Geology Reference
In-Depth Information
Hydraulic Properties of the Different Sets of Fractures
Fracture Aperture, Fracture Permeability and Fracture Storativity
These parameters refer to single fracture and have to be calibrated. There is
poor site specific information about realistic bracketing values, since no tests
between packers are available, and since even the slug tests we are looking
at refer to open hole sections that may intersect more than one fracture (2
to 5 intersections are frequently detected in the present numerical models).
Fracture aperture are therefore set to 10
-2
m, and we assume that the infilling
materials, some weathering by products trapped in between the natural rough
fracture walls, have a porosity of about 30%. We investigate fracture
permeabilities ranging in between 10
-2
ms
-1
and 10
-4
ms
-1
. Dealing with
storativity, we started with values close to that of confined aquifers, about
10
-4
to 10
-5
, and move towards values more appropriate to unconfined
situations, i.e. some per cent.
Calibration Procedure and Model Outputs
Size and Shape of the Modelled Volume
The modelled volume of rock is a vertical cylinder, 50 m in diameter and
30 m in elevation. A vertical bore-hole is simulated along the vertical cylinder
axis. The open hole section is 22 m, centered at the mid-height of the block.
This value reflects the average aquifer thickness, which was observed on the
25 tested IFP wells.
Inner and Outer Prescribed Conditions at the Model Boundaries
The outer surface represents an open boundary where hydrostatic conditions
prevail all the time. These conditions are applied at all the fractures (i.e. at
the fracture centre in the numerics) that intersect this outer boundary. Along
the central bore-hole is the inner boundary, where a transient chart is
prescribed. Two phases are described. During the first two seconds, the
hydraulic head at the well linearly increases from 0 to 0.49 m, the theoretical
maximum value of the change in water level. Then from 2 to 2100 seconds,
the well is subject to a 'no flow' condition and the head perturbation dissipates
in the fracture network. The initial hydraulic head distribution is uniform
and set to 0 m.
Calibration Strategy and Results
The scenario is applied and we present the period during which the pressure
perturbation vanishes in the fracture network, in a normalized way, starting
from 0 at the peak time and increasing to 1 as time increases. Series of 10
equi-consistent network alternatives are generated and we produce a set of
