Agenda - Council Work Session - 10/20/2020 - Special

Home

5.1.8.1 A local model limited to an approximate radius around the city limits was extracted from the regional seven -county rnodel using telescopic mesh refinement with the Groundwater Vistas software. Constant and general head boundaries around the Limits of the model along with wells, rivers and lakes, and infiltration, provided the model boundary conditions. The model grid was refined around the City of Ramsey wells. Variable grid spacing was used, ranging from approximately 2 meters near the City of Ramsey wells to approximately 500 meters at the edge of the grid. Prior to their use in the delineations, the following modifications were incorporated in the refined models: • Local areas of modified horizontal conductivity were included in the model. • The pumping rates for baseline no pumping), maximum present day use, and projected 2040 demand were inputted into scenarios of the model. To determine the water contours of the aquifer and the resulting cone of depressions multiple model runs using multiple flow rates were inputted into the city wells. Baseline conditions were established creating a model that input no pumping from the City wells. This represents static aquifer water levels without influence of the City wells. Water elevations from this baseline condition is depicted on Figure 12 in Appendix D. The results from this model run are verified and match MnDNR hydrogeoiogic atlas potentionnetric surface predictions depicted in Figure 7 and Figure 9 in Appendix D. The second model run input pumping values from June 12, 2019 to predict the cone of depression caused from 8 hours of pumping 4 wells. The resulting head values from this model were subtracted from the baseline model. Results from this calculation are shown in Figure 13 in Appendix D. This Figure depicts the cone of depression created on June 12, 2019. To check accuracy of the model results Well 3 a non -pumping observation well saw water levels drop approximate 12-15 feet from baseline conditions, essentially matching modeled results. The third model run adjusted June 12, 2019 pumping wells to be increased to projected 2040 demand, The resulting cone of depression is depicted on Figure 14 in Appendix D. Results indicated almost double the drawdown depicted from the second model run. Well 3 was again input as a non -pumping observation well and would observe 40 feet of drawdown under these conditions. Model Calibration A qualitative evaluation of the calibration can be made by comparing the simulated potentiometric surface (Figure 12 in Appendix D) with observed water level targets obtained from the IV1WI database and Minnesota Department of Natural Resources Potentiornetric Surfaces (Figure 7 and Figure 9 in Appendix D). Upon review, the calibrated flow model generally captures the major features of the groundwater flow system along with the elevation, shape, magnitude, and gradient of the MWI database observed flow field, A quantitative measure by which to evaluate the success obtained during calibration is to compare the root mean square of the residuals (RIV1S) and the maximum observed head difference of the calibration dataset. The calibration dataset included water level information from wells in an approximate 16 mile radius of the cityls wells. The root mean square residual of the calibration for layers 6 and 7 for the model was approximately 5.15 meters with a Normalized FEASIBILITY STUDY -DRAFT RAMSY 154354 Page 22