(Closed to new input, but retained for reference value)
This page is established as agreed to in the Aquifer Committee during the meeting of September 22, 2000. This page is intended for scientists and other specialists to initially share and exchange their ideas and views about the aquifer. The contents posted reflect only the views of their authors. They do not represent the opinion(s), position(s), view(s), or consensus of the Georgia Ports Authority; the Stakeholder's Evaluation Group; or the Aquifer Committee.
Content for this page should be submitted to the chair of the Aquifer Committee, Christopher J. Schuberth, email: schubech@mail.armstrong.edu; telephone (912) 921-7332.
Information forwarded by the chair will be posted as received on this page. The author will be included for information along with the date of posting. New information posting will not be announced by the webmaster. Interested parties should check the page contents periodically.
From Chuck Watson on 11/7/00:
Some thoughts on numerical modeling of potential impacts of the deepening project on the acquifer. In the proposed scopes of work which are being discussed, one of the elements being discussed is the simulation of the impact of the project via numerical models. Since numerical models are "my field", here are a few thoughts that element:
1) Rather than concentrating on developing or finding a single "best" model, multiple models and approaches should be run. Even simple models can provide valuable information, especially as they are less expensive to run than more complex ones, and can help shed light on sensitive aspects of the process. With some exceptions, I would rather see 100 runs of several simpler models with well understood, statistically verifiable performance than 1 or 2 runs of a single "wundermodel" with less well understood performance and interactions.
2) Sensitivity tests should be run on *all* input data and parameters (especially calibration coefficients), not just a selected subset. This data should then be analyzed to provide projection limits. This work should proceed in conjuction with (or even in advance of, using existing data) additional field data collection, so that more effort can be concentrated on sensitive areas, and measurements made within the sensitivity of the model(s).
3) For models which require calibration, the validation data should be ENTIRELY INDEPENDENT of calibration data. BOTH data sets should include extreme event data if at all possible, but if little extreme event data exists, it should be used ONLY for validation.
4) Simulations should be verified and run for long enough time to assess long term impacts - maybe 50 years? This implies multiple runs with various groundwater pumping scenarios.
5) The model(s) should have demonstrated prognostic ability. In other words, proving the model works for a static situation is nice, but does not prove that it will react correctly if the static situation is changed. These must be "blind" progs against data which was not used in calibration (see 3).
Chuck Watson cwatson@methaz.com
From Fred Rich on 11/7/00:
Tell me what is wrong about the proposed procedure detailed by Hydrovision? I have yet to hear anyone state clearly why that approach is not workable. By contrast, I have heard a number of comments about the unreliability of slug tests as a means of determining the permeability and tranmissivity values that we are interested in. Perhaps there is only one really good way to go, i.e., by following the advice of Hydrovision. Perhaps the reason there is no discussion is that there is nothing to discuss. I'd like to know myself.
Another issue has come to the fore in recent weeks. I have been contacted by Ms. Pat Ortmeyer, Field Director for Nuclear Waste Issues, Women's Action for New Directions, and she is interested in the contaminant content of the sediments in the Savannah River, particularly as they might relate to the Savannah River Site. I have provided her with copies of minutes, and the statement made by the five scientists that relate to the concern over aquifer integrity. She is in the process of organizing a hearing, under the auspices of Georgia State Senator Regina Thomas to gather information dealing with the presence, and mobility of industrial wastes in the Savannah River sediments. We will probably be hearing more from Ms.Ortmeyer and Senator Thomas, and they will be watching closely as the aquifer committee and its experts either do, or don't do something.
Fred Rich frich@gsaix2.cc.gasou.edu
From Jim Reichard on 11/8/00:
1) I would like to reiterate that performing permeability tests on cores from the aquitard only gives us the hydraulic conductivity of the material's matrix, hence a MINIMUM conductivity value. The reason for this is due to the low probability that some type of large-scale heterogeneity (i.e. fracture) will be present in the small sample used in the analysis. Obviously, if fractures exist in the subsurface then the field (i.e. bulk conductivity) may be much greater than the matrix conductivity determined via the cores. Since we are concerned with the potential rate of saltwater moving downward through the aquitard, we must not use minimum (matrix) values of conductivity, but rather bulk values which are more representative of the true field conditions.
2) At the last meeting there was some discussion as to the type of in-situ aquifer test that would give us the most accurate determination of the bulk (i.e. field) conductivity of the aquitard. The two basic types of aquifer tests are referred to as slug and pumping tests. An important difference between the two is that the radius of influence of a pumping test is much greater than that of a slug test, hence the conductivity determination from a pumping test is generally superior. In spite of this, it was suggested that slug tests be performed, as they are the method of choice used to satisfy regulatory requirements at sites where groundwater contamination has occurred. It should be kept in mind that one of the primary reasons slug tests are used in the environmental field is because they can be performed without removing water from the formation being tested. This can be very important when one considers the fact that during pumping tests large volumes of water are removed from the subsurface. If the site is contaminated, then this water must be disposed of properly. The other reason slug tests are common is because they require only a single well and are simple to carry out, hence are relatively inexpensive. In the situation we are dealing with the Floridan aquifer, neither of these two considerations applies. There is no contaminated water to deal with and the additional cost and complexity of pumping tests are very trivial compared to what is at stake. It should be mentioned that slug tests could be performed on the observation wells installed in the aquitard; wells which would be required for the pumping tests. Such duel use of these wells could provide additional conductivity data that may prove useful.
3) Fred Rich recently asked about the merits of Hydrovision's proposal. From a hydrogeologic perspective I feel that their proposal is quite sound and well thought out. However, this doesn't mean it can't be improved - that is the purpose of the expert's page! For example, Chuck Watson's recent modeling suggestions could be incorporated into our final proposal. What he suggests are just good modeling practices and are more detailed than that what is described in the Hydrovision proposal. The only problem might be in the availability of the various data sets needed to do everything he suggests. However, due to the critical nature of the project, no effort should be spared to gather the required data.
