HomeMy WebLinkAboutNC0025348_Report_20180514/V(Mp2 -50
New Bern Wastewater
Reclamation Facility
Predictive Modeling Analysis;
Martin Marietta Quarry (East Lake)
Water Reclamation Facility,
City of New Bern,
Craven County, North Carolina
Prepared For:
Rivers and Associates, Inc.
107 East 2ntl Street
Greenville, NC 27858
Prepared By
Groundwater Management Associates, Inc.
4300 Sapphire Court, Suite 100
Greenville, North Carolina 27834
And
-a Groundwater Management Associates, Inc.
2205-A Candun Drive
Apex, North Carolina 27523
GMA Project #21342
May 14, 2018
` N C
PER
_,
TABLE OF CONTENTS
EXECUTIVESUMMARY.......................................................................................... ES-1
1.0
INTRODUCTION.................................................................................................1
2.0
SCOPE OF WORK...............................................................................................1
3.0
REGIONAL HYDROGEOLOGIC SETTING .............................................................2
3.1
Regional Geology.............................................................................................. 2
3.2
Hydrogeology of the Site.................................................................................. 5
4.0
REVIEW OF PREVIOUS STUDIES.......................................................................6
._
4.1
The Pohlig Model............................................................................................6
4.2
Matyiko Master's Thesis.................................................................................6
4.3
Additional Data Sources.................................................................................7
a..
5.0
PREDICTIVE MODELING....................................................................................7
5.1
Grid and Layer Design....................................................................................9
5.2
Model Boundaries........................................................................................10
5.3
Model Input.................................................................................................11
5.4
Steady -State Calibration..............................................................................12
5.5
Sensitivity Analysis......................................................................................13
5.6
Predicted Groundwater Flow Patterns.........................................................13
5.7
MT3DMS Transport Modeling.......................................................................13
5.7.1 M73DMS Settings.................................................................................14
OEM
5.7.2 Nitrogen Loading Scenarios......................................................................15
5.7.3 M73DMS Predictions................................................................................15
5.8
Model Limitations.........................................................................................18
NOW
6.0
CONCLUSIONS AND RECOMMENDATIONS......................................................18
7.0
REPORT CERTIFICATION
.................................................................................19
8.0
LIST OF REFERENCES......................................................................................21
MIND
FIGURE
Figure 1: New Bern Wastewater Reclamation Facility
0000 Figure 2: Model Grid Design
Figure 3: Cross Section through East Lake showing Vertical Distribution of Model Layers
Figure 4: Recharge
FUN Figure 5: Layer 1 Boundaries
Figure 6: Layer 2 Boundaries
Figure 7: Layer 3 Boundaries
FM Figure 8: Layer 5 Boundaries
Figure 9: Modeled Horizontal Hydraulic Conductivity Distributions
Figure 10: Observation Well Locations
RON Figure 11: Model Calibration Plots
Figure 12: Modeled Potentiometric Surface for the Surficial Aquifer
Figure 13: Modeled Potentiometric Surface for the River Bend Aquifer
fRIN Figure 14: Modeled Potentiometric Surface for the Upper Spring Garden Aquifer
Page i
Figure 15: Modeled Potentiometric Surface for the Lower Spring Garden Semi -Confiner
Figure 16: Modeled Potentiometric Surface for the Lower Spring Garden Aquifer
Figure 17: Modeled Groundwater Flow within a Vertical Section through the East Lake
Figure 18: Maximum Predicted Total Nitrogen (TN) Concentrations above Background During
2016
Figure 19: Chronological Total Nitrogen Levels in Monitoring Wells Surrounding the East Lake
Figure 20: Predicted Total Nitrogen (TN) Concentrations above Background after 15 Years of
Daily Discharge of 5 MGD of Effluent containing 5.0 mg/L of TN to the East Lake
APPENDICES
.. Appendix I: 2016 East Lake Usage Records
me
am
am
MR
am
an
am
am
fm
Page ii
Predictive Modeling Analysis, East Lake Wastewater Reclamation Facility
am
EXECUTIVE SUMMARY
'^ GMA has developed a three-dimensional, six -layer groundwater flow and transport model for
the East Lake Wastewater Reclamation Facility operated by the City of New Bern. The
groundwater flow model was constructed using the Groundwater Modeling System interface
(GMS 10.2.3) with the United States Geological Survey (USGS) groundwater model, MODFLOW-
2000. MODFLOW is a modular, three-dimensional groundwater -flow model code that simulates
groundwater flow using a finite -difference method applied to a block -centered rectangular grid.
Model assumptions were based upon a wealth of regional and local hydrogeologic data available
to GMA. Model input assumptions and conceptual model design elements were derived
predominantly from the following sources: GMA's prior work on the City of New Bern Castle
Hayne Aquifer Welifield, data provided by Martin Marietta Aggregates from the Clarks Quarry,
design data provided by Rivers and Associates, Inc. for the East Lake Wastewater Reclamation
facility, drilling and testing data collected by GMA from the Mackilwean Turf Farm, published
data from an ECU master's thesis prepared by Sara Matyiko, results of prior regional modeling
performed by Ken Pohlig of the NCDENR, and 10 years of groundwater and effluent monitoring
data provided by the City of New Bern.
GMA calibrated the groundwater model to simulate steady-state flow based on average
elevations of groundwater and lake levels reported for 2016. Hydraulic conductivities, recharge,
and river conductance values were calibrated using a "trial -and -error" approach to test the
overall soundness of the conceptual model and to ensure that calibrated values represented
real -world ranges. The model calibration has an acceptably high correlation coefficient
(r2=0.92), and the root mean square of residuals was low (RMSR=0.79 ft). The groundwater -
flow model provides spatial representations of head pressure variations and groundwater flow
for the Castle Hayne Aquifer System and the Surficial aquifer for an approximately 44 square
mile (mil) area. After model calibration, GMA used MT3DMS (Modular Transport, 3-
Dimensional, Multi -Species model) in conjunction with MODFLOW to predict nitrogen
concentrations moving from East Lake into and through the groundwater system.
Results from the MODFLOW groundwater -flow model and MT3DMS simulations indicate the
Im following:
• Groundwater pumping associated with the nearby Martin Marietta Clarks Quarry and the
City of New Bern production wells were determined to not have an effect on the
groundwater flow from the East Lake Wastewater Reclamation Facility.
The East Lake is functioning as a recharge boundary that drives groundwater radially
and downward under the induced, artificial head created by the wastewater facility.
• The induced head beneath the East Lake results in a groundwater divide in the upper
layers of the Castle Hayne Aquifer beneath the lake. Groundwater recharge on the east
side of that divide discharges to the Neuse River, and groundwater recharge west of the
divide discharges to the West Lake.
,R ES-1
IF_
Predictive Modeling Analysis, East Lake Wastewater Reclamation Facility
OR
• The majority of groundwater flow out of East Lake is concentrated at the sides of the
basin where hydraulic gradients are the steepest and where aquifer permeability is the
highest. An estimated 62% of the groundwater recharge exiting the East Lake
ultimately discharges to the Neuse River. The remaining 38% discharges to the West
Lake, where it then flows as surface water back to the Neuse River, or it is lost as
evapotranspiration. Travel times for the majority of groundwater flow from the recharge
at East Lake to the areas of discharge to the Neuse River or to the West Lake range
from 66 days to 5 years.
• Model results indicate that approximately 4% of the total groundwater recharge entering
from the East Lake is directed downward through deeper, less permeable, layers of the
Castle Hayne Aquifer System before eventually discharging at the Neuse River or the
West Lake. Travel times for this deeper portion of the groundwater flow from the East
Lake are on the order of 100+ years.
• Historical water -quality monitoring of wells surrounding the East Lake facility do not
indicate an increase in total nitrogen (TN) concentrations above the background
concentrations that were measured prior to the facility's first use. Further, base -level
�.
(i.e., preexisting) concentrations in some monitoring wells (e.g. EL1 and EL3) prior to
use of the facility exceed the measured TN concentrations in the East Lake after the
start of facility use. The field monitoring data indicate, on average, that the addition of
wastewater into the East Lake is not resulting in a recognizable increase in TN
concentrations in the groundwater monitoring wells surrounding the facility.
• During 2016, effluent was discharged to the East Lake reclamation facility during the
months of May and November (with a small amount discharged during October).
Consistent with historical water -quality monitoring, MT3DMS groundwater transport
simulations under this usage scenario predict no measurable increase over base level TN
concentrations at the groundwater monitoring well locations. These MT3DMS
simulations were performed without applying a retardation factor, a very conservative
assumption. The results suggest that, under present usage patterns, TN added to the
East Lake is being rapidly diluted in the groundwater system to concentrations that
cannot be distinguished from the naturally occurring background concentrations.
• Per request of the City of New Bern and Rivers and Associates, Inc., GMA ran additional
MT3DMS simulations that modeled effluent discharge to the East Lake at a rate of 5.0
million gallons per day (MGD) and a concentration of 5.0 mg/L of TN for four equally
FM
spaced months per year. This scenario represents the realistic maximum time period
that the City would discharge to the East Lake. MT3DMS simulations predicted that,
after discharging under this scenario for 15 years, TN loading to the Neuse River would
be 8,963 pounds of TN per year, or 35% of the total input.
• As an additional run, GMA also modeled effluent discharge to the East Lake at a rate of
5.0 MGD and a concentration of 5.0 mg/L of TN every day for 15 years. Under this
usage scenario, after 15 years, the model predicted an annual TN load to the Neuse
River of 29,558 pounds per year, or 39% of the TN input
• Because the MT3DMS runs assume no retardation factor for TN movement through the
groundwater, all reductions in concentration in these simulations are due to dilution
only. Previous research indicates that denitrification represents a sink of approximately
5% of the nitrate -nitrogen load to the Neuse River and its tributaries. Applying this
PM ES-2
Predictive Modeling Analysis, East Lake Wastewater Reclamation Facility
research to the model results, GMA predicts an annual nitrogen load of 8,515 pounds
per year of TN to the Neuse River under the maximum potential usage of the
— reclamation facility (quarterly wastewater inputs). Because this load is only 30% of the
TN added to the system, the load represents a potential TN credit of 70%.
