EXP(X*U / (2.* Ex)) * KO(2.* BETA2) * (q / Havg)
CCO = CCO / (PI * RHO * SQRT(Ex * Ey)) / Cavg
DR = 1./CCO
WRITE(IOUT,13) XX,CCO,DR
13 FORMAT(/5X,'Concentration Ratio At 1,F5.1,' MILES=',Fll.6/
/5X,'Dilution Ratio At Same Distance = 1 TO',F15.6/)
C... CONCENTRATION PROFILE
WRITE(*,FMT) 'Computing Profile ...'
DX = X / 50.
X = 0.
DO 50 I=1,51
IF (X.EQ.0.) THEN
IF (Y.EQ.0.) CXY = Cavg / DRO
IF (Y.GT.O.) CXY = 0.
ELSE
BETA2 = SQRT((Ey*X*X + Ex*Y*Y)*(U*U*Ey + 4.*Ex*Ey*K))
BETA2 = BETA2/(4.*Ex*Ey)
CXY = EXP(X*U / (2.* Ex)) * KO(2.* BETA2) * (q / Havg)
CXY = CXY / (PI * RHO * SQRT(Ex * Ey))
ENDIF
XAXIS(I) = X / 5280.
YAXIS(I) = CXY * 1000.
WRITE(IOUT,51) XAXIS(I),CXY
X = X + DX
50 CONTINUE
51 FORMAT(1OX,'DISTANCE=',F10.4,' MILES1,2X,1CONC.=',F10.6)
CLOSE(IN)
CLOSE(IOUT)
WRITE(*,FMT) 'Closing Input,0utput Files ...'
WRITE(*,FMT) 'ENTER TITLE ... A30
READ(*,91) LABL
91 FORMAT(A)
WRITE(*,FMT) 'ENTER Y-AXIS LABEL ... A30
READ(*,91) LABLY

WRITE(*,FMT) 'ENTER X-AXIS LABEL ...
READ(*,91) LABLX
WRITE(*,FMT) 'ENTER LEGEND LABEL ...
READ(*,91) NAMEL
WRITE(*,FMT) 'ENTER LEGEND X,Y COORD.
WRITE(*,FMT) '1....2....'
READ(*,93) XC,YC
WRITE(*,FMT) 'ENTER MIXING DISTANCE,
WRITE(*,FMT) '1....2....'
READ(*,93) XL,XR
93 FORMAT(2F5.0)
WRITE(*,FMT) 'Plotting ...`
CALL PLOTS(XAXIS,YAXIS,5I,XL,XR)
C--------------------------
C RESET MODE TO 80x25
C--------------------------
NUL = CHAR(0)
ESC = CHAR(27)
BEL = CHAR(7)
READ(*,*)
WRITE(*,19) NUL,ESC,BEL
19 FORMAT(2A1,1[=3h1,A1)
C---------------------------------
STOP
END
A30
A6 .'
M ...'
MILES ... XL,XR
REAL*8 FUNCTION KO(X)
C.... Modified Bessel Function of Order Zero, Second Kind
C.... Polynomial Approximation
C.... Ref.: Abramowitz and Stegun, p. 379.
REAL*8 IO
IF (X.GT.O. .AND. X.LE.2.) THEN
X2 = X/2.
KO = - ALOG(X2) * IO(X) - 0.57721566
KO = KO + 0.42278420 * X2 ** 2 + 0.23069756 * X2 ** 4
KO = KO + 0.03488590 * X2 ** 6 + 0.00262698 * X2 ** 8
KO = KO + 0.00010750 * X2 ** 10
KO = KO + 0.00000740 * X2 ** 12
RETURN
ENDIF
XX2 = 2. / X
KO = 1.25331414 - 0.07832358 * XX2 + 0.02189568 * XX2 ** 2
KO = KO - 0.01062446 * XX2 ** 3 + 0.00587872 * XX2 ** 4
KO = KO - 0.00251540 * XX2 ** 5 + 0.00053208 * XX2 ** 6
KO = KO / SQRT(X) / EXP(X)
RETURN
END