Jim Reichard - jreich@gasou.edu
From Bill McLemore on 11/8/00:
To: Chuck Watson:
Before discussing a modeling approach it is important to remember that there are no acceptable models. Another group of stakeholders has already determined that the exisitng models are not appropriate for assessing the salt-water intrusion on the Upper Floridan Aquifer in the Savannah area. In other words, exisiting models and variations thereof no longer are acceptable in the geological and engineering community for assessing salt-water intrusion in the Savannah area. Modeling experts have recommended several new models for consideration and there appears to be a general concensus that SUTRA 3-D is the most appropriate model to assess salt-water intrusion in coastal Georgia. Currently the USGS is attempting to do this as part of the Sound Science Initiative. The USGS, however, has not actually done this. Until (1) the Upper Floridan Technical Advisory Committee accepts SUTRA 3-D and (2) the Director of the USGS approves SUTRA 3-D, this model can only be said to be "under-development". Obtaining both acceptances is several years away. While we hope SUTRA 3-D will be good and viable model, there is no absolute certainty of this. Further, data being gathered to support the USGS's SUTRA 3-D is regional in nature not specific to the issue of the Savannah Harbor deepening. Acceptance of models is an expensive and time consuming process (i.e., multi-year). Going into an office and plugging numbers into a computer will lead nowhere.
To: Fred Rich:
There are good and bad points in any method of measuring hydraulic conductivity. There is no "best" way. Laboratory measurements may be superior to aquifer tests and vice versa. Hydraulic conductivity measurements should be designed to address the conditions at hand. For an assessment of slug tests, I suggest you read "The Design and Performance of Slug Tests", 1998, by James Butler, Lewis Publishers. This is excellent book on slug tests and provides techniques for obtaining quality measurments and properly interpreting the measurements. In his concluding paragraph, Butler writes: " A slug test may, therefore, be one of the most appropriate approaches for estimation o fhydraulic conductivity at many sites. A properly designed, performed, and analyzed series of slug tests can yield estimates that arequite reasonable representations of the bulk hydraulic properties of the formation". A skilled hydrogeologist can obtain reliable hydraulic conductivity measurements from slug tests. The methodology is used quite commonly.
Bill McLemore Bill_McLemore@mail.dnr.state.ga.us
From Chuck Watson on 11/9/00:
There is a real problem in that, as Bill McLemore pointed out, a formal acceptance and evaluation process of a model can easily take several years. I have extensive experience with both sides of the model creation and acceptance process, both as a developer of complex numerical models and by sitting on commissions, auditing models, and in developing standards for the application and acceptable use of models. Given the rapid pace of development of numerical models, that process just about insures that not only are accepted models technically obsolete, but the standards used to gauge acceptance are often obsolete as well. Another problem is that models "accepted" for one narrow use may not be valid for another, even closely related, use. Therefore, it is often the case that models which have not been formally "accepted" or "evaluated" must be used in the policy decision process. This isn't necessarily a problem as long as the normal standards of analysis are followed. Anyone presenting scientific results, be they chemical analysis, numerical models, or reading entrails (which is often pretty much the same thing!) must have a defensible methodology, should conduct sensitivity tests, and present confidence bounds sufficient for the user of the data to be able to assess its validity.
The data collection process is a critical area for this discussion. Subsurface geology must be inferred from a statistically very sparse set of samples, and a data base of the subsurface features developed for usein the analysis and modeling process. This may be the largest source of uncertainty in the modeling effort, and could present an insurmountable problem. If , as Fred Rich et. al. have suggested, there are meter-scale fractures sufficient to cause substantial variability in the miocene's permeability, a sufficiently dense sampling scheme to characterize the layer for decently accurate modeling may be impractical (unless, of course,you want the county to look like that golf course in the movie "CaddyShack", where a gopher had dug thousands of holes . . . ).
In designing the scope of work, it should be kept in mind that numerical modeling may not be appropriate or practical for this problem. If sufficient data can not be collected to validate the model(s) and establish prediction limits over the range of required outputs, and if there is a great amount of uncertainly in the input data, then it would be pointless to churn out numbers even using an "accepted" model. However, I wouldn't either accept or reject a model simply because it had or handn't been evaluated as part of a formal process, as opposed to being demonstrated to be adequate for the task at hand (which should be done as part of the scopt of work regardless of any independent approval).
Chuck Watson cwatson@methaz.com
From Bill McLemore on 11/9/00:
To: Jim Reichard:
Actual field data do not support your comment regarding laboratory measurements of hydraulic conductivity in the Miocene Aquitard. While, what you say is true if there are significant discontinuities; however, this is not the case we are dealing with. The Corps had extremely good core recovery and discontinuities were few and far between. In fact, they had over 400 of coring with 100% recovery. The Corps' core descriptions are very precise and provide considerable information on subsurface lithologies. Because of the very high core recovery, it is possible for a competent professional to select representative samples for laboratory testing and it is possible to obtain representative measurements of hydraulic conductivity.
The Corps' scientific approach underwent colleague review back in 1997 and deemed to be acceptable by both EPD and the USGS. Further, laboratory measurements of hydraulic conductivity of Miocene were used by the USGS in development of their numeric models. Laboratory measurements of hydraulic conductivity are routinely used in the professional practice of geology and are accepted by regulatory agencies.
As I pointed out to Fred in an earlier e-mail, there is no "right" or "wrong" way for performing a hydraulic conductivity measurements.
Aquifer or pump tests are not inherently superior to slug test or to laboratory tests. All methods have their pros and cons. Aquifer tests, however, are expensive and if they do not meet the governing assumptions are relatively easy to discredit. If the governing assumptions can not be demonstrated to be valid, then an aquifer tests is merely an merely an expensive boon-doggle. At this time, no-one has provided any information to demonstrate that an aquifer test would or would not provide any meaningful data.
In environmental decision making the key word is "demonstrate"; basically this means that every idea or concept has to be demonstrated. If you can't demonstrate it, it just won't fly.
Bill McLemore - Bill_McLemore@mail.dnr.state.ga.us
From Chris Schuberth on 11/10/00:
The discussions among the experts so far posted on this page are of profound importance. Each contributor applies his/her expertise to address some specific aspect of the hydrology of the Floridan aquifer ranging from the realities and technical aspects of dynamic fluid modeling to differing opinions on the details of the mechanical aspects of aquifer performance. Through all this dialogue, I suggest that we don't lose sight of our principal object and that we don't get caught up in technical details, unless, of course, resolving these technical details one way or the other leads us to a better and clearer understanding in reaching our primary goal. And what is our goal? To carefully and reflectively craft a recommendation(s) to SEG based on how to best answer the question whether further deepening will or will not, or to what degree the proposed deepening, adversely impacts the regional freshwater supply provided by the Floridan aquifer. Four official documents, each essentually trying to answer this central question, now lie on the table: (1) USACE 1998 report; (2) Camille Ransom's Memorandum in which he asks three critical questions; (3) the Five Signatory Scientists Memorandum; and (4) the "second opinion" HydroVision report prepared for The City of Savannah of USACE 1998 report and submitted for our review. These four documents have not been arranged in any priority but are randomly listed just the way they popped into my head. We all understand that USACE 1998 report has been, and continues to be, the springboard, the baseline, document. Chris Schuberth - schubech@mail.armstrong.edu
From Chris Hemingway representing Golder Associates on 11/10/00:
Below is a summary of the opinions of Golder Associates’ hydrogeologists regarding the tasks necessary to study saline leakage into the Floridan Aquifer induced by the deepening of the harbor.