Available field data show close agreement with the MODFLOW 2000 groundwater -flow model
simulations and the MT3DMS results. GMA asserts that the groundwater -flow model that we
have produced is a reasonable representation of the groundwater conditions associated with
the City of New Bern Wastewater Reclamation Facility. However, groundwater models must
incorporate simplifying assumptions about the nature of groundwater flow, and as a result
FM some local heterogeneities of the groundwater system may not be accurately represented. For
example, this model assumes that groundwater flow through the aquifer beneath the lake
occurs solely as homogeneous flow and does not reflect preferential flow paths (i.e.,
heterogeneities) that may exist within the limestone. Although there is some degree of
uncertainty in the model predictions, this model reasonably simulates groundwater flow and
advective transport of total nitrogen added to the East Lake of the New Bern Wastewater
Reclamation facility based on the best available data.
RM
FM
am
FM
MM
OM
MR
,M ES-3
ma
1.0 INTRODUCTION
— The City of New Bern (the City) operates a wastewater reclamation facility at the location of the
former Martin Marietta Glenburnie limestone quarry. The quarry ceased operations in 1996.
The site, now known as the East Lake Reclamation Facility, is located along the south bank of
the Neuse River on the north side of New Bern in Craven County, NC (Figure 1). The site
includes a 145-acre abandoned quarry lake, called the East Lake, which receives treated
wastewater effluent. An eleven foot high berm constructed around the lake allows for treated
wastewater to be pumped to the lake where it then infiltrates into the subsurface. The
wastewater moves away from the quarry lake, through the groundwater system where it is
diluted, and eventually discharges into the Neuse River.
In 2004, the North Carolina Division of Water Quality (NCDWQ) granted the City of New Bern a
non -discharge permit in addition to their existing NPDES permit for discharging into the Neuse
River. The non -discharge permit allowed the city to discharge tertiary treated wastewater into
the quarry disposal system, and/or water could be irrigated onto the nearby Mackilwean turf
farm. The quarry portion of the permit was 'performance limited' rather than'flowrate limited',
and required the City to monitor the wastewater effluent water quality, the East Lake water
quality and water level, and the water quality within a network of compliance monitoring wells
surrounding the facility.
RM
The 2004 non -discharge permit did not provide the City nitrogen credit for wastewater diverted
to the East Lake. In accordance with the City of New Bern's non -discharge permit, to apply for
FM nitrogen credit, the City is to provide estimates of subsurface total nitrogen loading to the
Neuse River that may result from flow out of the quarry lake. To accomplish this, Groundwater
Management Associates (GMA) has developed a groundwater -flow model to simulate the
'm directions, rates, and concentrations of total nitrogen (TN) moving through the groundwater
system from the quarry lake to a point of discharge.
fm
2.0 SCOPE OF WORK
Rivers and Associates, Inc. contracted GMA to evaluate the nitrogen loading of the Neuse River
from the New Bern wastewater reclamation facility. GMA accomplished this work in three
stages:
• Stage 1 Site Characterization and Records Review — GMA reviewed available
records to establish a baseline understanding of the operation of the facility, the
hydrogeologic data, and prior modeling. Specific records reviewed included operational
and monitoring records of effluent disposal to the quarry lake, prior hydrogeologic
studies, including a thesis and a previous mounding modeled prepared by GMA, and
prior nitrogen loading evaluations developed in 2000 by Ken Pohlig of the NC
FM Page 1
nm
a.o
Department of Environment and Natural Resources (now the NC Department of
Environmental Quality).
Stage 2 Groundwater flow modeling — GMA developed a 3-dimensional finite -
difference groundwater flow model (USGS MODFLOW-2000) using Groundwater
Modeling System (GMS) software. Model properties were derived from prior
hydrogeologic studies and information gathered during the records review process. The
groundwater model was calibrated to simulate steady-state flow based on average
elevations of groundwater and lake levels reported for 2016.
• Stage 3 Groundwater transport modeling — Once the model was calibrated to local
water -level data from the existing monitoring well network and the quarry lake levels,
GMA used the MT3DMS module of MODFLOW to predict how quarry use would affect
nitrogen concentrations in the adjacent groundwater system.
FM
This report provides modeling predictions of groundwater flow paths and TN loading estimates
to the Neuse River resulting from the addition of tertiary treated wastewater additions to the
fm East Quarry Lake reclamation facility.
'R 3.0 REGIONAL HYDROGEOLOGIC SETRNG
IM The facility (Figure 1) lies within the Coastal Plain Physiographic Province of North Carolina
(NCGS, 1985). North Carolina's Coastal Plain is a broad, relatively flat physiographic province
separating the hilly Piedmont province from the Atlantic Ocean. Local elevation ranges from
FM about 35 feet above sea level (ASL) in upland areas west of the facility to almost sea level at
the shore of the Neuse River. Land surface topography in Craven County is primarily a product
of Neogene and Quaternary fluctuations in sea level that repeatedly inundated and exposed the
land over the past 23 million years (Horton and Zullo, 1991). These sea -level cycles sculpted
the land surface into broad, relatively flat marine terraces bounded by low escarpments that
represent former shorelines. Streams and rivers have incised these terraces to create the
current topographic character of the area.
rRa The regional hydrogeology of the area has been described in published reports by GMA (2014),
Fine (2008), Lautier (2001), Giese and others (1997), Winner and Coble (1996), and Lyke and
Winner (1990). For the purposes of this investigation, relevant aquifer and confining unit
fm descriptions presented in GMA (2014) have been adopted, and these are presented below.
MW 3.1 Regional Geology
The Coastal Plain Province is underlain by marine, estuarine, and terrestrial sediments that were
deposited along the continental margin over the past 200 million years. The East Lake
Reclamation Facility is underlain by approximately 1,700 feet of Cretaceous to Recent aged
sediments and sedimentary rocks that were deposited on top of pre -Mesozoic aged (>250
MW
,e Page 2
M
million years) basement rocks. Basement rocks in the vicinity of the facility site are mapped as
granite of the Carteret Batholith (Lawrence and Hoffman, 1993).
The Mesozoic -aged sediments beneath the property are dominantly clastic in nature and include
sequences of silt and clay interbedded with sand and gravel zones and minor amounts of shell.
These sediments are associated with deltaic and marginal marine depositional environments
that predominated along the eastern margin of North America from about 145 to 66 million
years ago. The Mesozoic sediments have been hydrostratigraphically subdivided into four
principal aquifers of the Cretaceous Aquifer System (CAS). The CAS includes (from deep to
shallow) the Lower Cape Fear Aquifer, the Upper Cape Fear Aquifer, the Black Creek Aquifer,
and the Peedee Aquifer.
FAR Overlying the CAS is a sequence of Cenozoic -aged (<66 million years) sediments of dominantly
marine origin. These include significant beds of sands, shelly clays, and fossiliferous sandy
limestones. These sediments have been stratigraphically subdivided into six mappable units,
including (from deep to shallow): the Paleocene Beaufort Formation, two members of the
Eocene Castle Hayne Formation, the Oligocene River Bend Formation, the Pliocene Duplin
Formation, and unnamed Pleistocene to recent marine terrace deposits (GMA, 2014).
The basal unit of the Castle Hayne Formation in Craven County is the Comfort Member. The
Comfort Member is a hard sandy skeletal limestone and is the most widespread of the Castle
Hayne units. The Spring Garden Member of the Eocene Castle Hayne Formation overlies the
Comfort within the study area. The Spring Garden Member is a thinly bedded, very sandy,
FAR shelly limestone to calcareous unconsolidated sand deposit that is interpreted to have been
deposited in a low -energy, shallow marine environment. The upper surface of the Spring
Garden unit is very uneven and is often thinly coated by a veneer of phosphate (Ward et al.
`m 1978).
f M9 The Oligocene River Bend Formation overlies the Castle Hayne Formation at the New Bern
Quarry. The unit dips to the east and south in the New Bern area and is about 10 feet thick at
the location of the New Bern Quarry (Ward et al. 1978). The River Bend Formation is
composed of limey sand, sandy shell, barnacle hash, and sandy limestone, with the most
prominent lithology being a pelecypod moldic sandy limestone. It is interpreted to be an inner
to mid -shelf deposit that formed below wave base. Near the base of the formation, the River
Bend typically contains little to no quartz sand and predominantly consists of barnacle plates
and molluscan molds in calcarenite matrix (Ward et al. 1978).
pan
Together, the Comfort and Spring Garden Members of the Castle Hayne Formation, and the
River Bend Formation comprise the Castle Hayne Aquifer System (CHAS) in Craven County.
Sm Remnants of the Pliocene -age Yorktown Formation occur as laterally discontinuous deposits of
fossiliferous sands and clays that often occur as paleo-channel fill material. Shallower deposits
RM are primarily Pleistocene in age and are composed of sand, silt, shelly debris, and some clay.
,e Page 3
rm
M"
MR
MR
am
MIN
FMI
FRO
MR
Table 1 presents a list of the principal aquifers that occur in Craven County and the associated
ages and formations that comprise these aquifers.
Table 1. Regional Aquifer Framework for Craven County (Adapted from GMA, 2014)
Hydrostratigraphic
Period
Ma
Geologic Age
Formation Name
Units
cRecent
Unnamed Holocene
Surfiicial Aquifer
v
o�
0 - 2.5
to
Pleistocene
James City Fm.
