Laserfiche WebLink
Bernhardt Pit Groundwater Evaluation <br /> May 2023 Page 14 of 31 <br /> Model Results <br /> We conducted two model runs to evaluate groundwater level changes caused by the Bernhardt pit <br /> slurry wall. <br /> Pre-slurry Wall Run <br /> The initial run pre-slurry wall run (Bernhardt_SS7) represents the pre-slurry wall conditions. <br /> Figure A8 shows the modeled water table contours (blue) compared to "target" contours shown <br /> in Figure A3. Figure A8 also shows the locations of"Constant Head" cells which establish the <br /> overall model gradient and the model grid. In this run the Bernhadt pit cells are"on"which allows <br /> water to flow through them. The purple "dry" cells are where the modeled water level are below <br /> the bottom of the aquifer. This makes sense because there are not wells (yellow dots) within the <br /> dry cell area. <br /> Observed vs. Modeled Heads <br /> We compared the predevelopment observed well water levels and their corresponding modeled <br /> cell value in the figure below and the calibration statistics are included in the table to the right. <br /> 4760 v=0.9998x Residual Mean 0.765455 <br /> R'=0.6342 Absolute Residual Mean 4.948347 <br /> 4750 • • Residual Std.Deviation 6.68418 <br /> E Sum of Squares 3394.8138 <br /> 4740 • RMS Error 6.727866 <br /> •.:`f • Min.Residual -23.832459 <br /> m 4730 • • Max.Residual 17.682257 <br /> m • �• • Number of Observations 75 <br /> A �•go •• • <br /> F 4720 •••i�i • Range in Observations 47.2 <br /> • •M Scaled Residual Std.Deviation 0.141614 <br /> 34710 Scaled Absolute Residual Mean 0.104838 <br /> o • • Scaled RMS Error 0.14254 <br /> m •.- <br /> f 4700 Scaled Residual Mean 0.016217 <br /> 4690 _. . <br /> 4690 4700 4710 4720 4730 4740 4750 4760 <br /> Observed Water Table ElevAon(fLmsl) <br /> If there were perfect match, the data would plot on a straight line with a perfect 1:1 relationship. <br /> The best fit line shown above has a slope of 0.9998:1 which is a very close to perfect when <br /> considering all the data(no SEO well data was excluded). The Root Mean Squared(RMS)error <br /> (6.73 ft) is the square of the difference between the observed and modeled values (aka"residual"). <br /> We feel this value is reasonable considering the data used covers a 60-year period that includes <br /> fluctuations caused by well pumping, river fluctuations and agricultural return flows. We did not <br /> spend a considerable time trying to obtain a better match heads because the model results (Figure <br /> A9)are insensitive to aquifer parameters as discussed below in"Model Sensitivity" section. <br /> Residuals <br /> McGrane Water Engineering,LLC <br /> 1669 Apple Valley Rd. 0 Lyons,CO 80540 0 Phone:(303)917-1247;email:dennis@mcgranewater.com <br />