THE KEY QUESTIONS
* Is salt water currently moving through the confining strata (Miocene sediments) overlying the Upper Floridan Aquifer?
* If so, how long has this been happening and at what rate is it happening?
* Would removing up to 10 feet from the total thickness of this confining stratum increase the risk of such leakage occurring or significantly increase the rate at which salt water is moving?
RECOMMENDED APPROACH
Golder believes the most expeditious approach to answering the key questions is to perform the following tasks in sequence, evaluating the need to continue the study after completion of each task:
1) Define the magnitude and extent of Miocene pore water salinity;
2) Determine the spacing, orientation, and other aspects of the fractures;
3) Determine the vertical permeability of the Miocene sediments;
4) Correlate historic drawdown in the Floridan and dredging in the Savannah River with the advancement of the salinity front in the Miocene sediments;
5) Predict the effects of additional dredging.
ADDITIONAL DISCUSSION
1) Pore Water Sampling
Drill a series of boreholes through the Miocene sediments underlying the Savannah River. Recover cores of these sediments at various depth intervals for and extract pore water to determine A) pore water salinity; B) illustrate a salinity profile; and C) estimate the rate at which a salt water front would move downward under a vertical hydraulic gradient. Some samples should come from a portion of the river in which the salt water front is not present
This approach would be very useful as a preliminary screening tool to determine if salt water leakage is indeed occurring. We may find that no significant vertical migration is occurring, in which case further investigation may not be necessary.
2) Fractures
Core samples collected for pore water sampling should be carefully examined for fractures prior to any other testing. As some investigators have stated that these fractures are nearly vertically oriented (high angle), Golder suggests oblique (horizontal) drilling in the Miocene and careful examination of the core. Fracture sets should be fully described so that their influence on the hydraulic characteristics of the Miocene sediments can be determined and the fracture density can be approximated. Additionally, some fractured and unfractured samples should be subjected to the laboratory permeability tests described in #3.
3) Vertical Permeability
A reasonable way to treat the confining stratum is by using a bulk vertical hydraulic conductivity (Kv) value to represent the entire unit. This value could be obtained from a carefully planned, representative array of lab permeability tests, or a pumping test. While we believe that a pumping test derived Kv value is typically more representative of the bulk aquifer/aquitard properties since it represents a combination of fracture and matrix conductivity, we question whether it would differ significantly from the mean value derived from lab permeability testing. This, of course, is dependent on the fracture spacing and transmissive properties of the fractures.
Furthermore, to properly stress a prolific aquifer such as the Floridan such that the contribution and effects of all other local pumping could be eliminated, and to elicit a response from observation wells at the top of the confining unit, would be extremely difficult. Also, in our experience, derivation of a Kv value from pumping test data is inherently problematic using standard techniques (multi-level piezometers or leakance estimates based on a flattening of a log/log drawdown curve) under ideal conditions.
Therefore, we suggest that before proceeding with a pumping test, the Kv should be assessed by laboratory tests of core samples. Samples should be carefully selected to represent the range of characteristics of the Miocene sediments (i.e. lithology, fracture density).
4) Historic Drawdown and Salt Water Leakage
It may be possible to determine the rate of downward salt-water migration by examining historic pumping records and harbor dredging records and comparing these with the salinity profile. However, the magnitude of the downward vertical gradient has likely changed over time along with the rate of movement of downward vertical leakage, so it may be difficult to make predictions based on existing conditions.
5) Simplistic Predictive Modeling or Calculations
Once an agreed upon, representative bulk Kv value is chosen, this value could then be used to perform some very simple modeling exercises for use in predictive scenarios to determine the effects of decreasing the thickness of the confining bed. Alternatively, some even simpler calculations based on Darcy’s Law could be performed to estimate the travel rates and determine the impact (if any) of the harbor deepening project. However, modeling is pointless without first gaining additional directly observed data. For a simple example, varying Kv by one to two orders of magnitude, which would be a typical sensitivity approach, would change the rate of saltwater movement 10 to 100 times, and could raise further questions rather than solving the problem.
Chris Hemingway representing Golder Associates - chris_hemingway@golder.com
From Rick Krause on 11/14/00:
After a few pertinent and germane postings, the thread took off on tangents. I suggest we focus on the stated problem--addressing the potential for leakage of saltwater from the Savannah River/Atlantic Ocean to the underlying Upper Floridan aquifer as a result of proposed dredging. I also suggest that we re-read the proposal submitted by the City of Savannah and posted by Ben Brewton for the SEG (see web address below), as the later postings are either not related to the problem, and/or could have been answered by reading the proposal. It does look like Golder read the the proposal; their "summary of opinions" is somewhat of a repackaging of it. The path to the proposal follows:
http://www.sysconn.com/harbor/SEG/General%20Info/Hydrovision%20proposal%20to%20City%20of%20Savannah.htm
There are two principal issues that need clarification--modeling, and slug tests and their alternatives.
Please note that the following addresses postings in this forum; it does not address the people posting them. I refer to the postings by name, only because that is the way they are posted. Basically I am addressing topics, issues, principles, not people, positions, personalities.
Regarding the use of models to predict the leakage of saltwater from the Savannah River/Atlantic Ocean to the Upper Floridan aquifer as a result of increased dredging, as suggested in the HydroVision scope--
First, to address Chuck Watson's posting of 11/7/00: All five of Chuck Watson's statements are valid (his point 1 is more of a preference or tenet, than a component of a modeling method or approach--see my points in addressing Bill McLemore below for more comment on this). Chuck Watson's points 2-5 are basically standard operating procedures and are de rigueur for any modeling activity. All these procedures are considered in the modeling suggested by HydroVision. Because of the general scope of the proposal as submitted, details such as those were not included in the written proposal, much in the same way that details for the other tasks were not included. Basically, there are no differences in methods and approach between Chuck Watson's detailed ones and those proposed by HydroVision.