Upper Castle Hayne CL
a�
2.5 - 5
Pliocene
Yorktown/Duplin Fm.
Yorktown Aquifer
a
Upper Castle Hayne CL
z
23 - 34
Oligocene
River Bend Fm.
Upper Castle Hayne
Aquifer
Upper Castle Hayne
Castle Hayne Fm.
Aquifer
(Spring Garden Mbr.)
Middle Castle Hayne CL
34 - 56
Eocene
Middle Castle Hayne
0
A uifer
q
a�
a
Lower Castle Hayne CL
(Comfort Member)
Lower Castle Hayne
Aquifer
Beaufort Fm.
Beaufort CL
56 - 66
Paleocene
Beaufort Aquifer
Yaupon Beach Fm.
Peedee CL
66 - 72
Upper Cretaceous
Peedee Fm.
Peedee CL
(Maastrichtian)
Peedee Aquifer
72 — 84
Upper Cretaceous
Black Creek Group
Black Creek CL
(Campanian)
Black Creek Aquifer
W
84 — 86
Upper Cretaceous
Middendorf Fm.
Black Creek Aquifer
401
a�
.
(Santonian)
(Pleasant Creek Fm.)
Upper Cape Fear CL
86 — 90
Upper Cretaceous
Upper Cape Fear CL
(Coniacian to
Cape Fear Fm.
Upper Cape Fear Aquifer
90 — 94
possibly Turonian)
Lower Cape Fear CL
Lower Cape Fear Aquifer
>252
Paleozoic and
Crystalline Basement
Bedrock Aquifer
Late Proterozoic
Ma — Million Years (mega-annum) CL — Confining Layer Fm. — Formation Mbr. — Member
,�, Page 4
no
3.2 Hydrogeology of the Site
The City of New Bern Reclamation Facility is located within the Neuse River drainage basin on
no the north side of the City of New Bern in Craven County, NC. The Reclamation Facility lies
immediately adjacent to the Neuse River, with the East Lake being located as close as 120 feet
F, from the southern river bank (Figure 1). The Neuse River flows from the northwest to the
southeast. Major bodies of water that discharge to the Neuse River within the model area
include the Trent River to the south and Bachelor Creek and one of its tributaries, Caswell
no Branch, to the north.
As presented in Table 1, the CHAS in Craven County is composed of three aquifer units: the
Lower Castle Hayne (Comfort), the Middle Castle Hayne (Spring Garden), and the Upper Castle
Hayne (River Bend). The middle and upper portions of the CHAS are the primary aquifers of
interest for this investigation. All of the units evaluated in this report generally thicken from
west to east. The Lower Castle Hayne confining unit is used as the lower limit of our
investigation. This confining unit, consisting predominantly of silt and clay, lies beneath the
Lower Spring Garden member and impedes the vertical flow of groundwater. Aquifers and
confining units deeper (older) than the Lower Castle Hayne confining unit were not included in
this modeling investigation because these deeper units are unlikely to be affected by the
addition of wastewater to the East Lake.
Me During pre -mining conditions, the groundwater flow direction at the facility location was to the
east-northeast towards the Neuse River. When the quarry was active, mine dewatering
lowered head beneath the quarry and facilitated movement of salty groundwater from the
Neuse River westward towards the quarry. Under present usage conditions, the addition of
effluent to the East Lake produces a local high water level that causes the downward movement
of groundwater and thus serves as a groundwater recharge area. At the subject site, the
Surficial aquifer and the Upper Castle Hayne Aquifer are unconfined, and the Middle Castle
Hayne Aquifer is semi -confined.
Major withdrawals in the study area are the Martin Marietta Clarks Quarry, located
approximately three miles west of the East Lake Reclamation Facility, and the City of New Bern
Castle Hayne Aquifer wellfield (Figure 1). The Clarks Quarry is an active limestone mine that is
permitted to withdraw 16 million gallons per day (MGD) from the Surficial and Castle Hayne
FOM aquifers as part of the mine's dewatering activities. Since 2002, average daily withdrawals from
the mine have ranged from 6.7 to 11.9 MGD. New Bern's Castle Hayne wellfield consists of
fifteen production wells scattered west of the site. The wellfield is permitted to withdrawal a
maximum of 5.5 MGD from the Castle Hayne aquifer. The New Bern wells are screened in the
lower portions of the Spring Garden member, and withdrawals are from the Middle Castle
Hayne aquifer unit. Maximum withdrawals from individual wells within the New Bern Castle
Hayne wellfield may range from 0.252 MDG to 0.720 MGD.
em
�, Page 5
4.0 REVIEW OF PREVIOUS STUDIES
4.1 The Pohlig Model
GMA reviewed a previous groundwater flow (MODFLOW) and transport model (MT3D) of the
facility developed by Ken Pohlig (Pohlig, 2000) of the NCDWQ. Pohlig constructed a five layer
regional model. The model simulated a wastewater flow to the quarry of 6.5 MGD with a total
nitrogen concentration of 4 mg/L. Nitrogen removal or uptake processes were not considered.
Pohlig's model predicted that the addition of treated wastewater to the reclamation facility
would cause a small flux of groundwater to migrate towards the Clarks Quarry with the rest
discharging to a 2-mile stretch of the Neuse River via the surficial and Castle Hayne aquifers.
Importantly, however, Pohlig modeled the entire reclamation facility property, a 550-acre
footprint, as the quarry lake receiving wastewater input. This may have originally been the plan
for the facility; however, an 11-foot high berm was constructed only around the 145-acre East
Lake, and only the East Lake receives wastewater. Thus, the Pohlig model greatly
overestimates the area of the Neuse River receiving wastewater discharge.
Pohlig also chose to model the nitrogen inputs from the reclamation facility operation to the
groundwater system as a constant concentration of total nitrogen in recharge assigned to
model cells within the quarry property footprint. This approach does not adequately simulate
the dilution of the wastewater effluent as it discharges to the East Lake. Surface water in the
East Lake is regularly diluted by precipitation, and it is this diluted concentration of total
nitrogen that actually enters the groundwater system beneath and around the East Lake. For
example, in 2016, the City discharged wastewater to the lake with total nitrogen concentrations
as high as 8.27 mg/L., but measured total nitrogen levels in the East Lake during 2016 were less
than 1.15 mg/L at all monitoring locations. Thus, GMA believes that Pohlig's estimates of
nitrogen loading to the Neuse River are erroneously high.
4.2 Matyiko Master's Thesis
two GMA also reviewed published data from an East Carolina University Master's thesis prepared by
Sara Matyiko, which described the hydrogeology of the CHAS at the former Glenburnie Quarry.
elm Matyiko also attempted to predict the hydraulic impacts of discharging tertiary treated
wastewater to the East Lake Reclamation Facility (Matyiko, 2001). Through drilling
observations and analysis of geophysical logs, Matyiko divided the site into surficial sediments
(0 - 8 feet below the land surface), an 'upper zone' consisting of alternating hard and soft gray
moldic limestone beds (from 8 — 66 feet below the land surface), a softer `middle zone'
composed of calcareous sandstone (from 66 — 83 feet depth), and a 'lower zone' consisting of
alternating sand, clay, and limestone beds (from 83 —162+ feet depth). Matyiko estimated
hydraulic conductivity, transmissivity, and storativity at the site using data collected during 24-
hour constant -rate aquifer pumping tests at wells around the facility. Testing indicated that the
transmissivity of the "upper zone', which corresponds to the Riverbend Formation, was as much
as two orders of magnitude greater than the lower units. Testing also confirmed that the CHAS
,R Page 6
MR
rs
was, in fact, hydraulically connected with both the Glenburnie Quarry lakes and the Neuse
River. Contrary to Ken Pohlig's model, Matyiko also documented a groundwater divide between
the Clarks Quarry and the reclamation facility.
4.3 Additional Data Sources
In addition to the data collected from the Pohlig model and the Matyiko thesis, GMA used
aquifer test data and drilling records from our previous work in the area to serve as the basis of
constructing the model. Geophysical logs and aquifer testing data from the New Bern Castle
Hayne Aquifer wellfield served as a major source for top of aquifer elevations and hydraulic
conductivity values within the study area. Drilling and testing data collected by GMA from the
Mackilwean Turf Farm, just north of the reclamation facility, provided the major source for
Surficial aquifer parameters (GMA, 2006).
5.0 PREDICTIVE MODELING
MR GMA developed a three-dimensional groundwater MODFLOW model using Groundwater
Modeling System (GMS) software to simulate groundwater flow in the vicinity of the East Lake.
MR GMS is a graphical user interface that facilitates model design and parameter input for
programming MODFLOW, a modular, 3-dimensional, finite -difference, groundwater -flow model
code developed by the United States Geological Survey (USGS) (Harbaugh and others, 2000).
O, GMA's model predicts hydraulic head in the Surficial, Riverbend, and Spring Garden aquifer
units under steady state conditions based on regional water -level information from 2016.
The purpose of this report section is to describe the development, calibration, and application of
the model to support non -discharge permit renewal for the City of New Bern. Specifically, GMA
performed model simulations to:
• Calibrate to baseline groundwater -flow conditions in the vicinity of the City of New Bern
fim Wastewater Reclamation Facility,
• Simulate the impacts on groundwater flow of effluent discharge into East Lake,
• Simulate the transport of total nitrogen emanating from East Lake and moving through
the groundwater system in response to effluent discharges to East Lake, and
• Estimate nitrogen loading to the Neuse River from treated wastewater inputs to East
Lake.
