MEMORANDUM
TO: Aquifer Committee Members
The City of Savannah is
recommending to the Aquifer Committee of the SEG that GPA fund additional scientific
studies to determine the risks associated with the Harbor deepening to the Floridan
Aquifer. The additional work recommended is
detailed in the attached scope of work. It is
also recommended that a joint partnership of two credible and knowledgeable experts in the
field perform the studies. These two experts
are Hydrovision of Atlanta and the Corps of Engineers (COE). Since the COE was involved in the initial Tier I
EIS study and report on the aquifer, it is recommended that the COE continue this work. In order to satisfy the varied interests on this
issue, Hydrovision will provide another level of credibility needed to answer the
scientific questions.
It is further recommended
that the aquifer study team provide a joint conclusion based on the data, expertise, and
professionalism of each of their organizations and individuals as to the appropriateness
of the Aquifer Committee's endorsement, or opposition,
to the deepening of the harbor relative only to the safety of the Aquifer.
It is further recommended
that the joint aquifer study team report directly to the Aquifer Committee with all
preliminary reports, updates, etc. and once the report is completed, the joint aquifer
study team will present the final report to both the Aquifer Committee and the Georgia
Ports Authority for review and further action.
I understand that the
9/8/2000 meeting has been canceled due to requests from various parties. This postponement may provide you the opportunity
to review this scope of work in order to prepare for discussion at our next meeting.
HJ/do
(See attachment below)
September 6, 2000
Mr. Harry Jue
Director, Water Operations
City of Savannah
706 Stiles Ave.
Savannah, GA 31402
RE: Proposal for an Assessment of the Potential
Effects of Savannah Harbor Expansion on the Upper Floridan Aquifer
Dear Harry:
HydroVision is pleased to
submit the following proposal for an assessment of the potential effects of Savannah
Harbor expansion on the quality of water in the Upper Floridan aquifer. The study proposed
herein would supplement the work done by the U.S. Army Corps of Engineers, (Potential
Ground-Water Impacts? Savannah Harbor Expansion Feasibility Study, March 1998). The
proposed work would answer the basic, yet critical questions posed by the City of
Savannah, the Aquifer Committee, the SEG, and all those having an interest in the subject:
Would dredging cause
leakage from the Savannah River/Atlantic Ocean to the Upper Floridan aquifer to occur or
increase as a result of harbor dredging?
· What level of confidence can be placed
on the results? and thus
· What degree of risk in contaminating
the aquifer would be incurred if the proposed dredging were to take place?
In order to address these
questions, we first need to determine the following:
· Is water presently leaking into or
through the confining unit separating the Savannah River/Atlantic Ocean and the Upper
Floridan aquifer?
· If leakage is occurring, at what rate
is it leaking and has it reached the Upper Floridan aquifer?
· If leakage has reached the Upper
Floridan aquifer, has it migrated downgradient and increased salinity in the aquifer?
To be able to answer all
these questions and to enhance the level of confidence in study results, additional study
is proposed that would include the following principal work elements:
· field determination of salinity with
depth in the strata of the confining unit separating the river and the aquifer;
· ground-water flow and variable density
solute (saltwater) transport modeling;
· field determination of hydraulic
properties, if results of first two tasks deem it necessary; and
· monitoring of head and salinity
before, during, and after dredging.
HydroVision would suggest
the following tasks:
Task I Evaluation and Analysis of Existing Data and
Previous Investigations
Data collected and
investigations conducted chiefly by the U.S. Army Corps of Engineers, and to a lesser
extent, the U.S. Geological Survey, Georgia Environmental Protection Division, South
Carolina Department of Health and Environmental Control and Department of Natural
Resources, Georgia Ports Authority, academia, and the work of consultants, would be
gathered, compiled, assimilated, and interpreted. Existing ground-water flow, solute
transport, and other pertinent models would be evaluated for applicability of both input
data sets and model results. Completing this task will help in the development of the
conceptual hydrogeologic framework and model and in determining optimal locations for
field data collection, including test-and monitoring-well drilling, coring, and aquifer
testing. An equally important purpose of this task is to enhance and increase the
available data that will be used in the proposed numerical models (Task III). This task is
subordinate to, and basically is used in support of, the remaining tasks.
Task II Determination of the Salinity of Water in the
Confining Unit? Extraction and Analysis of Fluid from Cores
Cores of strata that
separate the Savannah River and the Upper Floridan aquifer would be taken, fluid extracted
from the core, and salinity of the fluid analyzed and determined. The main purpose of this
task is to determine salinity distribution in the confining unit separating the Savannah
River/Atlantic Ocean from the Upper Floridan aquifer. Completion of this task will help
answer questions as to whether leakage is presently occurring, and if it is, at what rate,
and whether it has reached the Upper Floridan aquifer.