Second, to address Bill McLemore's posting of 11/8/00: Bill McLemore's statement: "...existing models are not appropriate for assessing the salt-water intrusion of the Upper Floridan aquifer in the Savannah area.", means that the existing, documented and published models that simulate ground-water flow can't be used to address saltwater intrusion. Of course this is true; they are regional models developed to address ground-water flow in the coastal area of Georgia and South Carolina. (To offer credentials as did Chuck Watson: this happens to be part of "my field", as I developed nearly all those models.) Those models are not solute transport models, and they were not designed to address saltwater intrusion, encroachment, or leakage, and certainly weren't developed to address the problem at hand, in any stretch of appropriate use. As such, Bill's statement and the one that follows regarding "existing models and variations thereof no longer are acceptable in the geological and engineering community for assessing salt-water intrusion in the Savannah area" has no bearing on the situation at hand, which would be to develop a saltwater transport model to specifically address the leakage of saltwater from the Savannah River to the Upper Floridan aquifer. Bill McLemore's statements about the proposed use of SUTRA 3-D by the USGS, its being "under development", and its "regional nature not specific to the issue of the Savannah Harbor deepening" is absolutely true. However, because of that, that has no bearing at all on the problem at hand. HydroVision does not propose the use of a regional, undocumented, unapproved, three-dimensional form of any model. HydroVision proposes to use a model capable of simulating saturated-unsaturated, density-dependent, ground-water flow and transport in two dimensions. The model of choice is SUTRA, although there are several other model codes that are capable of addressing this problem. The simulation geometry proposed by HydroVision is two dimensional, not three. Two dimensional flow, vertical leakage, would be simulated along the river reach, and if solute (saltwater) is present in the aquifer, two-dimensional lateral flow and transport would be used to simulate it. The two-dimensional version of SUTRA, the original, is approved, documented, and published. In addition, I have discussed the use of SUTRA with other USGS experts, including the author of the SUTRA code, Cliff Voss, as well as others who have used it and other models, such as Dr. Ming Kuo-Lee, Auburn University. They agree that this would be an appropriate use of SUTRA. Referring back to Chuck Watson; note that this is not a "wundermodel", but a powerful, albeit sophisticated, code that can easily be used "off the shelf".
Finally, to address Chuck Watson's posting of 11/9/00 on modeling issues: As was the case for Chuck's first posting, most of this discussion relates to standard operating procedures, methods, and approaches to modeling. These topics, such as data input importance, acceptance and validation, sensitivity, prediction, and the like, are basic and common truisms in modeling, and are well taken and important, and would be addressed in the modeling procedure. They are either stated in the proposal or inferred in it and not specifically stated owing to the proposal's limited scope and scale as mentioned above. Chuck's reference to the "problem...[of] a formal acceptance and evaluation process" should be addressed in the preceding paragraph and a reading of the proposal. Again, as stated above, although SUTRA is the model of choice among those who have written and used model codes, it is not the only code that could be used.
Slug Tests, Aquifer Tests, Laboratory Tests:
Regarding the use of slug tests compared to other means of in situ hydraulic testing as suggested in the HydroVision scope--
As I stated in the September 22 meeting in response to the suggestion that slug tests be used: Slug tests are inappropriate for the problem at hand because 1) the mathematics and theory of the application are based on homogeneous and isotropic media having primary porosity and lateral flow. (I believe we all can agree that the media is none of the above.) 2) Results of a slug test will be horizontal hydraulic conductivity of a lateral water-bearing zone, not vertical hydraulic conductivity of an aquitard. Let's remember that leakage of water from the river to the aquifer will be a vertical process, not a lateral flow one. We are seeking information and data on vertical leakage, not lateral flow. 3) Even laterally, a slug test evaluates an extremely small lateral extent of the "aquifer"--In contrast, an in situ hydraulic test suggested in the proposal tests a large volume of strata, which would include those parts of the strata having secondary porosity/permeability (fractures), and which results in vertical hydraulic properties for a large volume of strata. Even if one were to violate the assumptions of homogeneity and isotropy, the testing of vertical properties (2), of a large volume of strata (3), is critically important. Slug tests don't address that.
Jim Reichard's posting, points 1 and 2 posted 11/8/00, are right on the money. His point 1 basically states that in situ tests, not lab-derived properties, are needed to test enough of the strata to capture features enhancing secondary permeability. His point 2 explains the appropriateness and lack thereof of slug tests, as stated above. (Jim, correct me if I'm wrong.)
Although Bill McLemore's posting "To Fred Rich" of 11/8/00 correctly states that "Hydraulic conductivity measurements should be designed to address the conditions at hand", his comments that follow do not address the conditions at hand, and instead imply that the use of slug tests for this problem would be appropriate. They are not, as stated above. I too, suggest a reading of Butler's book on slug tests. I believe it would be easy to conclude that slug tests are totally inappropriate for addressing this problem, as stated above. Yes, slug test have great importance and utility in some applications, just not this one.
Note that the proposal suggests in situ testing of the strata only after certain conditions in simulation results are met or not met. (Please read the proposal). If in situ testing is determined necessary, then it should be conducted as proposed, and not using slug tests. Governing assumptions would be met, and not violated as they would be in applying a slug test. Aquifer testing as proposed would result in estimates of vertical and horizontal hydraulic conductivity of the "aquifer"; specific yield, storage, and effective porosity of the "aquifer"; and vertical hydraulic conductivity and storage of the "aquitard". I understand that "aquifers" and "aquitards" are not clearly defined, but their comparative presence can be estimated from results of work done and published by the Corps in their March 1998 report. This information would be used to select well and piezometer placement. Aquifer tests are comparatively expensive, but they may not be necessary, as stated above. Slug tests also are expensive, and would result in a "boon-doggle" that we want to avoid.
As stated at the beginning, the posting from Golder Associates is somewhat of a re-packaging of the HydroVision proposal. I see very little in the five main tasks described in the Golder posting that is not covered in the HydroVision proposal. The "Key Questions" and "Recommended Approach" are the same. There are some important differences in the details, however, and for completeness, comments on their 5 points follow:
1. Pore Water Sampling--This is identical to that proposed by HydroVision.
2. Fractures--HydroVision did not propose horizontal drilling. "Describing" fractures, if found, won't determine "their influence on the hydraulic characteristics of the sediments". In situ aquifer testing would be needed for that, if scope and funding allowed. Even if horizontal drilling and coring captured a fracture, it would only be about 4 inches in diameter (in height), and still not be representative of the strata above and below it. The fracture could extend; it could terminate; it could be transmissive; it may not be. The in situ strata would still not be tested. One would be chasing fractures horizontally, in 360 degrees, instead of vertically. Also, couldn't horizontal drilling be quite expensive and end up as an expensive "boon-doggle"?