Groundwater flow for the period of January 1, 2016 to December 31, 2016 was simulated for
four aquifer units potentially affected by the addition of wastewater to the reclamation facility
quarry — the Surficial, the Riverbend, and the Upper and Lower Spring Garden units. Because
headwater streams and shallow surface water features far from the reclamation facility are
unlikely to affect ground -water flow patterns in the vicinity of East Lake, these features were
not explicitly modeled. Therefore, simulations for the Surficial aquifer relate only to the areas in
,M Page 7
n
r�
the refined portion of the model in the vicinity of the reclamation facility. Table 2 lists a
summary of the assumptions and design elements used to develop the model.
ow
ROM
r"
FM
em
rmm
MR
M
rM
on
M
i aDie z. bummarlf
or the moaeii uesin ana Assumptions
Design and Assumptions
Parameters
Area
Area surrounding the proposed lake including the Neuse River to the northeast
and the Clarks Quarry to the west (see Figure 1)
Code & Solver
MODFLOW 2000 — PCG Solver with GMS 10.2.3
Calibration Period
Steady state model calibrated to average 2016 head values
Dimensions &
Model origin: x= 2,558,000, y = 488,000 (NAD83 State Plane NC - Feet)
Orientation
X extent: 40,000 ft Y extent: 30,500 ft
Grid rotated 320 west consistent with predominant flow direction.
Grid Spacing
101 rows & 120 columns (Figure 2)
Cell size: ranging from approximately 167 ft x 167 ft to 500 ft x 500 ft
Layers
6 layers: SurFcial aquifer (SA), River Bend (RB), Upper Spring Garden (USG),
Lower Spring Garden Semi -Confining Unit (LSGCU), Lower Spring Garden
(LSG), Lower Castle Hayne Confining Unit (LCHCU) (Figure 3)
Surfaces
Based on City of New Bern well logs, DWR well logs, and site borehole data
BOUNDARIES
No -flow
Unless otherwise specified, model extents were left as no flow boundaries by
Boundaries
default.
Inactive Cells
Cells within the Clarks Quarry excavation area that were excavated after the
date of the NED source data were set to inactive in order to simulate
excavation. Inactive cells function as no -flow boundaries.
Groundwater
Groundwater recharge set at 0.002 ft/day (8.76 in/yr) for majority of model.
Recharge
Range of recharge used in the model: 0.001 ft/day to 0.005 ft/day (4.38 in/yr
to 22 in/yr). All recharge applied to the top of layer 1. Recharge details shown
in Figure 4.
Constant Head
No constant head cells were used in this model.
General Head
General head boundaries (GHBs) were used in the SA (Layer 1) to simulate the
lakes in and around the reclamation facility and to simulate lateral flow to the
Trent River, south of the model. GHBs were set in the Riverbend Formation
(Layer 2) based on regional groundwater contouring available from the NCDWR
website for 2015 (NCDWR, 2017).
Drains
Drain cells were used to model the Clarks Quarry in the western portion of the
model area. Drain cell elevations (set in layer 3) were assigned based on water
level data from the Clarks Quarry monitoring wells.
Rivers
The Neuse River and perennial streams to the east of the reclamation facility
were set as MODFLOW River boundaries with elevations trending linearly
downstream between two topographic elevations.
Pumping Wells
Pumping well data were gathered from the NCDWR website (NCDWR, 2017b).
,M Page 8
5.1 Grid and Layer Design
_ The study area is the East Lake Wastewater Reclamation Facility located at the former New
Bern Martin Marietta Quarry in Craven County. To ensure realistic hydrologic boundaries for the
groundwater model, and to reduce the potential for introducing non-realistic boundary effects
during modeling, the model was expanded to an area of approximately 44 mi2 (Figure 1). The
large areal extent of the model also ensured that the simulation included the closest large-scale
groundwater users (e.g., the Martin Marietta Clarks Quarry and the City of New Bern wellfield)
so that we could ascertain any predicted interactions between these large groundwater users
and groundwater recharge from the East Lake.
The finite -difference method requires that the model be discretized horizontally into a two-
dimensional grid (Figure 2) and vertically into layers (Figure 3), resulting in a three-dimensional
array of cells known as the model grid. GMA developed a 6-layer static groundwater model that
represents the units underlying the wastewater treatment facility.
Table 3: Model Layers
Model Layer
Lithology
Hydrostratigraphic Unit
Layer 1
Surficial Sands and Clays
Surficial Aquifer
Layer 2
Riverbend Fm. (moldic limestone)
Upper Castle Hayne Aquifer
Layer 3
Upper Spring Garden (skeletal limestone)
Upper Castle Hayne Aquifer
Layer 4
Lower Spring Garden (sand)
Middle Castle Hayne Semi -
Confining Unit
Layer 5
Lower Spring Garden (limestone)
Middle Castle Hayne Aquifer
Layer 6
Lower Castle Hayne (clay)
Lower Castle Hayne Confining Unit
Elevations for the top of each model layer were constructed by contouring interpreted data
using GMS. Hydrogeologic data used to construct layer surfaces in the model were primarily
obtained from drilling data collected by GMA within the model area, from Clarks Quarry well
_ records, and from the DWR online ground water database
(https://www.ncwater.org/?pa(ie=20). The elevation of the land surface was based on the
USGS 1/3 arc -second National Elevation Dataset available online
(https://nationalmai).00v/elevation.htmi). Data were correlated to construct a three-
dimensional representation of aquifers and confining units.
The six hydrostratigraphic layers modeled are listed in Table 3. The model grid has 101 rows
and 120 columns (Figure 2). The spatial discretization of the model grid determines the
_ resolution. GMA has designed a model grid that is tighter in the area surrounding the East Lake
to increase the resolution in this area. This increased resolution more accurately simulates
groundwater flow from the lake to the nearby Neuse River, located between 120 and 1200feet
to the northeast. Cell dimensions range from approximately 167 feet by 167 feet
,� Page 9
M"
em
(approximately 0.64 acres) in the refined model area to approximately 500 feet by 500 feet
(approximately 5.7 acres) outside of the refined area. Principal grid axes (rotated 32' west)
"" align with the predominant groundwater flow direction and are generally parallel the Neuse
River.
5.2 Model Boundaries
Boundary conditions are necessary to define how the model interacts both internally and with
areas outside the model domain. Incorrectly assigned or unrealistic boundary conditions are
often the greatest source of error in groundwater modeling. To minimize any potential errors in
boundary specification, GMA utilized boundaries corresponding to natural physical boundaries,
wherever possible. A summary of the boundaries specified in this model is provided below and
boundaries are shown in Figures 5-8. Unless otherwise specified below, model extents were
left as no flow boundaries by default. However, the extent of the model was chosen to be large
enough so that model edge boundaries would only minimally affect the solutions for the area of
interest. Assumptions used to construct the model are summarized in Table 2.
General Head Boundaries
GMA used general head boundaries (GHBs) to simulate the lakes within the model area and to
simulate lateral boundary flows to and from distant boundaries located outside of the model
domain. GHBs are assigned by defining a hydraulic head and a conductance value. A GHB will
remove or add water to adjacent cells with lower hydraulic head, based on an assigned
hydraulic head, and a threshold conductance, expressed in square feet per day. The flow to
adjacent cells is not allowed to exceed the conductance of the general head boundary.
Because the primary purpose of this model was to predict nitrogen loading to the Neuse River
r, and because East Lake water levels can be artificially maintained, GMA modeled the East Lake
as a fixed lake level model using a GHB (Figure 5). As a surface water feature, East Lake is not
represented explicitly within the model (i.e. the lake itself is not represented by grid cells);
'm rather a general head representing the artificially maintained head within the lake was assigned
to the thin section of layer 1 corresponding to the lake bottom. In the initial calibrated steady-
state model, the elevation of the East Lake water surface was set at 5.71 feet ASL, which was
the average lake level in 2016.
,e" GHBs were also used in the Surficial aquifer layer to simulate the surrounding lakes (West Lake,
North Lake, and the private lake located north of North Lake) and to simulate lateral flow to the
Trent River, which is located outside of the model domain to the south. GMA used regional
groundwater contouring available from the NCDWR Ground Water Management Branch Map
Interface (NCDWR, 2017) for 2015 to assign GHBs to the UCHA (Figure 6).
fm Page 10
MR
Rivers and Drains
fm The Neuse River was simulated as a MODFLOW river boundary with elevations trending linearly
downstream (Figure 5). During model calibration, GMA determined that accurate simulation of
water levels in the monitoring wells at the reclamation facility required incorporating the
no channels and creeks that lie between the quarry lakes and the Neuse River. These features
were also modeled as river boundaries.
` l Drain cells were used to simulate the Clarks Quarry in the western corner of the model extent.
Both river and drain cells represent head dependent flux boundaries; but of the two, only river
cells are capable of recharging the groundwater system. In MODFLOW, if the head in a drain
cell falls below the specified elevation, the flux from the drain model cell drops to zero. In
T other words, drains can only take water out of the model.
MR
5.3 Model Input
Model input included recharge estimates (Eimers et al., 1994 and Giese et al., 1997), hydraulic
conductivity values for the aquifers and confining units modeled (Matyiko, 2001; GMA 2008-
2013), withdrawals by pumping, and average observed groundwater levels for 2016.
