Basic HTML Version
OH – ET – VA - LL: Analysis of Dynamic Data in Shale Gas Reservoirs – Part 1 – Version 2 (December 2010)
p 19/24
11 – Adding unconsolidation
Here again unconsolidation is an effect that can be introduced in a relatively simple manner in
the full non-linear numerical model through the definition of k(p) and
(p) laws, whereas it just
cannot be handled properly in analytical models nor in straight-line analysis, to our knowledge.
Let us start from the reference multi-fractured well model with desorption (§3, with k=1E-4
md) and add unconsolidation to it, with the following relationships:
In the above, c
r
is the rock compressibility (taken to be 3E-6 psia
-1
here), and p
0
is the
reference pressure (taken to be p
i
= 5000 psia). The equations above are simple models that
are typically applied to CBM reservoirs – so we have no reason to use them here, except that
they seem to be widely used by the industry for shale gas. Nevertheless, the pure qualitative
analysis leads to the conclusion that ... unconsolidation as we define it has no influence on the
reference case simulation:
We would need to increase the rock compressibility to unrealistic values to see any influence at
all. In the simple model we have used the fractures are not affected by unconsolidation (their
conductivity is infinite, anyway), only the matrix permeability / porosity can change;
furthermore those parameters do not vary by more than 10% during the simulation... In other
words, our current model may not be well adapted to this shale rock + hydraulic fractures
geometry, and we need to modify it - incorporate variability on fracture conductivity, for
instance - before to go further...
Summary:
At this stage we have the unconsolidation in the model but it does not seem to
have a noticeable effect. We may be using incorrect parameters and this remains to be
checked.