It is likely that a
floating, jack-up drilling rig would be used to take the cores. Surface casing would be
set to and slightly into the riverbed or seafloor, and ambient river or ocean water washed
out using freshwater. Freshwater would be used to mix with the drilling/coring mud during
coring,
and be under sufficient
pressure to exclude river or ocean water. Cores would be taken using HQ wireline equipment
and methods. Fluid would be extracted from the core while still on the drilling rig, thus
eliminating the preparation of the core for shipment to the laboratory. Onsite fluid
extraction also would result in a more representative sample of the pore water. Analysis
of the water, including that for chloride concentration, would be conducted in a
laboratory. Any core not used for fluid extraction would be saved and preserved as
necessary for any further analysis, as part of this, or any subsequent study. All cores
would be used for lithologic and stratigraphic analyses and as "ground truth"
for correlation with other nearby data from cores, borehole geophysical logs, and
subbottom geophysical survey collected and reported in the Corps report.
After coring the hole,
borehole geophysical data would be collected in the cored hole from riverbed or seafloor
to the top of limestone, near the top of the Upper Floridan aquifer. The hole would be
abandoned by sealing using Bentonite from bottom to riverbed/seafloor, and surface casing
recovered.
It is proposed that a
profile of salinity with depth be estimated by taking about 6-10 cores in the strata of
the confining unit that separates the Savannah River and the Upper Floridan aquifer. Cores
used for fluid extraction would be comparatively closely spaced near the top of the
confining unit and be more widely spaced with increasing depth. Also, fewer cores would be
taken where the confining unit is thinnest; and more cores taken where the confining unit
is thickest.
Profiles of salinity are
important at a minimum three locations: 1) where the confining unit is thinnest, probably
along the reach between station +9+000 and ?3+000; 2) where the vertical hydraulic
conductivity is greatest, probably having a paleochannel; and 3) where the difference in
head between that in the river (Savannah River stage) and aquifer (Upper Floridan aquifer
head) is greatest, probably near pumping at International Paper Co. These three locations
would be at sites that most likely have an extreme value for one of the three variables in
the ground-water flow (leakage) equation. It is possible that another location may be
selected that might combine two or more of the worst-case hydrologic variables, and
possibly additional consideration of density of the contributing water (Savannah River,
Atlantic Ocean).
Successful completion of
this task will answer the question as to whether the Savannah River is presently leaking
water down into and/or through the confining unit. Results of this task, and of Task III,
discussed next, also should give an indication as to the rate of such movement and
quantity of the leakage, and its proportion to the ambient, laterally flowingfreshwater in
the Upper Floridan aquifer.
Task III Stream-Aquifer Leakage and Solute Transport
Modeling
An estimate of the areally
distributed volumetric flow rate of leakage of brackish or saline water from the Savannah
River/Atlantic Ocean through the intervening confining unit to the Upper Floridan aquifer
would be computed by developing and using a ground-water flow and solute transport model
of stream-aquifer flow (leakage).
Leakage is driven by the
difference in water-level between the Savannah River stage or Atlantic Ocean (sea level)
and the head in the Upper Floridan aquifer; thickness, vertical hydraulic conductivity,
and specific storage of the confining unit; width (wetted perimeter) and length of the
river segment under consideration, and time. All these parameters vary spatially in this
heterogeneous setting, and will be considered as such in the model. In addition, leakage
is governed by density differences between the river water or seawater (more dense) and
aquifer water (less dense) and the hydrodynamic dispersion and advection of the brackish
or saline water as it migrates through strata of the confining unit. Accordingly, a model
having the ability to handle variable density flow and transport would be employed. The reach of the river and offshore channel would
be discretized into segments governed by stream geometry and data variability and desired
resolution and accuracy.
Leakage that has taken
place to date and the resulting salinity in the sediments observed today are the
consequences of the hydrologic conditions described in the preceding paragraph as they
have changed with time. For example, the difference in water level between the river and
the aquifer has changed markedly since the early 1900's as a result of pumping-induced
head reduction in the Upper Floridan aquifer. In addition, thickness of the confining unit
has already been changed as dredging in the past has reduced that thickness incrementally.
These historic conditions would have to be considered in model calibration. Calibration of
the ground-water flow and transport (leakage) model would be made against field data of
salinity of pore water in the confining unit strata that would be gathered as proposed in
Task II.
Leakage and salinity would
be computed for historic hydrologic conditions as well as for present-day hydrologic
conditions. "Historic hydrologic conditions" would incorporate data on the
potentiometric surface at selected times and thickness of the confining unit bracketed by
past dredging activity. "Present-day hydrologic conditions" are those that would
incorporate all hydrologic data gathered as part of the Feasibility Study report (U.S.