3. Vertical Permeability--Bulk hydraulic conductivity could not be determined from "representative array of lab permeability tests". A comparatively poor estimate, on the slow end of Kv, could be had using lab results. If that's all the scope and funding were to allow, then that would have to do. It is true, as Bill McLemore's posting of 11/9/00 "To: Jim Reichard" states, that lab permeabilities were used by USGS in development of their numeric models. It is important to note that the scope of those studies did not include collection of additional data, so we used what we had. We did not use the data without thought, however; instead, they were used as slow ends of a range of possible field values, and recognized as not representing the field, especially regionally.
The statement regarding difficulty of stressing the prolific Floridan sufficient to elicit a measurable response in the strata overlying it is well taken. If the aquifer test were conducted, a large volume of water would have to be removed from the Floridan.
The HydroVision proposal takes into account the use of Kv from the laboratory analyses already taken. Those data would be used in simulation as described in the proposal. I think we all agree that collecting additional core is not cost effective, especially considering their inherent utility.
4. Historic Drawdown and Salt Water Leakage--This task is proposed as part of Tasks II and III in the proposal. This task is important when considering model calibration, acceptance, and verification, and would be subject to sensitivity analyses and assignment of confidence of results.
5. Simplistic Predictive Modeling or Calculations--This task also is proposed, chiefly in Task III. However, Golder's alternative method of calculating leakage using Darcy's Law won't work. All assumptions in the equation are violated. Simple modeling would consider all assumptions and not violate them. "Gaining additional directly observed data", as being deemed important in Golder's posting is important, and is in the proposal of HydroVision's.
Rick Krause - rick.krause@hydrovision.com
From Chris Hemingway representing Golder Associates 11/16/00:
As pointed out, Golder’s opinion of the necessary approach is similar to HydroVision’s proposal, as well as Camille Ransom’s suggestions. In general, we all seem to be in agreement with the general approach and the sequence of tasks, but there is still some concern about the specific methods .
1) Preliminary Modeling
Golder believes that preliminary modeling performed after a pore-water/coring investigation as HydroVision has suggested can be useful for:
* Estimating the rate and magnitude of past leakage if the pore water analysis indicates substantial salinity in the Miocene pore water; and
* Quantifying the effort required for a pumping test to determine if it is a feasible alternative.
However, keep in mind that the key datum for these simulations is a Kv value, and relatively small changes in this value could result in large differences in predicted travel rates. At this point, the range of Kv determined by the Corps study would be the only useable values, unless additional lab testing was performed along with the pore water analysis.
Also, we believe that although Darcy’s Law and equations based on Darcy’s Law do not account for multi-phase (density) flow, these equations can be applied to approximate vertical leakage through the confining layer.
2) Horizontal Drilling
We agree that horizontal drilling would be expensive; however, a well planned investigation (i.e. two or three transects perpendicular and parallel to the river) will intercept fractures if they are present. Based on Law Engineering’s investigation of fractures in coastal sediments, we interpret that fractures observed in the Miocene sediments at the outcrop area near Statesboro were "numerous" and other coastal plain sediments investigated have a maximum spacing of approximately 1 meter.
We agree that the transmissive properties of the fractures would not be easy to evaluate based on such a small sample (except in perhaps a relative comparison to non-fractured sediments). However, we do believe this investigation would put to rest the persistent question of whether or not fractures are present in the Miocene sediments underlying the river. Considering these points we ask the other committee members two questions:
* Does anyone strongly believe that direct observation of Miocene sediments obtained through horizontal drilling would not provide useful and essential information on fracture occurrence and distribution?
* Is it obvious that the additional information gained would not be justified in comparison to the cost of obtaining it?
3) Pumping Test
While we agree that laboratory derived permeability data is not the best way to evaluate vertical permeability on a large scale, we need to determine if a pumping test is a practical option. An evaluation of the volume of water removed and the approximate time that pumping must occur to elicit response in observation wells within or above the confining strata could be modeled relatively easily as a preliminary exercise. Again, one of the parameters controlling the modeled response of the system is a Kv value, which is what we are searching for in the first place. The suggested pumping test design, analytical approach, and preliminary modeling input
Chris Hemingway representing Golder Associates - chris_hemingway@golder.com
From Chuck Watson on 11/16/00:
In considering how to proceed from here, I think Chris' comments of 10 Nov should be highlighted. As much as those of us on the research side of the house like doing projects simply to learn more neat stuff, we need to keep in mind the end goal of answering the question, Will the deepening of the harbor adversely impact the aquifer? And the corollary: can a study be constructed which will reasonably answer that question?
A key question to me is if the uncertainty in the modeling process is such that the end answer will be meaningful from a policy standpoint. As I understand the Hydrovision proposal, the plan is to conduct modeling and sensitivity tests using existing data plus a vertical profile of salinity at selected critical locations, then proceed to a pump test if the simulation indicates it is warranted. One question for Rick Krause is, wouldn't it make sense to do that test in any event, and include the data in further model simulations if different (also as a confidence check - simulate the pumping, then see if the simulated and actual results matched)? I am assuming in the comments here and which follow that our present data on the spatial variability of the permeability of the confining unit is relatively sparse, especially with respect to meter scale features - the "fractures". The existing data doesn't tell us that - the pump test might.
My concern is that given the uncertain knowledge of the confining unit, even with the existing data and pump test there will remain considerable uncertainty on the impact of the dredging. Hypothetical situation: assume there is a thin layer at -48ft that caps a paleochannel of permeable sediment. The most vigorous pump test would fail to disclose this, as would modeling based on existing data or reasonable extrapolations thereof. The project proceeds, the layer is broken, and we have a problem. While that may not be the best example (geologists - is such a scenario even possible?), the point is that we may not have narrowed the uncertainty all that much at the end of the exercise than it is now. If such a scenario is possible (or similar, where there is a failure mode which may be undetected by the study, but could plausibly exist), we may in the end be left with a study which indicates that nothing was found, but we haven't really assessed the potential risk from the project.
So my questions to the hydrogeologists are: 1) What is the existing uncertainty in the field data? 2) What impact does that have on the modeling? 3) How will the proposed field work reduce that uncertainty? 4) Will the final uncertainty be such that a decision will be substantially improved due to a study? 5) How will the study help assess or detect potential realistic worse and bad case scenarios? Is the risk function smooth, or discontinuous due to failure modes? 6) Would it be possible to construct and run a worse case simulation early, perhaps as an initial study, to set an upper bound? If that showed minimal impact, then we are worried about nothing. Of course, if it caused an artesian salt water geyser in Macon, we really haven't learned much other than how bad things could get (and maybe raised a question or two about the modeling!).
