Core-to-Log Comparison—What's A Good Match?

Abstract

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractIt is common practice in the petrophysical community to either calibrate log porosities to core measurements or to use the porosities, permeabilities and water saturations measured on core to verify that the log interpretation models used are appropriate. This paper advocates the latter approach, since such "tested" or "verified" models are more likely to be valid away from the cored wells and intervals. Using the core data to verify the log interpretation models also allows the log and core data to be compared and the uncertainties in the logderived properties to be quantified. This paper describes quantitative methodologies for comparing log-derived porosities, permeabilities and water saturations with their core-based equivalents. The methodology uses the differences between the log and corebased estimates to quantify uncertainty ranges at any arbitrary level (e.g. P90, P50 and P10).Quantitative comparisons are at their most useful when the uncertainty ranges derived can be placed within a known framework so that it is apparent whether or not the match is a "good" one. This paper uses experience gained from a number of Fields spread through a variety of Basins to show sort of comparisons are common. From this "Field experience" it is possible to determine whether or not log-to-core matches are within the expected range i.e. whether or not the match is "good". If it is not "good" then additional refinements may be required to improve the comparison to an acceptable level. In any case, the uncertainties associated with the log-derived estimates can be quantified.In general, the mean difference between log and core porosities should be less than 0.6 p.u., that between log and core permeabilities should be less than a factor of 1.5 and between log and capillary pressure (saturation-height) based water saturations should be less than 6 s.u.