|
DATE: September 28, 2000 On September 12, 2000, the Minnehaha Creek
Watershed District received a report from Kelton Barr of Kelton
Barr Consulting, Inc. describing potential groundwater impacts from
the Highway 55/62 interchange and related construction. Mr. Barr concluded
that the proposed southwest loop pond might result in a permanent diversion
of one-third or more of the current base flow to Camp Coidwater Springs.
Due to the significant nature of Mr. Barr's findings, on September 15,
2000, the Watershed District hired Dr. Donald Segiel from Syracuse University
to review Kelton Barr's report. Dr. Segiel's report (attached) was received
by the MCWD on September 28, 2000. Dr. Segiel is nationally recognized authority
on groundwater; he has served as a expert witness to the US. Senate, a
Fellow of the Geological Society of America, a Birdsall Distinguished
Lecturer, and is currently serving on the U. S. Geological Survey Committee
on Hydrologic Research (Dr. Segiel's Curricula Vita is attached). Dr. Segiel concurs with Kelton Barr's, September 12, 2000 report. Based on his review of Mr. Barr's report and the SEH (Short, Elliot, Hendrickson) study, Dr. Segiel concludes that dewatering the unconsolidated sediments at the interchange area may seriously inhibit normal flow to Coldwater Camp Spring. Although, he agrees that Mr. Barr's estimate on flow reduction of 33 percent is plausible, he indicates that the SEH study was sufficiently flawed that a more accurate assessment of how much the spring volume may change is not possible with the data available. Dr. Segiel's report is attached. Copies of Mr. Segiel's report will be faxed
and mailed to the Lowtr Minnesota River Watershed District, the Department
of Natural Resources, the Department of Transportation and other interested
parties on September 29th 2000.
September 28, 2000
224 Stolp Avenue Mr. Eric Evenson, Administrator
Dear Mr. Evenson: This is a letter report presenting my initial review of Mr. Kelton Barr's assessment (Barr, 2000a) of the potential impact from the proposed Highway 55/62 exchange and related construction on groundwater discharge at Camp Coldwater Spring. To prepare my report, I largely relied on Mr. Barr's memorandum (Barr, 2000a), supporting documents cited within it, and Barr's summary report on the Minnehaha Creek Watershed District (Barr, 2000,b), I also draw upon my Own knowledge of the study area. I lived not far from Minnehaha Falls for about eight years and studied Twin Cities geology both while pursuing my doctorate in hydrogeology at the University of Minnesota, and when subsequently I was employed by the St. Paul District of the U.S. Geological Survey before I joined the Syracuse University faculty. Kelton Barr's conclusions
that the construction and subsequent dewatering will measurably change
the volume of flow at Camp Coldwater Spring certainly are sound. The most
salient evidence supporting Barr's concern is the obvious NE-SW trending
set of major fractures connecting the groundwater system in the Platteville
formation near Barr calculated from data on hydraulic head, aquifer
transmissivities (SEH, 2000), effective cross-sectional width in the construction
area, and spring discharge rates that about 20 to 30 percent of groundwater
flow to the spring might be lost by the construction and permanent dewatering
in the spring recharge area. These calculations certainty are plausible.