REAL*8 FUNCTION IO(X)
C.... Modified Bessel Function of Order Zero, First Kind
C.... Polynomial Approximation
C.... Ref: Abramowitz and Stegun, p. 378
T = X/3.75
IF (X.GE.(-3.75) .OR. X.LT.3.75) THEN
IO = 1. + 3.5156229 * T * T + 3.0899424 * T ** 4
IO = IO + 1.2067492 * T ** 6 + 0.2659732 * T ** 8
IO = IO + 0.0360768 * T ** 10 + 0.0045813 * T ** 12
RETURN
ENDIF
IO = 0.39894228 + 0.01328592 / T + 0.00225319 / T ** 2
IO = IO - 0.00157565 / T ** 3 + 0.00916281 / T ** 4
IO = IO - 0.02057706 / T ** 5 + 0.02635537 / T ** 6
IO = IO - 0.01647633 / T ** 7 + 0.00392377 / T ** 8
IO = IO / SQRT(X) * EXP(X)
RETURN
END
SUBROUTINE FILES(IO)
C...............................................
C..... TO OPEN FILE FILIO AS I/O FILE IO .....
C...............................................
CHARACTER*20 FILIO,NAME*3
WRITE(*,'(//A\)') ' ENTER NAME OF I/O FILE '
READ (*,'(A)') FILIO
WRITE(*,'(A\)') ' ENTER I/O FILE UNIT NUMBER (I2) : '
READ (*,'(BN,I2)') IO
WRITE(*,'(A,I2)') ' FORMATTED I/O FILE UNIT NUMBER = ',IO
WRITE(*,'(A\)') ' ENTER FILE STATUS (1=NEW,O=OLD) : '
READ (*,'(BN,I1)') IS
IF(IS.EQ.1) THEN
OPEN (IO,FILE=FILIO,STATUS='NEW')
NAME _ 'NEW'
WRITE(*,11) FILIO,NAME
ENDIF
IF(IS.EQ.0) THEN
OPEN (IO,FILE=FILIO,STATUS='OLD')
NAME _ 'OLD'
WRITE(*,11) FILIO,NAME
ENDIF
11 FORMAT(1X,'FILE ___> ',A,' IS ',A)
RETURN
END

SUBROUTINE PLOTS(X,Y,N,XL,XR)
implicit integer*2 (a -z)
integer*2 status
REAL X(N),Y(N),XL,XR,XC,YC
REAL XD(2),YD(2)
CHARACTER*1 TITL1(30),TITLY(30),TITLX(30),NAME(6)
COMMON TITLI,TITLY,TITLX,NAME,XC,YC
C Dummy curve to fill-in bottom area.
XD(1)=XL
XD(2)=XR
YD(1)=0.0
YD(2)=0.0
C PLOT: Open Plotting System
status = popnps()
C PLOT: Assign Plotting System Output Device
status = ppsots(7,'DISPLAY')
C Define TWO data sets (dummy = 2)
C PLOT: Define X/Y data set.
status = pdsxy(1,N,X,Y)
status = pdsxy(2,2,XD,YD)
C Define all data sets to be line charts.
C PLOT: Define data set chart type to be line.
DO 100 i=1,2
status = pline(i)
100 CONTINUE
C -----PLOT: Set existence and location of legend.
status = plloc(1,1,XC,YC)
C Set legend alignment
status = plalin(2,2)
C Set legend font (Complex Roman Font).
status = pltfnt(102)
C Set legend height.
status = plthgt(2)
C Name data set for legend.
status = pdsnms(1,6,NAME)
C----------------------------------------------------------
C Set to fill region between data sets 1 and 2 (dummy).
C PLOT: Fill between two data sets.
status = pdspar(2,1)
C Set the color of the dummy to be color index 4.

C PLOT:
Set output primitive color for data set.
status
= pdsclr(2,4)
C PLOT:
Set output primitive style for data set.
C status
= pdsstl(2,5)
C PLOT:
Set visibility of data set (hide dummy).
status
= pdsvis(2,0)
C Set color index 4 to be blue-green (75%,25%).
E PLOT:
Set color representation.
status
= pscrep(4,0.0,25.0,75.0)
C PLOT:
Define axis title string.
status
= ptaxs(1,30,TITLX)
status
= ptaxs(2,30,TITLY)
C PLOT:
Define title string.
status
= pthds(30,TITL1)
C PLOT: Output the currently defined chart.
C THIS DRAWS THE CHART.
status = ppltit()
C------------------------------------
C PLOT: Close output device.
status = pclios(7,'DISPLAY')
C PLOT: Close plotting system.
status = pclsps()
RETURN
END