Army Corps of Engineers, 1998), hydraulic head data from the May 1998 potentiometric
surface of the U.S. Geological Survey (Peck and others, 1999), and salinity data gathered
in Task II.
Following calibration,
leakage and salinity would be computed for conditions that might result from harbor
expansion, in which a known thickness of sediment of the confining unit is dredged and
thus reduced in thickness, and hence, increased in leakance. Comparison of leakage between
existing and proposed conditions in the reach of the Savannah River considered for
dredging would give estimates of the increase in leakage expected to result from dredging.
Confidence in these results would be a function of the goodness-of-fit between observed
data and values simulated by the model for present-day conditions.
A second model would be
developed to simulate the migration of brackish or saline water as it enters and moves
downgradient in the Upper Floridan aquifer. This model also would account for the relative
densities of the leaking river water and the ambient, freshwater of the Upper Floridan
aquifer. The model would be coupled with the ground-water flow and solute transport
(leakage) model. Calibration of the solute transport model would be made against existing
field data of salinity of water from the Upper Floridan aquifer at well locations in the
vicinity of the Savannah River and the pumping center in Savannah.
A sensitivity analysis will
be conducted using the models. Sensitivity analyses are made to test the appropriateness
and importance of the various input parameters used in the models. Input parameters, such
as head of the Upper Floridan aquifer and stage of the Savannah River, would be varied
within a reasonable range of plausible field values. This range, or end points of the
range, would be input into the models, simulations made, and the resulting rates of
leakage and flow noted and assessed. One valuable application of sensitivity analysis is
that, combined with results of conditional simulation, it can provide important
information that can be used to direct field data collection, making that effort extremely
cost and time effective. For example, vertical hydraulic conductivity is the least known
of the input variables, and the most difficult to obtain. What is not known at this time,
is its importance within its plausible range of field values to the leakage computation
(ground-water flow and transport simulation). It is well established that field
determination of this parameter by using aquifer testing would be needed to get the
"best" field value. What is not known, however, is how much "better"
those determinations would be over those already available (having been determined in the
laboratory). The proposed ground-water flow and transport (leakage) model would be used to
evaluate the cost-effectiveness of obtaining additional field data.
Vertical hydraulic
conductivity would be varied in a range beyond that expected to occur in the field to test
leakage rates that are representative of a worst-case scenario (overly large values of
vertical hydraulic conductivity resulting in simulated leakage that would be greater than
plausible). Vertical hydraulic conductivity input to the model would be greater than that
expected in the field, and substantially greater than that determined in the laboratory.
These values could represent the effects of possible faults, cracks, and other
permeability enhancing features in the strata of the confining unit. If simulation results
using these values indicate that the hypothetical, worst-case leakage and transport were
acceptable (no adverse leakage of saline water into the Upper Floridan aquifer), then
there would be little reason to collect additional hydrologic field data.
Another form of sensitivity
analysis would be conducted to represent a varying range of future hydrologic conditions
and/or "what-if" scenarios. For example, if it is projected that pumpage on
Tybee Island were to be increased, then the resulting lowered head in the Upper Floridan
aquifer would be incorporated into the models and predicted leakage and downgradient flow
would be simulated.
Task IV Hydrogeologic Field Testing
The need for hydrogeologic
field testing for determination of in situ hydraulic characteristics would depend on the
results derived from Task III, Stream Aquifer Leakage and Solute Transport Modeling. The
hydrogeologic field testing would be conducted if more accurate vertical hydraulic
conductivity of the confining unit were needed. Of course, other hydraulic characteristics
and aquifer and confining unit data would be gathered, analyzed, and interpreted if this
task is conducted. This task would have three elements: A? Test-Observation Well
Installation; B? Aquifer Testing; and C? Aquifer Test Data Interpretation and Analysis.
Element A? Test-Observation
Well Installation
Installation of 4 wells is
planned, and would be contracted out to a licensed water-well contractor. These wells
would be constructed specifically for use in the aquifer tests, and would consist of the
following:
· Two wells open to the Upper Floridan
aquifer: both wells drilled through sediment and into the top of the aquifer, to about 150
feet below land surface; cased and pressure grouted to land surface; then drilled open
hole to a depth of about 360 feet below land surface.
One well, to be used as the pumped well, would have minimum 8-inch steel casing,
and drilled open hole about 7 7/8 in. in diameter. The other well would be used as an
observation well and have 4-inch steel casing, and drilled open hole about 3 7/8 in. in
diameter.
