Traditional log analysis is an inverse procedure in which an
interpretation of reservoir fluid volumetrics and pore properties
is made from indirect physical measurements of a logging tool.
When logged zones from a well reservoir interval are plotted on
a Pickett resistivity - porosity crossplot, they form a trajectory
whose trace is dictated by petrofacies and relative height in
the hydrocarbon column. The trajectory graphs the log response
"effect" rather than the reservoir variables "cause"
so that multiple interpretations can be possible when attempting
the inverse path from log response effect to reservoir cause.
Even when correct, the interpretation of reservoir properties
is usually limited to a qualitative assessment rather than a'
quantitative characterization. Forward-modeling offers the alternative
strategy in which reservoir models of capillary pressure can be
converted to logging measurements and plotted on a Pickett plot
for comparison with measured data. Forward-modeling provides realizations
rather than solutions, and so must be viewed as critically as
traditional inverse interpretation techniques. However, the models
are generated from numerical rock measurements, so that useful
reservoir characterizations can result from good matches of model
to logging data. An ideal strategy combines both inverse interpretation
and forward modeling within an integrated analysis.