M"
RecharcL
The discharge from the surficial aquifer to the small tributaries outside of the immediate
reclamation facility vicinity was not simulated as part of this model. Instead, recharge in this
area was reduced to represent only the fraction estimated to infiltrate to the deeper confined
MR aquifers, a practice employed in previous models within the region (e.g. Eimers et al., 1994 and
Geise et al., 1997). The distribution of recharge throughout the modeled area is shown in
Figure 4. Recharge was reduced in the dewatered area at the Clarks Quarry. GMA also
determined during model calibration that increased recharge rates were needed in the
topographical high area just west of the reclamation facility. All modeled recharge rates were
within the range of published estimates of groundwater recharge to the unconfined Surficial
aquifer in this region (e.g. 5-20 in/year (0.001-0.005 ft/day), Heath 1980).
ran Hydraulic Conductivity
Hydraulic conductivity (K) is a measure of an aquifer's ability to transmit water. More
,M specifically, hydraulic conductivity is a measure of the volume of water transmitted in a unit of
time through a unit area of the aquifer measured at right angles to the direction of flow under a
hydraulic gradient of one. Hydraulic conductivity is equal to the transmissivity of an aquifer
divided by the aquifer thickness (Heath, 1983). GMA assigned uniform values of hydraulic
conductivity to the Upper Spring Garden (layer 3), the Lower Spring Garden Semi -Confiner
(layer 4), and the Lower Castle Hayne confining layer (layer 6) (Table 4). The distribution of
hydraulic conductivities applied to the surficial aquifer (layer 1), the Riverbend portion of the
Upper Castle Hayne Aquifer (layer 2), and the Middle Castle Hayne Aquifer (layer 5) is shown in
Figure 9. Modeled K values were based on measured site specific data from aquifer testing
029 Page 11
Om
where available (e.g. Matyiko 2001, New Bern Castle Hayne wellfield data). Estimates of
transmissivity for the MCHA were known from multiple aquifer tests conducted at the City of
'W New Bern Castle Hayne wellfield. Using these data, GMA assigned a distribution of horizontal K
(Kh) values for the MCHA aquifer (Figure 9).
MR
Values of vertical hydraulic conductivity (K„) are generally small and are typically at least an
order of magnitude less than the Kh. GMA modeled vertical hydraulic conductivity (K„) as one
order of magnitude less than Kh for all layers except the Upper Spring Garden (layer 3), which
was slightly lowered during the calibration process (Table 4).
on Table 4: Hydraulic Conductivity and Porosity of the Model Layers
em
fm
Model
Aquifer/Confining Unit
Kh (ft/day)
Kv (ft/day)
Porosity
Layer
Layer 1
Surficial Aquifer SA
See Figure 9
0.25
Layer 2
River Bend RB
See Fi ure 9
0.15
Layer 3
Upper Spring Garden USG
3.0
0.08
0.15
Layer 4
Lower Spring Garden Semi -
1.0
0.1
0.15
Confining Layer LSGSU
Layer 5
Lower Spring Garden (LSG)
See Fi ure 9
0.3
Layer 6
Lower Castle Hayne
0.001
0.0001
0.2
ConfiningLayer LCHCU
MR Groundwater Withdrawals
The removal of groundwater via pumping results in a lowering of hydraulic head near the
pumping wells and must be accounted for during modeling. Pumping wells withdrawing within
the study area were identified using NCDWR Central Coastal Plain Capacity Use Area permitting
Mel records and GMA's regional knowledge (Figures 6-8). The majority of pumping wells located
within the model area belong to the City of New Bern's Castle Hayne Aquifer wellfield.
Estimated daily pumping rates were based on average pumping data for 2016, which were
available from the NCDWR website (NCDWR, 2017b).
5.4 Steady -State Calibration
Steady-state flow was simulated to represent conditions in 2016. GMA assigned average head
observations from 2016 to 16 monitoring and inactive production wells within the study area for
Pa use during model calibration (Figure 10). Head observations were included for wells screened
within the Surficial aquifer (n=1), the Riverbend Formation (UCHA; n=8), the Upper Spring
Garden (UCHA; n=4), and the Middle Castle Hayne aquifers (n=3). Although multiple
monitoring wells exist at the Clarks Quarry site, GMA eliminated all but one of the Clarks Quarry
wells from the calibration well dataset. GMA did not use these wells due to the coarseness of
the grid in this region of the model relative to the complexities of the excavation geometry and
pumping schedules. Calibration of the model to wells between the Clarks Quarry and the East
Lake Reclamation facility (e.g. well PW6 and the Craven 30 Supply well) ensure that the cone of
MIR depression from the mine is well simulated.
,M Page 12
Model input, specifically, hydraulic conductivities, confining -unit vertical conductance, and
recharge rates, were adjusted during model calibration so that simulated heads best matched
measured head data for 2016. To calibrate the model, parameter estimates were manually
adjusted and optimized to increase the goodness -of -fit between observed and calculated heads.
All parameter adjustments were kept within realistic limits according to available site specific
and published information. A comparison between modeled and observed groundwater levels
indicate a very good fit (r = 0.921, Figure 11). The differences between the observed and
calculated heads ranged from 0.03 feet to 1.54 feet. The mean difference was 0.64 feet and
the root mean square residual was 0.79 feet. Calibrated head data were used to prepare
potentiometric surface maps for the layers 1 through 5 to simulate groundwater head elevations
and flow directions within these aquifers (Figures 12-16).
5.5 Sensitivity Analysis
As part of the steady-state model calibration process, GMA performed a preliminary sensitivity
analysis to evaluate the relation between parameter variability and the modeled hydraulic head
in the model. Model response was tested for sensitivity to river conductance values, recharge,
and hydraulic conductivity values using the PEST (Parameter ESTimation) utility. PEST analysis
indicated that the model was sensitive to horizontal hydraulic conductivity values in the
Riverbend and Upper Spring Garden layers and that the model was also sensitive to recharge
values.
5.6 Predicted Groundwater Flow Patterns
The simulated groundwater elevations for the Surficial, Riverbend, and Upper Spring Garden
indicate that the East Lake is functioning as a recharge boundary that drives groundwater
radially and downward under the induced, artificial head created by the wastewater facility
(Figures 12-17). The recharge effects from the East Lake diminish with depth, but a subtle
mound is evident in the model extending down into Layer 6 (the Lower Spring Garden unit).
The induced head beneath the East Lake results in a groundwater divide in the upper layers of
the Castle Hayne Aquifer beneath the lake (Figure 17). Groundwater recharge on the east side
of that divide discharges to the Neuse River, and groundwater recharge west of the divide
discharges to the West Lake. Flow budget analysis indicates that approximately 62% of the
groundwater recharge exiting the East Lake ultimately discharges to the Neuse River. The
remaining 38% discharges to the West Lake, where it then flows as surface water back to the
Neuse River, or it is lost as evapotranspiration.
As shown on the potentiometric surface map for the surficial aquifer (Figure 12), the majority of
groundwater flow out of East Lake is concentrated at the sides of the basin where hydraulic
VIM gradients are the steepest and where aquifer permeability is the highest. Travel times for the
majority of groundwater flow from the recharge at East Lake to the areas of discharge to the
Neuse River or to the West Lake range from 66 days to 5 years. Model results indicate that
approximately 4% of the total groundwater recharge entering from the East Lake is directed
,� Page 13
downward through deeper, less permeable, layers of the Castle Hayne Aquifer System before
eventually discharging at the Neuse River or the West Lake. Travel times for this deeper
portion of the groundwater flow from the East Lake are on the order of 100+ years.
Model results indicate that groundwater pumping associated with dewatering at the Martin
Marietta Clarks Quarry does not have an effect on the groundwater flow from the East Lake
Reclamation Facility. The cone of depression predicted by this model does not extend to the
facility property. The model also predicts that pumping from the City of New Bern's production
wells does not affect groundwater flow from the East Lake. New Bern's Castle Hayne
production wells are screened in the productive limestone section of the Lower Spring Garden
unit. Impacts from pumping extend upward into the Lower Spring Garden Semi -confining unit,
but do not greatly perpetuate beyond that (Figures 13-16).
5.7 MT3DMS Transport Modeling
MT3DMS (Modular Transport, 3-Dimensional, Multi -Species model) is a modular three-
dimensional transport model for the simulation of advection, dispersion/diffusion, and chemical
reactions of dissolved constituents in groundwater flow systems (Zheng and Wang, 1999).
MT3DMS can be used in conjunction with MODFLOW to predict concentrations of a solute
plume moving through groundwater. GMA used MT3DMS in conjunction with the calibrated
MODFLOW model developed for the site to predict TN concentrations moving from East Lake
into and through the groundwater system.
In this study, the flow simulation was steady state (i.e. with one stress period and one time
step), but the transport simulation was transient (i.e. with multiple stress periods and time
steps). Because the flow model was steady state, the velocity and groundwater sink/source
information from the MODFLOW model is only updated once at the beginning of the simulation
' -9 and remains the same, while changes in sink/source concentrations are determined by the
stress periods (time intervals in which external sinks/sources are constant) set up with the
transport model. GMA modeled nitrogen additions to the East Lake using a mass loading
approach. Nitrogen mass was added to the general head boundary representing the surface of
the East Lake. Initial nitrate concentrations were assumed to be zero at the beginning of all
simulations. GMA also assumed that all of the nitrogen added to the lake enters the
groundwater system and that degradation of nitrogen did not occur — an extremely
conservative approach.
5. Z 1 M73DMS Seft'nQs
Dispersion is the spreading of a solute over a greater area than would be predicted solely from
advective transport. Dispersion is caused by differences in the velocity that water travels at the
pore level as well as differences in the rate at which water travels through different strata in the
flow path (Fetter, 1980). Dispersion is important because it allows the solute to be diluted by
unaffected groundwater and affects the first arrival time of a plume. Longitudinal dispersivities
for the model were estimated using published field dispersivities for similar aquifer types
Page 14
rim
PER
(Gelhar et al. 1992). Horizontal dispersivities were assumed to be one order of magnitude
smaller than longitudinal dispersivities, and vertical dispersivities were assumed to be an order
of magnitude smaller than that (Gelhar et al. 1992).