I am not suggesting here that there is not a need for further study - just that we think through what we will do with the answers. Although you can't prove a negative, we certainly need to greatly reduce the uncertainty associated with the premise that the project may harm the aquifer. If the study is unable to do that it would raise a serious policy dilemma.
Chuck Watson - cwatson@methaz.com
From Bill McLemore on 11/17/00:
We need to keep in mind that there are no data that indicate that fractures influence ground-water flow in the Savannah River area--this is merely a concept that has been raised by sincere persons, whom I respect--the concept is not backed up by any measurements or data. There are data that contradict this idea; namely:
(1) Field observations by myself indicate that while some fractures are near vertical, many if not most have a non-near vertical dip component. Photographs in the published literature also illustrate this. (2) The Corps had very good core recovery overall with over 400 feet of 100% recovery. The Corps core descriptions are extremely good, if steeply fractures had been present, the Corps would have seen and recorded them. Only two steeply dipping fractures were encountered. If a near vertical fracture set had an 85 degree dip (which is quite steep) and a one meter spacing (the largest spacing in the published literature), then a core hole would intercept one of these fractures about every 40 feet--this means that the Corps should have encountered a steeply dipping fracture every 40 feet or have encountered many more than they actually did. The fact that they only encountered 2 fractures strongly indicates that steeply dipping fractures or fractures at somewhat lower angles are not present. (3) A field analysis was performed by Law on fractures and their hydrologic significance. Law concluded, in a written report, that the fractures were not a significant pathway for ground-water movement. One of the Law scientist (Dr. Joe Wilson) is an expert in the impact of fractures on ground-water movement and developed considerable experience on this matter as part of the siting studies done for the Low Level Radioactive Waste Repository where fracture permeability was a major issue. The other Law scientist (Dr. Jim Wallace) assisted the Upper Floridan Technical Advisory Committee in the development of Sound Science Initiative and has considerable experience in coastal Georgia ground water issues. (4) Dr. Robert Carver, retired professor of geology at UGA, reviewed the core data as well as the only published report dealing with fractures and concluded, in a written report, that there was no evidence that fractures were at the site and that regional ground-water data did not support the concept that the Charleston Earthquake created fractures that affected ground water. Dr. Carver initiated hydrogeological studies at UGA in the 1960s and has worked on coastal Georgia ground-water issues in the Savannah area in the past. (5) The issue of contaminant migration was assessed at the LCP Superfund site in Brunswick. One objective of this investigation was whether there could be migration of pollutants through the Miocene into the Upper Floridan. This was found not to be the case and pollutants "pooled" on top of the Miocene as it is an impermeable barrier. Continuous coring at the LCP site did not indicate fractures; the logs of these core holes are very detailed. One of the individuals directing this work was the former chief geologist for the USGS's Georgia office. (6) The PSAR and FSAR work at Plant Hatch did not reveal any evidence of fractures even though a special set of continuous corings were constructed to look for faults (a type of fracture). At Hatch, the primary structures are in the Miocene. Nuclear power plant investigations are extremely detailed and features such as fractures that can have an adverse impact on plant safety are addressed in detail. (7) Based on information in the Corps report, lithostatic pressures can be estimated. In the Savannah harbor area, lithostatic pressure will exceed hydrostatic pressure by about 15 pounds per square foot per foot. This means that there will be a considerable downward force (load) forcing fractures together (i.e., healing). (8) Blow count data from Corps drilling and soil descriptions and soil property information in the Corps report indicate that the Miocene materials have some plasticity and fractures should be tight (i.e., closed).
Every PE or PG that has looked at the issue of Miocene fractures as a ground-water pathway has dismissed such a pathway as being significant. There are written reports by these PEs and PGs stating such. No PE or PG has contradicted this position to date. There are no written reports by anyone that indicate that Miocene fractures are significant pathways for ground-water movement. The work of PEs and PGs is more credible than the work on non-PEs or non-PGs. [Note: the rendering of professional ground-water opinions is restricted by law to PEs and PGs).
Contrary to what has been said at a previous SEG meeting, absence of evidence is evidence of absence--it is not proof of absence; but it is very definitely evidence.
The bottom line is that the concept of Miocene fractures are significant pathways for ground-water movement in the Savannah area can not be demonstration with scientific facts. On the other hand, the concept that Miocene fractures are not significant pathways can be demonstrated with scientific facts.
Investigation of fractures as part of any recommended scope of work is inappropriate; and the Aquifer Committee should not include them in any scope of work.
Bill McLemore - Bill_McLemore@mail.dnr.state.ga.us
From Jim Reichard on 11/17/00:
I would like to voice my support for the points made by Rick Krause in his detailed post. As a hydrogeologist, I can attest that Rick's assessments are based on sound hydrogeologic principles. Moreover, almost all of what he describes is really quite basic and should be familiar to those well trained in the field. Based on what I have read of his past work on the Floridan aquifer and the comments in his recent post, I'd venture to say there is no one in the region more knowledgeable than Rick about the science of hydrogeology, including the modeling aspect. Therefore, I strongly recommend we use the Hydrovision proposal as the basis for investigating the potential of saltwater migration through the Miocene aquitard. Keeping in mind, of course, that the results from such an investigation will be used to strengthen the USACE 1998 report, which serves as the baseline document.
Jim Reichard - jreich@gasou.edu
Letter from Bill McLemore of 11/8/00:
"This letter by Bill McLemore is addressed to David Schaller of GPA who had consulted under statutory responsibilities with the Georgia Geological Survey about the concerns raised for potential contamination of the Upper Floridan Aquifer. Dr. McLemore reviewed the 1998 USACE report and this letter are his recommendations."
Chris Schuberth, Chair, Aquifer Committee
The letter was received only in hard copy. In the interest of accuracy, it has been scanned into Acrobat format. Please click here to link to the letter.
