116 N. Baptiste, R.P.
Chapuis / Engineering Geology 184 (2015) 111118
1.E+00
soil K3 value when K3 is higher than K1. The results place the asymptote
best, US, LAW
worst, US, LAW at about 50 K1, thus about 5 104 m/s for this screen and the limited
usual, US, AW clogging condition of case 1.
1.E-01 worst, US, AW If K1 is higher than K2 the obtained results (Figure 5) are also similar
usual, CS, AW
highly CS, AW to those of Fig. 4, but they now depend only on the K2 value: Ktest starts
measured K in m/s
1.E-02 to underestimate K3 when K3 is higher than K2, and the asymptote or
measuring capacity has a lower value than that of the case with the
same K1 value but without lter sand. Thus, the lter sand is a major
1.E-03 obstruction yielding the K3 underestimate.
When K2 is higher than K1, the inuence of K2 on Ktest is small but not
negligible. The head losses in the lter sand become negligible when K2
1.E-04 is about 30 times higher than K1. For lower K2 values, the Ktest value is
inuenced by both K1 and K2, the asymptote value being lower than
that obtained for the case without lter sand. The eld measuring ca-
1.E-05
pacity is in the range of 105 to 103 m/s.
1.E-06
5.2.3. Numerical examples of variable-head tests
1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00
A few variable-head tests were numerically simulated using the
K 3 (tested soil) in m/s code Seep/W (Geoslope, 2007). The screen had an internal radius r1i
of 2.5 cm and an external radius r1o of 3.0 cm, whereas r2i was
Fig. 4. Inuence of screen (K1) on the Ktest value when there is no lter pack: analytical re-
3.0 cm and r2o was 7.5 cm for the lter pack. A few combinations of
sults (Eq. (5)) and numerical results are similar; worst for K1 = 6 105 m/s; usual for
K1 = 104 m/s; best for K1 = 2 103 m/s; US = unclogged screen; CS = clogged screen; K1 (screen) with K2 (lter) and K3 (soil) were analysed. The screen
AW = aerated water; LAW = less aerated water. was 60 cm long, the lter pack was 80 cm long, and it was in the
middle of a 2 m thick aquifer. The zone inuenced by the slug test
unclogged K1 value, the eld measuring capacity is in the range of 105 was numerically found to be very limited, which conrmed the similar
to 5 103 m/s. ndings of Guyonnet et al. (1993). The models had an external bound-
ary at 3 m from the centre of the MW axis. The MW pipe was treated
5.2.2. Filter pack around the screen as a rigid and very pervious material in which the functions for volu-
The obtained Ktest depends up on K3 (tested soil), K2 (lter pack) and metric water content and unsaturated permeability were those of
K1 (screen). A few cases are used here to show by how much Ktest differs a pipe element (Chapuis, 2009a). Numerically, the slug test was
from K3 for constant-head tests (Figure 5). The screens 0.006d and modelled in three steps. The rst step gives the initial conditions, a
0.010d+ are used with lter sands SF50, SF00 and SF1. Five cases are constant hydraulic head within the aquifer. The second step is rapid
for unclogged screens, the K1 value being that given by the tests in the inlling (2 s) of the MW solid pipe. The third step is the slug test,
water tank (K1 in the range of 104 m/s, see Figure 3). To represent the- the water level in the solid pipe slowly returning to equilibrium. The
oretical eld clogging conditions, the K1 values for the clogged screens numerical data yielded straight velocity graphs (Chapuis, 2001;
were taken as 10 5 m/s, to investigate the eld range of reduced K1 Chapuis and Chenaf, 2002), which yielded perfectly straight semi-log
value by a factor of about 10. graphs (Hvorslev, 1951).
If K1 (screen) is smaller than any K2 (lter sand), then K2 has little Numerical results for slug tests with an unclogged screen (K1 = 2
inuence and K1 has the chief inuence on Ktest. The results (Figure 5) 104 m/s) and a clogged screen (K1 = 105 m/s) appear in Fig. 6.
are similar to those of Fig. 4: the Ktest value underestimates the tested They gave Ktest values which are similar to those predicted by Eq. (5)
1.E-01 1.E-02
US 0.006d- SF50 Eq. (5) - K1 = 2e-4 m/s
US 0.010d+ SF50 FEM - K1 = 2E-4 m/s
US 0.006d- SF00 Eq. (5) - K1 = 1E-5 m/s
1.E-02
US 0.010d+ SF00 1.E-03 FEM - K1 = 1E-5 m/s
equality
obtained K (test) in m/s
US 0.010d+ SF1
CS SF00
1.E-03
CS SF1
K test in m/s
1.E-04
1.E-04
1.E-05
1.E-05
1.E-06 1.E-06
1.E-07
1.E-07
1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01
1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02
tested soil K 3 in m/s
K 3 (soil) in m/s
Fig. 5. Inuence of K1 (screen) and K2 (lter sand) on Ktest for constant-head tests with
fully penetrating screens. US = unclogged screen, CS = clogged screen with K1 = Fig. 6. Inuence of K1 (screen) and K2 (lter sand) on Ktest for variable-head (slug) tests
105 m/s; 0.006d is the type of screen; SF50 the type of lter sand. and one case of partially penetrating screen (FEM = nite element method).