Page 350 - Standard Handbook Petroleum Natural Gas Engineering VOLUME2

Estimation of Waterflood Residual Oil Saturation 317

successfully used, a test in a 100% water saturated formation yielded oil satura-
tions of 40%-60% [353]. These obviously incorrect values may have resulted from
incomplete displacement of formation water close to the wellbore by the injected
low salinity water.
Accuracy of the nuclear magnetism log depends largely on the signal-to-noise
ratio which can be improved by making repeated readings. Newer versions of
the NML tool have improved signal-to-noise characteristics. As with other log-
inject-log techniques, the possibility of fluid drift in the formation should be
considered. Fluid drift should not be sufficient to move injected paramagnetic
ions away from the wellbore which would cause erroneously high values of Sor [349].

Summary of Methods for Estimating Residual Oil
Economics of primary and secondary recovery processes are usually suf-
ficiently attractive to permit considerable error in the estimation of recoverable
reserves. However, for tertiary recovery the amount of oil remaining in a
reservoir and its distribution must be known with reasonable confidence. Firstly,
a reliable estimate of residual content is extremely important to technical
evaluation of field tests. Secondly, the high front-end costs of tertiary processes
are such that overestimates of residual oil saturation could have disastrous
economic consequences. Thus, a well-planned effort to measure residual oil
saturation is a necessity before any tertiary recovery application. Under favorable
circumstances, accuracies of e% of reservoir pore space can be achieved. In general,
accuracies will not be this good, but values within fi% are considered necessary.
There is no absolute measure of residual oil saturation for a reservoir. When
evaluating a tertiary prospect, a combination of methods should be used which
provide information on both amount and distribution. Evaluation will normally
begin with material balances using information that is already available. Fre-
quently quoted nation-wide estimates of the amount of residual oil that is
potentially available for tertiary recovery are based mainly on material balance.
Comparisons of material balance with other methods of determining residual
show unacceptable scatter, and on average, the material balance gives saturations
which are too high by about 9% (by pore volume). This corresponds to about a
30%-50% overestimate in amount of residual oil. It has been suggested that a
much higher ratio (say 2 to 1 or more) in residual oil determined by material
balance to that given by other methods is an indication that the reservoir
contains extensive regions of high oil saturation and may, therefore, be a good
prospect for infill drilling.
In addition to material balance, other estimates of residual oil from resistivity
logs and laboratory waterflood tests may also be available. However, in general,
none of the conventional methods of determining residual oil saturation-analysis
of conventional cores, laboratory displacement tests, conventional logging,
material balance-are considered sufficiently reliable in themselves. They can pre
vide useful guides as to whether a tertiary prospect should be investigated further.
Over the past 15-20 years there has been increased field testing of a number
of more sophisticated techniques: pressure coring, tracer tests, and the various
types of log-inject-log procedures. These methods vary in the conditions under
which they can be applied and the type of information they provide.
Pressure coring and the sponge core technique provide information on the
vertical distribution of residual oil and also have the advantage that the core
analysis procedures directly demonstrate the presence of oil. Because of the
possibility of flushing, results may tend to be conservatively low. For pressure