This report was prepared by Jordan, Jones & Goulding, Inc. (JJ&G) for the Georgia Environmental Protection Division (EPD) to document the results of the Savannah Harbor Aquifer Working Group meeting on March 13 and 14, 2001. This report summarizes Mr. Harold E. Gill's observations and professional opinions as a hydrogeological consultant to EPD.
The following reports and information from the Sysconn.com/harbor/SEG web site were reviewed prior to the meeting:
The Aquifer Working Group Meeting Report is included for completeness as Appendix B. The major areas of discussion and concern are summarized below:
It was agreed by the working group that an initial phase (Phase I) of limited fieldwork and analysis should be completed. The limited fieldwork would include:
Considerable discussion focused on the range of reported hydraulic conductivity for the Miocene sediments and how to better define this parameter. All seemed to be in agreement that the scale of a pumping test would be preferable to that afforded by laboratory tests.
I discussed this problem with Robert Faye (retired S.E. Regional Groundwater specialist for the USGS) at the 2001 Georgia Water Resources Conference in Athens. His recommendations with which I concur are quoted below:
"Per our conversations, I will offer some thoughts regarding the computation or determination of an effective or representative leakance of the confining units that overtie the Upper Floridan Aquifer at Savannah. These comments refer specifically to the "problems" posed by the future dredging and possible deepening of the Savannah ship channel.
1. A simulation approach perhaps should be considered before planning and executing any field tests to determine leakance. Several models exist at various levels of resolution that include the Floridan Aquifer at Savannah and each of these models includes an assigned leakance for the confining unit that overties the Floridan Aquifer. To the best of my knowledge all of these models were calibrated to a steady-state condition for both predevelopment and recent conditions. Some revision of the discretization pattern would probably be required to better represent the Savannah River along the reach of proposed dredging. Once a model is selected and the discretization is modified, then various tests can be accomplished to simulate the assumed changes in leakance caused by dredging and deepening. If part of the top of the aquifer is considered to be completely exposed to the river, this area can be "seeded" with particles and, using particle tracking, simulate the distribution of river water that is induced into the aquifer. The area of exposure and the leakance in that area can also be varied to test various scenarios of conditions and pumping.
2. An alternative approach might also be useful. I recall that data for several aquifer tests conducted in the Upper Floridan Aquifer are on file at the Georgia District. If any of these tests are of sufficient detail and duration, we can use one of the calibrated models (discussed in #1 above), to simulate the water-level response observed during the aquifer tests and calibrate to a leakance in the immediate vicinity of the test site. This approach would also require some rediscritization of the model grid as well as a thorough review of the aquifer test data. Such a review would be necessary to determine if the quality of the data and characteristics of the water-level response are reasonable and sufficient to contribute to a useful or even possible evaluation of leakance. Probably a test code such as AQTESOLV or other such would be very useful in accomplishing this task."
In 1996, pumpage was significantly reduced at the Union Camp (International Paper) well field between July 13 to 22, 1996. During the 9-day period pumpage was reduced by 12,625 gal/min (18.2 Mgal/d). Recovery data was analyzed by the USGA (Warner and others, 1999) in four observation wells, two south of the Savannah River, one on Hutchinson Island and one in South Carolina. This data can be used in conjunction with the simulation recommendations by Robert Faye to further the understanding of leakance through the Miocene confining. The U.S. Geological Survey for public acceptance should carry out this work as an independent scientific agency.
Considerable discussion was included in the information from the SEG web site about the significance of vertical fractures in the Miocene sediments in the Savannah Harbor area as possible conduits for vertical leakage of saline water to the Upper Floridan aquifer. I have spent the last 25 years in Georgia investigating the impact of fractures and faulting on groundwater flow in the Floridan and Miocene aquifer systems in Coastal Georgia. If the fractured or faulted rocks are carbonates the fractures may have been solution enlarged. In this case they can become conduits for vertical migration of saline water. That is the problem of upconeing of saltwater into the Floridan aquifer from depth in the Brunswick area. The shallower clastic Miocene sediments in the Brunswick area have been displaced as much as 50 feet over short horizontal distances with no noticeable impact of enhanced vertical flow across confining layers.
It is my professional opinion that the groundwater modeling and particle tracking recommended by Robert Faye and the Phase I work recommended by the Savannah Harbor Aquifer Working Group be given high priority for completion. The U.S. Geological Survey as an independent scientific agency should conduct the groundwater modeling and particle tracking. The Miocene pore water sampling dating and completion of the additional cluster wells are necessary to establish current conditions in the Savannah Harbor area. This work should continue with the Corps of Engineers and SCDHEC (Camille Ransom).
The marine seismic surveys recommended by Dr. Jim Henry are necessary to establish the areal extent of the paleochannels in the vicinity of the proposed harbor expansion. Completion of a monitor well in the paleochannel at the Bull River rotosonic site (60 to 75 feet) is needed to sample pore water and determine aquifer parameters. Addition of an Upper Floridan well at the Bull River site will provide additional information about vertical leakage (cluster well site).