Letter from Harry Jue of 11/28/00:
I have read with guarded interest Dr. McLemore's review of the 1998 USACE report and recommended additional studies to assess the potential impact of harbor deepening on the Upper Floridan Aquifer. I have stated from the beginning that Savannah will support the harbor deepening project if the GPA demonstrates that there will be no harmful effects on the water source supplies, water quality and structural integrity of established infrastructure. As a member of the SEG and Harbor Committee, Savannah has participated in numerous discussions, critiqued numerous reports and recommendations by the SEG to assess the impacts of the deepening. Thousands of study dollars are committed to studying striped bass to sand at Tybee Island. I agree that these are important issues that require further study. The SEG has recommended that these studies be included and GPA has agreed to fund these studies as the local sponsor. However, I am puzzled over the resentment to the recommended study that Savannah has presented to the Aquifer Committee. We are not talking about a large sum of money. Our estimate is $200,000 or less. From the beginning I made it clear that this is not a personal issue; however to date one can read the replies which are defensive and offensive in nature. Although cloaked in a guise as a scientific report, there is no doubt that a concerted effort is being made to discredit Savannah recommendations for additional studies. Dr. McLemore clearly made a statement at an Aquifer Committee meeting in October. He stated, EPD does not support the Hydrovision report. He further states EPD will not use private consultants to justify or critique EPD policy. We are not talking about EPD policy in this matter. We are talking about protecting the water quality and long term viability of this region's resource. I do not find it appropriate to further critique Dr. McLemore's critique of the Hydrovision recommendations. I can only find two sentences on page three of this five page review where Dr. McLemore recommends additional work. He further states that if harbor deepening occurs that a program of long term groundwater monitoring would be appropriate. It seems to me that this is totally unnecessary and contradicts an earlier statement which indicated that the proposed deepening would not adversely impact the quality of water in the Upper Floridan Aquifer based on the considerable amount of data. The logical solution is to prevent the event from happening. He also further states that if new data demonstrates that public water supplies would be significantly adversely affected, then deepening would not be appropriate. I assume that this conclusion will be reached with the two sentence recommended study on page three.
Both members of the SEG and Aquifer Committee are aware that Hydrovision was retained by the City of Savannah to assist in critiquing the USACOE report. You are also aware that Hydrovision has requested that their proposal and scope of work be withdrawn from consideration by the SEG and Aquifer Committee. For obvious reasons as outlined in their letter dated November 28, 2000, Hydrovision has withdrawn due to both professional and economic reasons. This is highly unfortunate since true science depends upon open eyes. In this case several governmental authorities have chosen to close their eyes which has resulted in more personal attacks than constructive criticism. I suggest as a path forward that Dr. McLemore should revisit his letter of November 8,2000 and clearly delineate what additional studies are necessary and recommended.
Harry Jue - hjue@ci.savannah.ga.us
Hydrovision's Withdrawal of Proposal of Scope of Work, 11/29/00:
November 28, 2000
Prof. Christopher J. Schuberth
Professor of Geology and Science Education
Armstrong Atlantic State University
11935 Abercorn Street
Savannah, GA 31419-1997
Dear Chris:
HydroVision regretfully requests that you withdraw from consideration by the Aquifer Committee the proposal: “An Assessment of the Potential Effects of Savannah Harbor Expansion on the Quality of Water in the Upper Floridan Aquifer” that we prepared and submitted to the City of Savannah; and thence the City submitted to the Aquifer Committee. We have discussed this with Mr. Harry Jue of the City, and he concurs with our request, albeit with similar regret. Although we believe that the proposed work, either all or in part, could answer the questions posed by the City, the Aquifer Committee, the SEG, and all those having an interest in the subject, we have no alternative but to withdraw the proposal. In addition, we ask for reasons of professional ethics that the Aquifer Committee not use any part of this proposal in any future work proposed by the Committee. Before proceeding, please understand that this action is in no way a reflection on, or result of, your handling of the Committee.
By way of explanation, it appears obvious that for some reason, the State Geologist has done everything in his power to thwart the proposal and discredit its authors and supporters. This situation first surfaced at the September 22, 2000, Aquifer Committee meeting when Dr. McLemore stated that the proposal by HydroVision was unacceptable to the State of Georgia—unequivocal and seemingly incontestable—without supporting explanation. It continued with his postings in the Aquifer Experts Information Exchange in which he countered tasks proposed in the HydroVision proposal, with methods known by hydrogeologic experts to have no chance of success.
Dr. McLemore then shifted his emphasis from technical to non-technical issues, and made statements such as “The work of PEs and PGs is more credible than the work [of] non-PEs or non-PGs”, and his statement regarding the illegality of "rendering of professional opinions" by someone not having a PE or PG. He further references what he must consider to be true ground-water professionals of “Ph.D. level and Georgia-registered” status. No, Rick Krause has neither a PG nor a Ph.D., but I don’t believe you will find anyone who is more of an expert on coastal groundwater issues than Rick, including those referenced by Dr. McLemore. Rick has worked more than 30 years with the USGS on coastal groundwater issues, and has extensive knowledge and experience on the very issues with which we are dealing. He has developed or assisted with the development of nearly every groundwater model documented for use on the coast, several of which were done for the State of Georgia. More importantly, it was Rick’s work and his day to day participation on coastal groundwater issues with Dr. McLemore that was used by the State in development of Georgia EPD’s Interim Strategy to Protect Coastal Georgia from Saltwater Intrusion. Rick is professional, ethical, and honest. It seems however that Dr. McLemore implies that Rick no longer has any of these skills and attributes, is no longer “credible”, and has lost all knowledge of coastal hydrology since leaving the USGS. This unprofessional attack on HydroVision is not acceptable to me, and I will not subject my employees to such an atmosphere.
Finally, Dr. McLemore's letter to Mr. Schaller states that although he sees nothing wrong with the Corps of Engineers 1998 report and that no work needs be done, he offers Georgia EPD, together with South Carolina DHEC and the USGS, to “provide technical guidance…in developing a technically sound work plan for additional studies”. In the letter, Dr. McLemore also makes comments on what should and should not be addressed. It appears that Georgia EPD proposes to develop a new scope of work, based in part on Dr. McLemore’s letter to Dave Schaller. (Please note that we are not herein critiquing Dr. McLemore’s letter/review to Mr. Schaller.)
We are a small business and have incurred substantial expense on behalf of the Aquifer Committee and SEG. We can no longer afford, financially, to continue this process.
I regret that the above situation has driven me to make this decision, but I see no alternative. At this time, we do not plan to attend the December 1, 2000, Aquifer Committee meeting. If you believe, however, that we can still be of service to you outside these constraints, please feel free to contact me at 770 446 5445 ext. 103, or Rick, at ext. 109.
Sincerely,
Ram Arora, Ph.D., P.G.
President
Copy to: Mr. Harry Jue, Director, Water and Wastewater, City of Savannah
From Fred Rich written December 6, 2000 (posted 2/5/01):
After giving the matter some consideration, I decided it would be well to comment on some of the remarks that were made relative to fracture systems in Bill McLemore's comments. I am not doing this to prolong the debate, but simply to provide information that might be of value to others that view this site.