Barr also shows that about 10% of the water being pumped during pumping
tests near the interchange site plausibly is water that would have gone
to the spring. Together) the fractures, pumping tests results, and water
level measurements compellingly document a hydrologic connection between
However, there still is significant uncertainty regarding how much water will be lost, within the calculated range or more. To explain why there remains such uncertainty, I have to partly critique the SEH (2000) study, which forms the essential database used to understand the hydrogeologic conditions at the interchange site. My critique necessarily is general; I have not prepared original maps or re-interpreted data. The SEH (2000) report has much to offer. It is comprehensive in many respects, and includes the results of a wide range of geologic, geophysical, and hydrologic analyses to address the issue. However, there are conceptual problems and errors in analysis that compromise the conclusions of the report. Briefly: 1. Conceptualization: The argument (p.3) that “Locally groundwater appears perched upon the Hidden Falls submember [of the Platteville Formation] “is arbitrary and unsubstantiated in the text. Why would more clay content in the Hidden Falls member, compared to the other members, affect the vertical fractures that might penetrate through it? I see no reason to argue for a hydraulic no- flow boundary at the base of the unconsolidated deposits near the interchange and to the spring without clearly documenting that such an unsaturated zone occurs. Indeed, were the unconsolidated deposits perched on the Platteville, then the potentiometric surface of the Platteville logically could be quite different that the water table configuration above it, particularly if the water table were leaking to the Platteville in discrete fracture zones. Yet SEH (2000) seems to have contoured all the water level data as if it were simply the watertable (Fig. 7), implying that there is excellent hydraulic connection between the unconsolidated materials and the platteville. I think SEH(2000) missed the opportunity to logically install nested water-table wells and piezometers where fractures actually occur to determine the extent to which recharge occurs through the fractures and towards the spring. 2 Aquifer tests: SEH(2000) misinterpreted both slug test and pumping test data. Many of the slug test data have a classic “double line” effect, indicating bore-hole problems and/or other complexities. In this case, the line fitted for interpretation is generally passed through the least steep segment of the plotted data, not the steepest. If the AQTESOLV program does not discriminate the double line effect, then why use it? I have not recalculated the hydraulic conductivity from the variable head tests using the appropriate line segments, but values for many of the tests would be smaller than determined from AQTESOLV. Also, slug test data could have been used to define where the best place would be along a fracture to put a pumping well (it should be located at the place of greatest hydraulic conductivity). Indeed, I find it striking that the pumping wells were not purposefully placed on fracture traces that might logically be connected to Camp Coldwater Spring. Would this not have been the best way to test whether dewatering might affect it Instead, aquifer tests were done in the more competent areas between the fractures where flow is inhibited. With respect to how the aquifer tests were done, I do not agree that hydraulic properties can be determined from changing water levels in the pumped wells by using equations designed for single monitoring wells adjacent to the pumping wells. Specific capacity data, in contrast, can be used for pumping wells to estimate transmissivity. SEH (2000) reports drawdown in monitoring wells (p.10-Il) but does not show recovery curves for these wells, which would have been more appropriate to obtain valid measurements of hydraulic properties. Indeed, there is no map showing the area of influence at all around the pumping wells; nor do I see data showing drawdown in the observation wells. Appendix H shows recovery
and drawdown data from the pumped wells, and many of the curve fits are
not good. This is not unsurprising. The unconsolidated aquifer is unconfined,
but most of the curves do not show the time when the effects of delayed
yield occur, after which drawdown or recovery data can be used to obtain
proper hydraulic properties. In short, the aquifer tests were not of sufficient
duration, even if one could use the pumping wells for the analysis as
if they were monitoring wells. Finally, the drawdown. and recovery observed
in the pumping wells were the combination of the drawdowns and recovery
of all three wells (P. 10.11). One cannot simply use the standard
equations appropriate for single monitoring wells located next to single
pumping wells for multiple pumping wells that hydraulically interfere
with one another; numerical analysis or image well analysis is needed.
3. These calculations are not an important part of the analysis, but I am curious as to why they were done (p.9). I have not seen such calculations done before in an engineering or hydrogeologic study of this kind. The amount of water removed from sediments in the interchange area Is not really important to the problem. What is important is how the dewatering affects the long-term delivery of water to the spring. 4. Briefly, it is clear in the SEH (2000) report that the groundwater model constructed to predict the effects of dewatering at the interchange could not be verified (experimentally shown to be correct). I also saw no indication of any sensitivity analysis to test the reliability of the initial calibration. Not including some sensitivity analysis is a major omission and atypical of what is now required of groundwater modeling. Because the model could not be validated, either the model design was wrong or the data put into it were wrong. Other questions on model design may be posed, but since the model in any event could not be verified, I will not raise them. In conclusion, I agree
with Kelton Barr (2000,a) that normal flow to Coldwater Camp Spring may
be seriously inhibited by dewatering the unconsolidated sediments at the
interchange area. I also believe that the SEH (2000) study was sufficiently
flawed that a more rigorous analysis of the potential impact on spring
volume is not possible from the data available at this time. Sincerly, Donald I. Siegel, PhD References Cited: Barr, K., 2000b, Minnehaha
Creek Watershed District Bluff Area Summary Report, Barr, K., 2000a, Memorandum:
Potential Impact of Construction and Dewatering at the Short Elliot Hendrickson,
(2000),, TH 55162 interchange: Hydrogeologic Evaluation, |