· Two wells screened in the Miocene sediment overlying the Upper Floridan: both wells would be used as observation wells, constructed of 2-inch diameter, schedule 80 PVC casing and screen. The screened intervals in the two wells would be about 90-100 feet below land surface, in the clay confining unit, and 70-80 feet below land surface, in water-bearing sand.
Based on the data review in
Task I and resulting location of the well installation and field testing, lithologic and
borehole geophysical log data may be sufficient and available for use in this study. This
would be the situation if the selected site were located near a site tested by the U.S.
Army Corps of Engineers during their initial study. If that were the case, additional
lithologic and geophysical data would not be needed. If that should not be the case,
HydroVision would need to sub-contract that work at additional cost to the study. The
additional work would be that of collecting continuous geologic core from one hole by the
water-well contractor, and borehole geophysical logging at least one hole by a company
capable of providing that service. It is likely that the U.S. Geological Survey may be
interested in collecting borehole geophysical log data at the selected site if their data
coverage in that area were sparse.
Element B? Aquifer Testing
Aquifer tests would be
conducted at the site and would be for the purpose of determining in situ hydraulic
properties of the aquifer and confining unit. (Note: the terms "aquifer tests and
aquifer testing are used because of their common usage; however, tests and testing will be
conducted of the entire hydrogeologic setting, which would include water-bearing and
confining strata.) Groundwater levels, river stage, and barometric pressure would be
monitored for about 48 hours to provide ambient hydraulic conditions. A step-drawdown
aquifer test, in which the pumping would be incremented and response observed, would be
conducted to gather preliminary data and information needed to determine the optimum
pumping rate for the main aquifer test. During the main aquifer test, the Upper Floridan
aquifer well would be pumped at a constant rate, the discharge measured, and the response
in water level in the three observation wells measured and recorded. In the event
additional wells are available in the area of the test, response in water level in the
well or wells also would be measured and recorded. The pumping phase of the test would be
a minimum 72 hours in duration. Pumping would cease and conditions monitored for at least
48 hours to assess recovery. The pump and power source would be supplied by the water-well
contractor.
Element C? Aquifer Test Data
Interpretation and Analysis
Aquifer test data would be
recorded, analyzed, and interpreted by HydroVision personnel. State-of-the-science methods
for interpretation of the aquifer-test data would be used. Hydraulic characteristics of
aquifer transmissivity and storage coefficient, and vertical hydraulic conductivity
and specific storage of the
confining unit, would be derived from the aquifer test results.
Task V Report Preparation
A report summarizing study
results would be prepared and copies provided.
Task VI Long-Term Monitoring
Monitoring of the
ground-water level in and salinity of the Upper Floridan aquifer would be done in selected
locations along the Savannah River before, during, and after dredging. Data from the
monitoring wells would be used as an "early-warning system" in which chiefly
salinity would be
determined periodically to
observe any possible increase in salinity thatmight be caused by the dredging. The
monitoring wells would be sampled and analyzed for salinity prior to dredging to provide
background conditions.
Five wells would be used
for this task, and would be constructed at locations determined by an analysis of the
conceptual hydrogeologic model and interpretation of results of simulation conducted as
part of Task II. Wells would be 4-inch wells constructed in a manner similar to the Upper
Floridan aquifer observation well planned for Task IV, Element A. The wells would be
capped securely, and 1-inch PVC drop pipe permanently installed in each well for ease in
sampling by blowing using an air compressor. Frequency of sampling would be determined by
the interpretation of results of simulation conducted as part of Task II, and by the
ongoing dredging activity. Maximum frequency probably would be during, and in the area
most likely influenced by, dredging. Frequency probably would be no greater than quarterly
but no less frequent than annually.
References
Peck, M.F., Clarke, J.S.,
Ransom, Camille III, and Richards, C.J., 1999,
Potentiometric surface of
the Upper Floridan aquifer in Georgia and
adjacent parts of Alabama,
Florida, and South Carolina, May 1998, and
water-level trends in
Georgia, 1980-98: Georgia Geologic Survey, Hydrologic
Atlas 22.
U.S. Army Corps of
Engineers, 1998, Potential ground-water impacts;
Savannah harbor expansion
feasibility study.
Schedule
The study would be
completed in about 6 months.
Cost
HydroVision would be glad
to cost out this study if you desire. It is likely that some of the components of work as
proposed might be conducted by other study partners.
In summary, we believe that
conducting the study proposed in this scope of work could answer all those questions posed
on page 1 of this document to the satisfaction of all involved. If you have any questions,
please feel free to call Rick Krause (X109) or me (X103). We appreciate the opportunity to
be able to assist you with this important water-resource evaluation.
Sincerely,
Ram Arora, Ph.D., P.G.
President