I have made more than 1600 direct field measurements of joints, small faults, and clastic dikes. The latest field site, at Rocky Ford, exemplifies the nature of fracture systems within the Miocene of the Inner Coastal Plain. Rocky Ford is on the Ogeechee River at the Jenkins-Bulloch County line. A silicified layer within the Coosawhatchie Fm. extends nearly across the river, and is extensively jointed. Among the 51 fractures I measured, 19 had dips of 86 degrees or more, with twelve having dips of 88-89 degrees. The remainder (32) had dips of 85 or less, with a range of 65-85. At a site on Sapelo Island, among 77 joints that we measured in humate-cemented sand, 32 had dips in excess of 85 degrees (19 were between 88-90), with the remainder lying between 68 and 85. These joints are illustrated in the paper by Bartholomew et al.(2000) in fieldtrip #3, SE-GSA. The joints may be spaced as little as 1 cm apart, or may be a meter apart; this is highly variable.
The field analysis performed by former Law geologist Joe Wilson was done as I guided him around Bulloch County. Two of the sites he visited are good ones to illustrate the fractures. Among 23 measured fractures at the two sites, 13 have dips of 86 degress or more, some have surface areas of a square meter, and a few are assembled in en echelon sets of 2, 3, or 6 with spacings of 5-20 cm.
The area near Plant Hatch, contrary to Bill McLemore's observations, is full of joints. This was made clear on a fieldtrip that I helped to lead for Gannett Fleming, Inc. in September 1999. Several sites, including the type section of the Cypesshead/Citronelle Fm. near Jesup, and Oglethorpe Bluff, Linden Bluff, and a roadcut near Twin Rivers Church were visited, and they are all extensively fractured. Early this fall I visited a Miocene vertebrate fossil locality on Game Management land on the banks of the Altamaha River in Wayne County. That site, and several others near it are just upstream of Plant Hatch. I saw the fossils (not previously reported from the Miocene) and enough joints to warrent returning to the site this winter. Jointed, silicified units within the Miocene define the shape of a small island in the river, and determine the morphology of the stream banks in several areas. The fact that the joints were not observed by the engineers that built the power plant is, perhaps, not unexpected in light of the generally poor knowledge we have of the geology of this region.
With reference to the PE's and PG's that may have familiarity with the joint systems in this area, it's safe to say that most of them are professionals from DuPont, Gannett Fleming, and the Savannah River Site that have been on fieldtrips in the last 14 months. None of those people have written any opinions that I am aware of because the subject was new to them all. Few PE's and PG's know anything about this issue, and none have done any work that amounts to more than an afternoon drive or a review of the five or six small bits of published information (much of it written by Bartholomew and Rich). Rather than relying on second-hand distillations of information that only a couple of us actually have in hand, it would be much better to gather more information. Hence the need for more research. We cannot afford to overlook,ignore, or minimize anything.
Respectfully,
Fred Rich
Fredrick J. Rich Department of Geology and Geography Georgia Southern University Statesboro, Georgia 30460-8149
Good day Chris,
The following remarks are presented by a hydrogeologist I know named John Seaberg. John is a graduate of the University of Minnesota hydrogeology program, and is a fellow I've known for a number of years. He has worked for the Twin Cities for several years in their groundwater modelling group, and I asked him if he would please look over the comments on the experts' page and render an opinion of his own. I am pleased to submit the following for inclusion on the second experts' page. I hope readers will view this as an objective appraisal that is offered by an experienced hydrogeologist who has no particular stake in this issue, and who has the credentials that Bill McClemore seems to think are essential.
Best wishes,
Fred
The following paragraphs comprise my comments regarding the discussion on the Aqufier Experts Information Exchange. I've read with interest the lively dialog on the "Aquifer Experts Information Exchange", and would like to offer my perspective on some of the issues.
In deciding the best methods by which to characterize the hydraulic properties of the Miocene sediments, it is important to consider the scale at which the system behavior is being evaluated. On 11/14/00, Rick Krause indicated a number of reasons why slug test data will not properly characterize the hydraulic characteristics of the Miocene sediments, not the least of which is that slug tests provide a measure of horizontal hydraulic characteristics rather than the vertical hydraulic conductivity, which is necessary for characterizing its leaky nature. Moreover, slug tests measure the properties on a scale so local that it could not be considered representative of the bulk aquifer properties over the scale of interest. The same can be said for laboratory test results on sediment cores. The application of spatial statistics, or geostatistics, would be required to ensure that the sampling distribution and density are adequate. I suspect that in all likelihood the number of samples required to provide a meaningful characterization of the medium would be prohibitive. Characterization and quantification of fractures from log data can be helpful, but will provide no data regarding how or if the nature, geometry, and distribution of the fractures affect vertical leakage. A properly designed and executed pumping test offers the only reliable means to evaluate the properties and behaviors of the hydrostratigraphic units on the scale of interest. It is possible that the Miocene sediments do indeed offer significant hydraulic resistance so that dredging activities will not impact the Floridan aquifer. If this is the case, it can be most effectively demonstrated through a proper pumping test.
I am pleased to see the use of simple models advocated because, as Chuck Watson stated on 11/7/00, they "can help shed light on sensitive aspects of the process." I recommend the application of simple models early on in the investigation, based on the available data. The use of a simple model early in an investigation can tip off researchers about the sensitivities of inputs. It also can be used to compare the relative responses of different scenarios, and to identify data gaps, thereby helping to guide the investigation so that limited resources are applied most efficiently. Such a model can be readily revised to include updated data as they become available, thereby improving our understanding of the flow system.
Chuck Watson's concern (11/16/00) that the impacts of dredging still will be uncertain, even if the hydraulic characteristics of the Miocene sediments are properly characterized, deserves further attention. Perhaps one approach would include direct measurement or estimation of the hydraulic resistance attributable to the bottom sediments of the river channel. Incorporating these data into a model would permit an evaluation of the role played by the river channel sediments in impeding flow. The model could be used to evaluate the system's sensitivity to the channel sediments and to test various scenarios under a range of hydraulic parameters assigned to them.
John K. Seaberg, PG, CPG, CGWP
John K. Seaberg
Metropolitan Groundwater Model
Minnesota Pollution Control Agency
Phone: 651.296.0550 Fax: 651.297.7709
520 Lafayette Road, St. Paul, MN 55155-4194
john.seaberg@pca.state.mn.us http://www.pca.state.mn.us/water/groundwater/metromodel.html