Advection refers to the transport of miscible solutes at the same velocity as the average
groundwater flow. GMA used the Third Order Total -Variation -diminishing ('TVD) scheme
(ULTIMATF) solver for the Advection package in MT3DMS. This TVD method is a mass
conservative, Eulerian solution technique that is generally more accurate at solving advection-
dominated problems (Zheng and Wang, 1999).
f' 5.72 Nitrogen Loading Scenarios
To evaluate the potential nitrogen loading to the Neuse River, GMA modeled nitrogen transport
FM resulting from wastewater discharge to the East Lake under three scenarios:
1) Actual wastewater discharge to the East Lake during 2016,
2) Quarterly wastewater discharge to the East Lake, which represents the practical
maximum usage of the reclamation facility by the City, and
3) Continuous discharge of wastewater to the East Lake every day for 15 years
During 2016, the City discharged wastewater to the East Lake during the months of May and
November (Appendix I). The 2016 usage pattern required that GMA simulate five stress periods
in the transport model to represent the two months of nitrogen inputs to East Lake. Per
request of the City of New Bern and Rivers and Associates, Inc., GMA ran a subsequent
MT3DMS simulation that modeled effluent discharge to the East Lake at a rate of 5.0 million
gallons per day (MGD) and at a concentration of 5.0 mg/L of TN for four equally -spaced months
per year. This quarterly usage scenario represents the realistic maximum time period that the
City would discharge to the East Lake. As an additional run, GMA also modeled effluent
discharge to the East Lake at a rate of 5.0 MGD and a concentration of 5.0 mg/L of TN every
day for 15 years.
5.7.3 M73DMS Predictions
MT3DMS groundwater transport simulations under the 2016 usage scenario predict no
measurable increase over base level TN concentrations at the groundwater monitoring well
locations (Figure 18). The results suggest that, under present usage patterns, TN added to the
East Lake is being rapidly diluted in the groundwater system to concentrations that cannot be
distinguished from the naturally occurring background concentrations.
rM
These findings are consistent with historical water -quality monitoring of wells surrounding the
East Lake facility. Historical monitoring well sampling does not indicate an increase in total
r' nitrogen (TN) concentrations above the background concentrations that were measured prior to
the facility's first use (Figure 19). Further, base -level (i.e., preexisting) concentrations in some
FM monitoring wells (e.g. EL1 and EL3) prior to use of the facility exceed the measured TN
,, Page 15
F"
MM
concentrations in the East Lake after the start of facility use. The field monitoring data indicate,
on average, that the addition of wastewater into the East Lake is not resulting in a recognizable
increase in TN concentrations in the groundwater monitoring wells surrounding the facility.
Even under the extreme assumption of daily wastewater discharge to the East Lake,
groundwater concentrations of TN at the point of discharge to the Neuse River are only
predicted to be 0.001 mg/L (Figure 20).
To predict mass loading of TN to the Neuse River, GMA evaluated the MT3DMS mass budget for
each model simulation. MT3DMS simulations predict that after 15 years under the quarterly
usage scenario, approximately 8,963 pounds of TN will be discharged to the Neuse River each
year (Table 5). This represents only 35% of the total annual nitrogen discharged to the East
Lake under a quarterly usage scenario. Thus, the city should be eligible for credit for at least
65% of the nitrogen added each year.
Under the unrealistic assumption that wastewater would be discharged to the reclamation
facility continuously for 15 years, the model predicts an annual TN load to the Neuse River of
29,558 pounds, which represents 39% of the TN input under this scenario.
Please note: we recognized a previous incorrect error in the technical memo (GMA,
2017) with regard to mass loading calculations. In the technical memo, the loading
calculations were erroneously predicted to be much lower than what we report
here. In this final report we are correcting those mass loading calculations.
As previously stated, GMA performed all MT3DMS simulations without applying a retardation
factor for nitrate, a very conservative assumption that ignores the denitrification processes that
almost certainly occur along the groundwater flow path from the East Lake to the Neuse River.
Denitrification is an anaerobic respiratory pathway that reduces nitrate (NO,,) to nitrous oxide
(N20) or di -nitrogen gas (N2). This microbially-facilitated process occurs under anoxic
conditions and requires a labile, organic carbon (C) source to serve as food for the denitrifying
rum bacteria. Research on the main stem of the Neuse River and its tributaries indicates that
denitrification represents a sink of approximately 5% of the nitrate -nitrogen load to these
waterbodies (Whalen et al., 2008). Applying this finding, nitrogen loading to the Neuse River
under the maximum potential usage of the reclamation facility (quarterly wastewater inputs), is
predicated to be 8,515 pounds per year (Table 5).
SM
rMq
ow
"M Page 16
rn
V
Table S. Summary of Nitrogen (N) Loading to the Neuse River from the Wastewater Reclamation Facility
MT3DMS Model Output
IN
OUT
Potential N
Estimated N
Minimum
Credit assuming
N added to
Estimated N
% of
loading
% of
Potential
a 5% Reduction
East Lake
loading to the
Neuse River
Total
to West Lake
Total
N Credit
in Load due to
Denitrification3
Modeling
(lb/yr)
(lb/yr)
Input
(and surrounding
Input
Scenario
lakes) (lb/yr)
Max every day
at 15 yearsi
76,176
29,558
39%
17,498
23%
61%
66%
Quarterly use
u rt years2
25,638
8,963
o
35 /0
5,521
0
22 /0
0
65 /0
0
70 /o
i Modeled effluent discharging to the East Lake at a rate of 5.0 MGD and a concentration of 5.0 mg/L of
TN every day for 15 years.
Z Modeled effluent discharging to the East Lake at a rate of 5.0 MGD and a concentration of 5.0 mg/L of
TN for four equally spaced months per year. This scenario represents the realistic maximum time period
that the City would discharge to the East Lake
3Reduction in load due to denitrification based on research conducted in the Neuse River by Whalen and others (2008).
rom
MR
5.8 Model Limitations
Mm The steady-state, finite -difference model described in this report reasonably simulated
groundwater flow and nitrogen loading from the East Lake into the groundwater system.
However, due to the inherent complexities of groundwater flow systems in both space and time,
MN and considering limitations on available data and computing capabilities, there are some model
limitations.
Any model is limited by the quantity and quality of the supporting data. This model has the
benefit of representing an area that is well studied. As with any model, however, there is a fair
degree of uncertainty in the hydraulic properties of the aquifers and confining units. Site -
specific parameters were used whenever available. At locations without known data, GMA used
hydraulic property values obtained from the literature or based on GMA's experience with the
hydrogeology of this region. This model, like all models, is also limited by grid spacing.
Although the grid spacing of this model adequately allows for simulation of the hydrologic
system within the area, data input and simulation results are averaged over the entire cell.
Consequently, small heterogeneities within the system, such as small streams or well clusters,
can result in discrepancies between modeled and observed values.
It should also be noted that because East Lake does not occupy space within the model grid,
the exchange of water with the underlying aquifer is assumed to occur only vertically through
the bottom of the lake. Furthermore, this model does not accurately predict the groundwater
flow patterns in the surFcial aquifer outside of the refined model area surrounding the
reclamation facility. Surface drainage features were only modeled in the refined area where
such features could potentially be affected, and model calibration only included Sunccial aquifer
head data from this area. Because this is a steady-state model calibrated to average
groundwater levels and pumping rates for 2016, the model also does not account for potential
seasonal variation in groundwater recharge or withdrawals rates.
Although there is uncertainty in the model predictions, GMA contends that this model
reasonably simulates groundwater flow and advective transport of dissolved constituents added
to the East Lake of the New Bern Wastewater Reclamation facility based on the best available
Mr data. We further contend that the model reasonably predicts nitrogen loading to the Neuse
River resulting from the input of wastewater to the lake.
MR
6.0 CONCLUSIONS AND RECOMMENDATIONS
r,
GMA has completed a predictive modeling analysis of groundwater flow and nitrogen loading
from the City of New Bern Wastewater Reclamation Facility's East Lake. GMA's groundwater
' model focused on determining groundwater recharge effects of wastewater additions to the
East Lake and, ultimately, nitrogen loading to the Neuse River as a result of groundwater
discharge. Based upon our modeling efforts, GMA concludes the following:
Page 18
era
MW
• Usage of the East Lake Reclamation Facility for wastewater disposal by the City of New
Bern delays and reduces total nitrogen inputs from the City's wastewater to the Neuse
River.
• The artificial head maintained in the East Lake drives groundwater radially and
downward and results in a groundwater divide in the upper layers of the Castle Hayne
Aquifer beneath the lake. Groundwater recharge on the east side of that divide
discharges to the Neuse River, and groundwater recharge west of the divide discharges
to the West Lake.
• An estimated 62% of the groundwater recharge exiting the East Lake ultimately
discharges to the Neuse River. The remaining 38% discharges to the West Lake, where
it then flows as surface water back to the Neuse River, or it is lost as evapotranspiration.
• Groundwater pumping associated with the nearby Martin Marietta Clarks Quarry and the
City of New Bern production wells do not have an effect on the groundwater flow from
the East Lake Wastewater Reclamation Facility.
• Historical water -quality monitoring of wells surrounding the East Lake facility do not
indicate an increase in TN concentrations above the background concentrations that
were measured prior to the facility's first use.
• Consistent with historical water -quality monitoring, MT31DMS groundwater transport
„tea
simulations predict little to no increase over base level TN concentrations at the
groundwater monitoring well locations. TN added to the East Lake is being rapidly
diluted in the groundwater system to concentrations that cannot be distinguished from
the naturally occurring background concentrations.
• MT3DMS simulations that modeled effluent discharge to the East Lake at a rate of 5.0
million gallons per day (MGD) and a concentration of 5.0 mg/L of TN for four equally
FWD
spaced months per year predict a TN load to the Neuse River of 8,963 pounds per year,
or 35% of the total annual input. This scenario represents the realistic maximum time
period that the City would discharge to the East Lake.
