This test case was simulated using the described

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: st are listed in table 3. All other parameters are identical to the base case for the Reilly problem. The transient simulation was run for a stress period of 150 days, using 12 time steps and a time-step multiplier of 1.3. The initial heads were defined by the solution generated in a preliminary steady-state stress period in which the multi-node well was assigned a net discharge of zero. This test case was simulated using the described correction for the presence of a seepage face and without the correction (assuming that driving force for flow between the aquifer and the well is always the head difference between the hn and Table 3. Modified parameters in Reilly problem for evaluating seepage face calculation in multi-node pumping well. [ft/d, feet per day; ft, feet; ft3/d, cubic feet per day] Parameter Specific storage (all layers) Specific yield (layer 1) KSKIN Q Value 3×10-4 ft-1 0.30 12.5 ft/d -10,000 ft3/d hWELL). As seen in figure 20, hWELL did not drop below the bottom elevation of the cell associated with the top node of the well (-10.0 ft) until the ninth time step (at about 65 days). (Note that this figure shows changes during the transient stress period, and not for the preliminary steady-state stress period.) From then on, there was a small difference in the computed value of hWELL between the uncorrected and corrected cases, with the head in the well being somewhat lower when the correction is made. During this time, the head in the top cell connected to the well (located in model layer 2) remained above the bottom elevation of the cell through the 11th time step (at about 114 days) while the computed water level in the well was below the bottom elevation of that cell—creating the conditions for a seepage face to develop. During the 12th and final time step, the head in this cell dropped below the bottom elevation of the cell (-10.0 ft), thereby disconnecting it from the well. Also during the final time step, hWELL dropped below the bottom elevation of the cell in layer 3 (-15.0 ft), thereby creating a seepage face condition in that cell, which contains the second node of the multi-node well. If hWELL drops below several layers of active nodes, then the model will allow the development of multiple seepage faces and compute adjusted inflows accordingly. The flow into the top node of the well was more substantially affected by the implementation of the correction (fig. 21) than were the heads. The inflow from the aquifer into node 1 of the well (located in model layer 2) decreased slightly over time Figure 19. Schematic cross section of A, an unconfined aquifer showing a multi-node well open to parts of the uppermost five model layers, and B, the relation of the screens (open intervals) to lower water levels at a later time (t1), when the water level in the well has fallen below the bottom elevation of model layer 1. Model Features and Processes 21 Figure 20. Computed changes in head in the multi-node pumping well (hWELL) and in the aquifer cell connected to the uppermost node of the multinode well (hn) for the modified Reilly problem, both with and without a correction for the development of a seepage face in the uppermost cell. Top node of well located in model layer 2. Results of preliminary steadystate stress period represent initial conditions for the 150day transient stress period shown in this figure. through the eighth time step as the head in the aquifer and the water level in the well both declined. However, once the water level in the well dropped below the bottom elevation of the cell, the rate of decrease in the inflow was much greater with the correction on, which resulted in a lower flow rate to the well compared with the simulation without the correction. Because the net discharge from the well is specified, as the inflow decreases in node 1 there are compensating increases in inflow at the other nodes of the multi-node well (fig. 22). Once the cell in model layer 2 goes dry (or when the head in the aquifer falls below the bottom elevation of the cell if the cell is not convertible, as in this example), the inflow to the well from that cell ceases. In this example, this occurs during the last time step (at 150 days, as seen in fig. 21). Overall, the correction for the development of a seepage face, as described above, yields changes that are logically consistent with expectations based on well hydraulics. Although this example problem shows only small effects of the seepage face, in other situations, the effects can be much greater. The calculations related to a seepage face are performed automatically by MNW2 whenever it detects the presence of conditions for which a seepage face is expected to occur, as described above. The model user does not have an option to control this feature. Constraints on Pumping Rate The range over which the water level in a well can potentially change may not be unbounded. For example, in discharging wells, pumping cannot continue if the water level drops below the depth of the pump settings and screen intakes. In recharging wells, the water lev...
View Full Document

This document was uploaded on 01/20/2014.

Ask a homework question - tutors are online