Field Tests - Rock Mechanics Based on an Anisotropic Jointed Rock Model - page 506

By means of specifi c plots, certain fl ow models can be identifi ed. Test results yielding linear
relationships between Δ h and ln(t) indicate radial fl ow. A linear relation between Δ h and

1/ h versus plot is linear, parallel fl ow conditions

exist (Hekel 1994). The most sensitive results can be obtained using double logarithmic
representations in which the derivative of pressure with respect to time is plotted versus
time. These are referred to as “diagnostic plots” (Bourdet et al. 1983).
The methods based on type curves and diagnostic plots are based on analytic solutions
of the transient fl ow equation, assuming a homogeneous isotropic rock mass with in-
fi nite extent or impermeable boundaries at some distance (Miehe et al. 1993, Hekel
1994). In more refi ned interpretation models, numerical methods such as the FEM are
applied in which arbitrary boundary conditions and also inhomogeneous and aniso-
tropic rock mass permeability can be accounted for (Stormont et al. 1991, Miehe et al.
1993, Wittke 1999, Wittke 2000a).
The interpretation of injection tests conducted with gas as test fl uid with respect to per-
meability requires evaluation methods that are based on the equation of transient gas
fl ow (Al-Hussainy & Ramy 1966). If the tests are carried out below the water table two-
phase fl ow of water and gas must also be accounted for (Wittke 1999, Wittke 2000a).
Figure 15.33 shows examples of the interpretation of pulse and constant pressure tests
by means of the FEM (Wittke 1999, Wittke 2000a). These tests were carried out in
horizontal boreholes drilled into the disturbed rock zone around a vertical shaft located
in rock salt at a depth of 340 m. By means of six packers fi ve test sections, with dis-
tances up to 120 cm from the wall of the shaft, were sealed. Each test section was tested
consecutively with brine as test fl uid. The distribution of permeability as a function of
the distance from the shaft wall shown in Fig. 15.33 (upper) yielded good agreements
between test and analysis results (Fig. 15.33, center and lower). The initial phase of the
constant pressure tests was infl uenced by the testing equipment and thus was not used
for the test evaluation (Wittke 1999, Wittke 2000a).

reveals spherically fl ow. If a Δ

Slug test and falling head test
In a slug test the test section in the borehole is sealed by a double packer system. A
standpipe inserted into the borehole allows free passage of water into the test section.
Then the water level in the test section is abruptly changed. Subsequently, the recov-
ery of the water level in the standpipe versus time is recorded until it approaches the
value prior to test.
In a drawdown test, lowering of the water level in the test section is obtained by
pulling a displacer. In a pouring test, rise of the water level in the test section is ac-
complished by immersion of a displacer. Slug tests can be conducted only below the
groundwater table.
In Fig. 15.34 (left) the layout of a slug test in an unconfi ned aquifer is illustrated (Hekel
1994). Using a water supply system, the water level in the standpipe is raised to an excess
head Δ h 0 , and Δ h(t) then is recorded until it approaches zero (Fig. 15.34, upper right).