Research on the Neuse River indicates that denitrification represents a sink of
approximately 5% of the nitrate -nitrogen load to the river. Therefore, under the
maximum potential usage of the reclamation facility (quarterly wastewater inputs), GMA
estimates an annual nitrogen load of only 30% of the TN added to the East Lake, which
corresponds to a potential TN credit of 70%.
7.0 REPORT CERTIFICATION
FM
This hydrogeologic characterization and predictive modeling report was prepared by
Groundwater Management Associates, Inc., a professional corporation licensed to practice
geology (NC Corporate License No. C-121) and engineering (NC Corporate License No. C-0854)
in North Carolina. We, Emma H. Shipley, James K. Holley, and Richard K. Spruill, North
Carolina Licensed Geologists for GMA, do certify that the information contained in this report is
correct and accurate to the best of our knowledge.
M
Ma Page 19
_ Emma Shipley, .G.
Project Hydrogeologist
James K. Holley, P.G. ej
Senior Hydrogeologist
Richard K. Spruill, Ph.D., P.G.
Principal Hydrogeologist
_
Page 20
rW
8.0 LIST OF REFERENCES
Eimers, JL, CC Daniel, and RW Coble. 1994. Hydrogeology and Simulation of Ground -water Flow
at U.S. Marine Corps Air Station, Cherry Point, North Carolina, 1987-90.
Fetter, C.W., 1988, "Applied Hydrogeology", Second Edition, Merrill Publishing Company,
Columbus, Ohio, p. 567.
�► Fine, JM, 2008, "Hydrogeologic Framework of Onslow County, North Carolina, 2008": U.S.
Geological Survey Scientific Investigations Map 3055, 1 sheet.
Giese, GL, IL Eimers, and RW Coble, 1997, "Simulation of Ground -Water Flow in the Coastal
Plain of North Carolina": U.S. Geological Survey Professional Paper 1404-M.
Gelhar, GL, C Welty, and KR Rehfeldt, 1992, "A Critical Review of Data on Field -Scale Dispersion
in Aquifers', Water Resources Research, vol. 28, no. 7, p. 1955-1974.
Groundwater Management Associates Inc. (GMA), 1998, "Hydrogeologic Evaluation, Abandoned
Glenburnie Quarry", prepared for Rivers and Associates, Consulting Engineer for the City
of New Bern, NC.
Groundwater Management Associates Inc. (GMA), 2006, "Monitoring Well Installations and the
Abandonment of Monitoring Wells at the Former martin Marietta Glenburnie Road
Quarry and the Mackilwean Turf Farm near New Bern, Craven County, North Carolina",
Prepared for the City of New Bern, NC, 5 pages plus figures.
fW Groundwater Management Associates Inc. (GMA), 2008-2013, Various test well and production
well reports for the 16 well sites for the City of New Bern wellfield.
Groundwater Management Associates Inc. (GMA), 2014, "Hydrogeologic Framework of Onslow
County, North Carolina", Prepared for the Onslow Water Resources Group, Jacksonville,
North Carolina, 75 pages plus figures, tables, and appendices.
M' Horton, JW, and VA Zullo, 1991, Geology of the Carolinas: Carolina Geological Society 50th
Anniversary Volume, 1st Ed., 424 pages.
Lawrence, DP, and CW Hoffmann, 1993, "Geology of basement rocks beneath the North
Carolina Coastal Plain", North Carolina Geological Survey, Bulletin 95, 60p, one Plate.
'm Lyke, WL and Winner, MD, 1990, Hydrogeology of aquifers in Cretaceous and younger rocks in
the vicinity of Onslow and southern Jones Counties, North Carolina": U.S. Geological
Survey Water -Resources Investigation Report 89-4128.
MM
Matyiko, S, 2001, "Hydrogeologic Evaluation of the Castle Hayne Aquifer System North of New
Bern, North Carolina: Implications for Disposal of Tertiary -treated Wastewater into the
'W Martin Marietta Glenburnie Quarry," Thesis, Master of Science in Geology, East Carolina
University, Greenville, North Carolina.
M
,M-9 Page 21
North Carolina Division of Water Resources (NCDWR), 2017, "Ground Water Management
Branch Map Interface" https://www.ncwater.ora/GWMS/`openlavers/ol.php?menulist=bl.
Accessed October 2017.
North Carolina Division of Water Resources (NCDWR), 2017b, "Central Coastal Plain Capacity
Use Area: Query the Data" htto://www.ncwater.orci/?page=49&menu=QuervtheData.
Accessed October 2017.
.. North Carolina Geological Survey (NCGS), 1985, Geologic Map of North Carolina, one plate.
Pohlig, KO, 2000, Groundwater Flow and Transport Analysis for a Water Reclamation Project.
NC Division of Water Quality Report.
Pollock, DW 2012. User Guide for MODPATH Version 6 — A Particle -Tracking Model for
.. MODFLOW: U.S. Geological Survey Techniques and Methods 6-A41, 58 p.
Ward, LW, Lawrence, D.R., and Blackwelder, B.W., 1978, Stratigraphic Revision of the Middle
Eocene, Oligocene, and Lower Miocene — Atlantic Coastal Plain of North Carolina,
Geological Survey Bulletin 1457-F,
Whalen, SC, Alperin, MI, Nie, Y, and EN Fischer, 2008, Denitrification in the mainstem Neuse
River and tributaries, USA, Fundamental and Applied Limnology, 171:249-261.
Winner, MD and Coble, RW, 1996, "Hydrogeologic Framework of the North Carolina Coastal
Plain Regional Aquifer System Analysis: U.S. Geological Survey Professional Paper 1404-
I.
Zheng, C and PP Wang. 1999. MT3DMS: A Modular Three -Dimensional Multispecies Transport
Model. US Army Corps of Engineers Contract Report SERDP-99.
Zheng, C. 2010. MT3DMS v5.3: Supplemental User's Guide. Technical Report for the U.S. Army
Corps of Engineers. 50 pages.
Page 22
FIGURES
'1
(y
A ^
At
BLUE RIDGE
COASTAL
PLAIN
I
I
PHYSfOGRAPHIC PROVINCES
b
_
e
i
f s
A
1 :I•
11 el
r ,
flr '
r
r •i F,
.YA
��M"
_
i"
tv"!'�'
0
��S prb ys�y��J
f4° is �
f liv 4Y� I MM•��'
k�.t�
File: ,
WASTEWATER
I
11N=4MI
P
P
A
Af
WEST EAST NEUSE
LAKE LAKE RIVER
LAYER
r-A--% r-41--I , =%
1 (SA)
SURFICIALAQUIFER
2 (LICHA
ERIVERBEtD FM.
3 (UCHA)
UPPER SPRING GARDEN
4 (MCHSC)
LOWER SPRING GARDEN (SAND)
LILL
5 (MCHA)
6 (LCHC)
- - - LOWERISPRING GARDEN (LIMESTONE)
`'" '' " `''"
LOWER CASTLE HAYNE (CLAY.)
VERTICAL EXAGGERATION = 25x
-LEGEND-
MODELAREA
GMA
HYDROSTRATIGRAPHICABBREVIATIONS:
—PROFILE
SA = SURFICIAL AQUIFER UCHA=UPPER CASTLE HAYNE AQUIFER
MCHSC = MIDDLE CASTLE HAYNE SEMI -CONFINER MCHA= MIDDLE CASTLE HAYNE AQUIFER-�
HYDROGRAPHY
LCHC = LOWER CASTLE HAYNE CONFINER
GROUNDWATER MINIOFMFM gf50<IgTFS, INC.
File: DRAWINGS/21342/
CROSS SECTION THROUGH EAST LAKE
DATE: 4/18/2018
FIG3 GRID LAYERS .MXD
SHOWING VERTICAL DISTRIBUTION OF MODEL LAYERS
PROJECT21342
NEW BERN, CRAVEN CO., NC
FIGURE 3
m
f .a
i
_ 1
i
t
CLARKS QUARRY & _
IEAST LAKE
I
L IP•i -1_
F'
.,7
LEGEND HEAD (FT MSL)
14
-2
-18 GMA
-34 0 7000'�
"iign�
��...'-
-50
1 IN = 7,000 FT
Z:\Drawings\21342\
FIG14_USG_POT.mxd
MODELED POTENTIOMETRIC SURFACE
FOR THE UPPER SPRING GARDEN AQUIFER
DATE: 4/18/2018
PROJECT: 21342
EAST LAKE RECLAMATION FACILITY, NEW BERN, CRAVEN, CO., NC
I FIGURE 14
LEGEND
PROJECT: 21342
i
v_*
EAST LAKE
i
HEAD (FT MSL)
nW 14
-2
-18 GMA
-50 � O
SO 1 IN = 7,000 FT
MODELED POTENTIOMETRIC SURFACE DATE: 4/18/2018
FOR THE LOWER SPRING GARDEN SEMI -CONFINER
EAST LAKE RECLAMATION FACILITY, NEW BERN, CRAVEN, CO., NC I FIGURE 15
E
R
pert;
1p
FIGURE 17: MODELED GROUNDWATER FLOW WITHIN A VERTICAL SECTION THROUGH THE EAST LAKE
WEST LAKE EAST LAKE
NEUSE RIVER
- -
AI
00,
RR.
4��
7�i
/T
-------
-
4,
1 Ox VERTICAL EXAGGERATION
500 FT
-LEGEND-
HEAD (FT MSQ NOTE: VIEW IS LOOKING NORTH
-10
t RIVER CELL
♦ GENERAL HEAD CELL
-
GMA
6
OSSEFFNM
4
GROUNDWATER MANAGEMENT ASSOCIATES, INC.
