PART 7 PERMEABILITY TESTS
7.1 Background
The tests for establishing the insitu horizontal permeability were carried
out with a self boring permeameter (PERM).
The PERM consists of a self boring pressuremeter whose expanding part is covered with a brass sleeve perforated by a regular arrangement of holes, in appearance much like a coarse colander. The diameter of the cutting shoe head and associated parts is matched to the diameter over the brass sleeve, so the probe is self bored into position with the same minimal disturbance arrangement as the conventional SBPM.
Once in position in the ground, water is pumped down into the space behind the inner wall of the brass sleeve, out through the holes and into the formation. The water is delivered at a constant rate of flow, and an equilibrium driving pressure is discovered for maintaining a constant flow throughout the soil. The pressure for a given flow is a function of the permeability and the geometry of the arrangement of parts.
The conventional solution for a permeameter test uses an equation of the following form derived from Darcy’s flow rule:
k = Q¥ /FH ...[Equ.51]
where
A practical version of Equation [51] is given by Hvorslev (quoted in Lambe & Whitman 1969) who offers a solution for the horizontal permeability from a well point-filter in a uniform soil, the model for the self boring permeameter test:
kh = [Q/2p LHc].Ln[(mL/D)+Ö (1+(mL/d)2)] ... [Equ.52]
where
Figure 24 shows the physical arrangement on which this solution is based. The shape factor F in equation [51] combines L, D and m. In practice these three parameters are only a starting point for deriving F and finite element studies have been carried out to modify the estimate (Tavenas et al 1990). In view of the rather large assumptions that have to be made in applying the solution it is doubtful whether minor adjustments to F greatly influence the precision of the final estimate. For these tests the following argument has been applied:
Assuming a likely variation in the ratio kh/kv is from 2 to 0.5, m takes a value between 1.4 and 0.7. It is reasonable to take m as being 1.
Equation [52] then reduces to
kh = [Q x 8.4 x 10-5]/Hc ...[Equ.53]
This gives an answer in metres with kh in metres per second.
In the system used on site, ‘head’ was actually a pressure (p) in kPa, and flow was in millilitres per hour. The version of equation [53] that can be applied directly to our field results to end up in units of m/sec is:
kh = [Q x 2.29 x 10-9]/Dp ...[Equ.54]
where Dp is the increase over the ambient PWP.
By setting different constant flow rates a range of pressures is discovered. Plotting flow rate against head gives a graph whose gradient can be used to give the permeability directly. Figure 25 is an example of raw test data, where pressure is plotted against time and a number of equilibrium pressures are recorded. Figure 26 is an example of the reduced data taken from this plot, where the pressures are converted to head of water and plotted against flow rate.
The gradient of this plot gives the permeability directly. Note the intersect on the pressure axis is the ambient pore water pressure. If the ambient pore water pressure is known then in theory only one flow rate will be sufficient to give the permeability. However it greatly increases confidence in the data if a number of pressure/flow co-ordinates plot the same trend.
Fig.25 An example of a self bored permeameter test, pressure vs time
Fig.26 Deriving permeability from head vs flow
7.2 Comments on the Permeameter tests
The solution assumes, among other things, directional isotropy, so
kh is constant. No disturbance, swelling, segregation or
consolidation of the material is allowed. In practice the application of
water pressure greater than the insitu ambient level has an unknown but assumed
to be small affect on the ground and other users of permeameters have reported
some consolidation taking place.
The plots of pressure against time do not, for the most part, look as convincing as figure 25. The best results seem to be achieved when the driving pressure changes rapidly to bring the test close to the steady state condition in the shortest possible time – once one pressure has been established, the value for others can be predicted so it is possible to accelerate the test. If the assumptions governing the use of equation [52] are correct then this accelerated procedure should make no difference to the time taken to reach steady state but the results indicate it does. There is no explanation yet for this effect.
Figure 27 is a more typical example from a test with only two steady state values achieved, at 160 and 300kPa.
![[Image]](Anal_27.gif)
Fig.27 A test with only two good readings of steady state pressure
To analyse this test, advantage is taken of the fact that the ambient pore water pressure is known. It is not the value of standing water in the borehole but the water table at approximately 8 metres known from other work carried out on site. For this test, therefore, there is a third value of steady state pressure, 30kPa for zero flow.
Figure 28 is the plot of the reduced data.
Fig.28 Deriving permeability for test PM2 T1
From equation [52] it can be seen that the shape factor is very sensitive to the length to diameter ratio of the equipment but relatively insensitive to the transformation ratio. This suggests that a future version of the self boring permeameter ought to have the ability to modify the length to diameter ratio. Tests at the same position in the soil using different geometry would allow the transformation ratio m to be calculated and hence the vertical permeability to be inferred.
Back to Index of Internal Reports