Abstract

An innovative method of log analysis has been developed for thin bed formation evaluation. It has been successfully applied to conventional, commonly available, log suites recorded at conventional sampling rates. The method produces geologic and engineering answers without the need for a dipmeter or other high resolution device. This paper documents the steps required to perform the processing, the applicability of the method and the verification of the results obtained to date.

Some previously published thinbed log analysis routines include algorithms for the enhancement of the vertical resolution of the raw logs as a preliminary to performing a modified conventional shaly sand analysis. Effectively these methods assume that thinly bedded sand/shale sequences are accretions of shaly sand with continuous, but rapidly varying, clay content. Thus standard (dispersed) shaly sand logic is applied, but on a much finer depth scale than is used in non-laminated rocks.

An alternative approach is to assume that individual laminae in a sand/shale sequence are pure and that laminated shaly sands are binary mixtures of pure constituents, i.e., sandwiches of clean sand and shale. In a given sedimentary environment either extreme is possible. In the absence of direct measurements from full core or a borehole imaging device it is recommended that both the binary approach and the high resolution dispersed shaly sand approach be taken and an informed judgement be made as to which one to rely on.

The current method for binary lithology analysis combines a number of steps including:

• Computation of shale content and effective porosity using conventional techniques.

• Raising the vertical resolution of the formation conductivity to that of the neutron/density data.

• Computation of sand/shale bed boundaries based on an adaptive filtering technique.

• Computation of the petrophysical properties of the clean sand portions of the binary mix using the enhanced conductivity, laminated logic and the results from previous steps.

• Accumulation of a sand count, and the integration of the hydrocarbon pore volume through the sand beds so defined.

• Preparation of an appropriate display format.

Evidence to date suggests that individual beds as thin as 3 inches (7.62 cms) can be resolved in this manner. Comparisons with core analysis, core photos and borehole imaging devices confirm the reliability of the method. Disadvantages of the method include reliance on a deconvolution process subject to error in the presence of noise. This disadvantage is partially compensated for by the wide applicability of the method to existing log data bases and the ease of porting the logic to simple log analysis platforms that perform at fixed sample rates.