This preview shows page 1. Sign up to view the full content.
Unformatted text preview: imation, the application of the correction is optional, and the
user must specify whether or not the heads or water levels in a
multinode well should be corrected according to the algorithm described in equation 25. (The effects of partial penetration are included in MNW2 calculations if the user specifies
a value of input variable PPFLAG greater than 0 in input item
2b.) If the user indicates that partial penetration corrections
should be made, and the open intervals of a vertical well are
specified in terms of the elevations of the tops and bottoms
of one or more open (or screened) intervals in a long well,
then MNW2 will calculate the length of open interval associated with each node of the multinode well and automatically Figure 17. Plot showing sensitivity of total intraborehole flow (Q)
to variations in the penetration fraction for the uppermost node of
the multinode well in the Reilly problem. 19 calculate the penetration fraction at all nodes of the well. If
the user indicates that partial penetration corrections should be
made, but nodes are specified in terms of model layers, then
the user must also explicitly define the penetration fraction
associated with every node of that multinode well (using
input variable PP in item 2d2 of the MNW2 input). Seepage Face
In the MNW2 Package, it is normally assumed that flow
between the well and each cell associated with the well is
driven by the head difference between the well and the respective cell. In an unconfined aquifer, however, it is possible
that the water level in the well can be computed to lie some
distance below the water table. Chenaf and Chapuis (2007)
note that when a well is pumping from an unconfined aquifer,
“…the water table usually does not join the water level in
the well. There is a seepage face inside the well, which is a
key element in evaluating the well performance.” Thus, in a
numerical model, if the computed water level in a well in an
unconfined aquifer drops in elevation over time, then its position may fall into the model layer underlying the cell containing the water table (fig. 19). This creates a seepage face and a
disconnect in the hydraulic continuity between the saturated
zone in the cell and the water in the well.
When the computed water level in the well falls below the
bottom elevation of a cell, the normal assumption in the model
that the flow is driven by the hydraulic gradient—defined by
the difference between the watertable elevation and the water
level in the well—would be an overestimation of the driving
force because of the loss of continuous saturation between the
water in the upper cell and the water in the well. Under such
conditions, a seepage face can be surmised to occur in the upper
part of the well screen—at least in the cell containing the water
table. When such a seepage face is detected by the model,
flow in that cell from the saturated zone into the multinode
well is simulated as a generalhead boundary, and the flow is Figure 18. Plot showing flux
into or out of each node of
the borehole for penetration
fractions of the top node of the
well of 0.0, 0.5, and 1.0 in the
Reilly test problem. Negative
values represent flow from the
aquifer into the borehole and
positive values represent flow
out of the borehole and into the
aquifer. 20 Revised MultiNode Well (MNW2) Package for MODFLOW GroundWater Flow Model computed on the basis of the celltowell conductance (based
on the saturated thickness of the cell containing the water table)
and the hydraulic gradient as defined by the difference between
watertable elevation in that grid cell and the elevation of the
bottom of the cell. Figure 19 also shows that a small seepage
face develops in model layer 2 in that hypothetical example.
However, the model does not treat this seepage face any differently than normal because the cell is fully saturated and the
water level in the well is above the bottom elevation of the cell,
so the governing hydraulic gradient would be the same over the
entire thickness of the cell.
If both the water level in the well and the head in the grid
cell fall below the bottom elevation of the cell, then the model
will assume that hydraulic continuity is lost. Therefore, no
flow is allowed between the aquifer and that particular node of
the multinode well under those conditions.
The effect of computing the presence of a seepage face is
illustrated with an example based on the Reilly test problem.
To generate a seepage face condition, the Reilly problem as
described above was modified to convert it to a transient flow
problem with a pumping rate (Q = 10,000 ft3/day) sufficient
to generate drawdown in the multinode well so that the water
level in the well would decline to a level below the bottom elevation of the first cell connected to the multinode well. Additionally, the hydraulic conductivity of the skin was set equal
to onetenth of that in the base case (KSKIN = 12.5 ft/day). The
modified parameters for this te...
View
Full
Document
This document was uploaded on 01/20/2014.
 Winter '14

Click to edit the document details