Laserfiche WebLink
Mr. Peter MacNicholl <br /> 6/12/2007 <br /> Page 6 <br /> 3. A model run using a conservative ion such as boron or chloride as a non <br /> retarded species and compared with particle tracks might be helpful in assessing <br /> the relative effects of hydraulic conductivity and retardation for the organic <br /> compounds. <br /> 4. Time-history plots of predicted versus observed VOC concentrations at a few key <br /> wells might also be helpful in assessing calibration. Likewise, model results <br /> shown in cross section located near sections A-A' and B-B' would be helpful. <br /> Section 6. Conclusions <br /> 1. Calibration needs to simultaneously address all aspects of Darcy's law, not just <br /> matching heads: the overall water balance must be correct, the gradients and <br /> flow directions, vertical and horizontal, must be within limits, and the permeability <br /> must be within reasonable range of observations. Matching of heads does not <br /> eliminate problems in overall water balance, or water locally moving in the wrong <br /> direction, or errors in the calibration dataset due to differences in pumping <br /> patterns from year to year. A short discussion of the other indicators of <br /> calibration should be added. <br /> Section 7.Conclusions and Recommendations: <br /> 1. The transport model does reproduce water levels to within a few percent, but the <br /> solution used is not unique, because most of the calibration was to values of <br /> head alone, and not to overall water balance. Improving estimates of the water <br /> balance, particularly areal recharge, subsurface inflow and outflow, would further <br /> constrain the solution and reduce uncertainty. <br /> 2. Because recharge has been shown to be a particularly important parameter, <br /> GSU agrees that more effort to quantify recharge is appropriate. An analysis of <br /> the effect of delaying recharge into the following year is requested. Additional <br /> refinement is also recommended during future use of the model. <br /> 3. Assumptions regarding other boundary conditions also need continuous <br /> reevaluation. The assumption of no flow across the model's bottom boundary is <br /> open to question, as are assumptions about flow rates across general head <br /> boundaries. <br />