Page 30 - Well Logging and Formation Evaluation

20 Well Logging and Formation Evaluation

at different rotational speeds (equivalent to different pressures) is mea-
sured. This technique also has the advantage that the sample if not
handled during the experiment.

1.6.4 Limitations of Core Measurements

There is a tendency among petrophysicists to treat measurements made
on cores as “gospel” and not to question the reservoir parameters so
derived in their petrophysical model. The following may give reasons why
the core data are not always correct:

• A core is a section of rock cut usually over only a subset of the reser-
voir in a particular part of a field. There is no a priori reason why it
should be representative of the reservoir as a whole. In particular, a core
cut in the water leg, where diagenetic processes may be occurring, is
not necessarily representative of the oil or gas legs in a reservoir.
• The coring and recovery process subjects the rock to stress and tem-
perature changes that may profoundly affect the rock structure.
• The plugging, cleaning, and drying process may completely change the
wettability of the plugs, making them unrepresentative of downhole
conditions.
• Resistivity measurements performed on plugs at ambient temperature,
using air as the nonwetting fluid, may be wholly unrepresentative of
reservoir conditions. Apart from the fact that the brine has a totally dif-
ferent resistivity at ambient temperatures, there may be other factors
affecting how easily the nonwetting phase may mingle with the wetting
phase. In fact, where experiments have been performed to measure m
and n under truly in-situ conditions, it was found that the values dif-
fered completely from those measured under ambient conditions.
• When measurements are made on a selection of, say, 10 SCAL plugs, it
will typically be found that the m, n, and P c behavior of all 10 will be
completely different. These are usually then averaged to obtain a repre-
sentative behavior for the reservoir. However, because of the variability,
if a new set of 10 plugs is averaged, the result will be completely differ-
ent. This calls into question the validity of any average drawn from 10
plugs that are taken to represent thousands of acre-feet of reservoir.

Overall, it is my conclusion that it is better to use core-derived values
than nothing at all, and a lot of valuable information about the reservoir
can be gained from core inspection. However, no core-derived average