2
10
File: DRAWINGS/21342/
NEW BERN WASTEWATER RECLAMATION FACILITY
DATE: 12/8/2017
VERT PROFILE Fig
PROJECT21342
CITY OF NEW BERN, CRAVEN CO., NC
FIGURE 18: MAXIMUM PREDICTED TOTAL NITROGEN (TN) CONCENTRATIONS ABOVE BACKGROUND DURING 2016
TN: 12/1/201612:00:01 AM
0.0005 mg/L
- 0.0004 mg/L
0.0003 mg/L
0.0002 mg/L
0.0001 mg/L
EAST LAKE
0.0 mg/L
1
WEST LAKE
NEUSE RIVER
1 Ox VERTICAL EXAGGERATION
LEGEND-
GMA
+ RIVER CELL
♦ GENERAL HEAD CELL
OROUNCW TER MANAOEMENI ASSOCI4iES, INC.
File: DRAWINGS/21342/
NEW BERN WASTEWATER RECLAMATION FACILITY
DATE: 12/8/2017
TN MT3DMS Fi
PROJECT21342
CITY OF NEW BERN, CRAVEN CO., NC
5
4.5
4
3.5
3
E
w 2.5
rn
0
z 2
0
r
1.5
1
0.5
0
FIGURE 19. CHRONOLOGICAL TOTAL NITROGEN LEVELS IN
MONITORING WELLS SURROUNDING THE EAST LAKE
Initiated
Operation of Shallow Well
Facility —�—Shallow Wells Avg
Baseline = 2.45 mg/I
t Deep Wells Avg
Deep Well Baseline
= 1.82 mg/I
✓d y�r✓O�pc ✓d q� ✓4 O� ✓d y� ✓G�p� ✓d yp ✓4jp� ✓d yq ✓4�p✓d y� ✓` 0�f✓d yp2f2y' �74�f✓oY y0S✓✓0 yb ✓G 0
'Sf 7S7'7�6�j7��0"�7j7 ?6E �76'000600'0O Oj0;0�'OjbB0d08'08OAOQO0'o9-O-O;OQ7
DATE
FIGURE 20: PREDICTED TOTAL NITROGEN (TN) CONCENTRATIONS ABOVE BACKGROUND AFTER 15 YEARS OF
DAILY DISCHARGE OF 5 MGD OF EFFLUENT CONTAINING 5.0 MG/L OF TN TO THE EAST LAKE
TN:
12/1/201612:00:01 AM
0.008 mg/L
0.006 mg/L
0.004 mg/L
0.002 mg/L
EAST LAKE
0.0 mg/L
WEST LAKE
NEUSE RIVER
MH
1 Ox VERTICAL EXAGGERATION
-LEGEND-
GMA
* RIVER CELL
♦ GENERAL HEAD CELL
GROUNOWRIFR MRNRGFMFNI RSSOCIRlES, INC.
File: DRAWINGS/21342/
NEW BERN WASTEWATER RECLAMATION FACILITY
DATE: 12/13/2017
TN MT3DMS Fig
PROJECT21342
CITY OF NEW BERN, CRAVEN CO., NC
APPENDIX I
m"
2016 EAST LAKE USAGE RECORDS
11
Jan 2016
1 Feb 2016
1 Mar2016
I Apr 2016
May 2016
1 Jun 2016
1 Jul 2018
1 Aug 2016
1 Sep 2016
1 Oct2016
I Nov 2016
1 Dec 2016
1 FINAL
Final
Measure
W WTP:
Influent
Total influent
157.37
183.13
137.77
111.79
115.94
124.68
128.87
123.96
168.91
182.03
106.90
121.10
1662.45
Total
Avg, influent daily
5.07
6.31
4.44
3.72
3.74
4.15
4.15
3.99
5.63
5.87
3.56
3.90
4.54
Average
Max. influent daily9.12
12.59
5.69
4.85
4.74
6.53
4.88
5.13
11.34
12.88
4.21
5.26
12.88
Maximum
Max Date
23Jan
5-Feb
4-Mar
15-A r
31-Ma
7-Jun
14Jul
3-
3Se
9-Oct
2-Nov
6-Dec
Total Inn. BOD tons Removed
152.6
125.33
743.85
119.521i120.77
123.64
118.66
109.57
137.46
152.14
131.24
150.43
1585.21
Total
Total Inn. TSS tons Removed
85.64
59.47
101.32
98.49
77.58
49.82
57.6
68.76
74.48
69.50
79.43
101.55
923.64
Total
Rainfall WWTP (")
3.22
5.85
3.49
1.8
5.39
4.35
9.98
7.W
10.56
6.61
1.09
3.06
64.56
Total
Total Effluent
157.81
183.46
139.19
110.61
115.97
126.56
127.69
127.74
170.83
182.74
106.72
121.81
1671.131
Total
Avg. effluent daily
5.09
6.32
4.49
3.68
3.74
4.21
4.11
4.21
5.69
5.89
3.55
3.92
4.57
Avem e
Max. effluent daily
9.04
12.65
5.72
4.87
4.81
6.79
5.21
5.24
11.44
12.60
4.27
5.40
12.60
Maximum
Max Date
Total to River
157.81
183.46
139.19
110.61
7.86
126.56
127.69
127.74
170.83
180.58
2.83
121.81
1456.97
Total
Total to Quarry
0.00
0.00
0.00
0.00
108.11
0.00
0.00
0.00
0.00
2.16
103.89
0.00
214.16
Total
Avg. Effluent BOD m
2
< 2
< 2
< 2
< 2
< 2
< 2
< 2
< 2
< 2
< 2
< 2
< 2
Average
Avg. Effluent TSS m 1L
2.5
< 2.5
< 2.5
< 2.5
< 2.5
< 2.5
< 2.5
< 2.5
< 2.5
< 2.5
< 2.5
< 2.5
< 2.5
Average
Avg. Effluent TN m
9.11
5.24
5.9
4.73
2.61
1.32
1.35
1.69
3.21
5.37
8.27
7.76
4.71
Average
Avg. Effluent Phos. m
1.12
0.428
0.68
0.732
0.304
0.4
0.682
1.44
1.04
1.12
1.42
1.24
0.88
Average
Rainfall Quarry
3.23
6.82
3.2
1.85
3.75
4.87
9.25
6.5
11.05
5.58
0.91
3.82
60.83
Total
Quarry Level
6.5
6.3
5.8
5.2
5.6
6.1
5.5
5.1
5.2
5.3
5.8
6.1
5.710
Average
Total um MMQ to Mackllvieen
O.00D
0.000
0.000
0.720
0.704
1.519
1.031
0.214
0.349
0.917
1.011
0.000
6.4651
Total
Total s rasd at MackInvean
I O.coul
U.ucul
0.000
0.7201
0.7utI
1.5191
1.0311
U.2141
0.3491
0.917
1.011
0.000
6.465
Total
Anything out of compliance
I No
I No
I No
I No
I No
I No
I No
I No
I No
I No
No
No11
�jl
am
LAYER 1:
k
SURFICIAL AQUIFER
S
LAYER 2:
UPPER CASTLE HAYNE
AQUIFER
.;x
LAYER 5:
MIDDLE CASTLE HAYNE
AQUIFER
LEGEND
HORIZONTAL HYDRAULIC CONDUCTIVITY (Kh; FEET/DAY)
GMA
_0.01 -18 -31
-63 80 -100
-150
-5.0 -23 -50
-65 -85 -113
-270
Z:\Drawings\21342\
MODELED HORIZONTAL HYDRAULIC CONDUCTIVITY
DATE: 4/18/2018
FIG 9 Kh.pdf
DISTRIBUTIONS
PROJECT. 21342
EAST LAKE RECLAMATION FACILITY, NEW BERN,
CRAVEN, CO., NC
FIGURE 9
m
m
Clarks MW5 PW3
EL6 PW8
EL5"e- EL4
PW6 ED
EL1
Clarks RS USG
OWt
Craven 30 Supply PW5
OW2
OW3
iP
'L
0i
J
LEGEND
OBSERVATION WELL USED FOR CALIBRATION GMA
0 7000'
1 IN = 7,000 FT
zFIGIOw_013S t842\
OBSERVATION WELLS
DATE: 4/18/2018
FIG10 OBS WELLS.mxd
PROJECT: 21342
EAST LAKE RECLAMATION FACILITY, NEW BERN, CRAVEN, CO., NC
FIGURE 10
Computed vs. Observed Values
Head
11
} EL1
10 X EL2
p EL3
B • EL4
❑ EL6
8 0 ■ ELB
PW3
7 ♦ PINS
o p PW8
8A PWB
. V Clarks RS USG
E
u 6 V Clarks MWS
U } OWt
4 + i OW2
�
O OW3
3 ■
•
• • Craven 30 Supply
2 • cb
1 r2=0.92
o
B
0 1 2 3 4 5 8 7 ll B 10 11
Observed
Residual vs. Observed Values
Head
2.0
} EL1
X EL2
1.6 0 EL3
EL4
❑ EL5
1.0 • ■ • o ■ EL6
;. PW3
0.6 ♦ PINS
o + p PW8
° •A PINS
0 A V Clarks RS USG
x ♦ Clarks MW5
+ OW1
x OW2
O OW3
•1.0 • Craven 30 Supply
•1.S
-2.0
2 3 4 5 6 7 8 B 10 11
Observed
STATISTICS
Mean Residual (Head, in feet) 0.30 GMA
Mean Absolute Residual (Head, in feet) 0.64
Root Mean Squared Residual (Head, in feet) 0.79
ZFIG10ingsbratio\
MODEL CALIBRATION PLOTS
DATE: 4/18/2018
FIG 10 Calibration.pdf
PROJECT: 21342
EAST LAKE RECLAMATION FACILITY, NEW BERN, CRAVEN, CO., NC
I FIGURE 11
4
M
00,
,z �IT
mil- V
'�. I
r 7w,�1,�•
` '4 ,
f
i
a
N