ABSTRACT

Successful reservoir description depends on identifying the form of hydraulic continuities and discontinuities in rock masses between wells. This ideally needs a combination of high resolution seismic, quantitative well correlation and core analysis. The quantitative correlation of wire-line logs is an important contribution in areas with little core and low resolution seismic. However the use of wireline logs for inter-well correlation and matching is often limited to a manual match of a small number of traces, looking for trace patterns, and characteristic trends. In the last few years several authors have published papers applying gene-typing algorithms to the correlation of geological sections. These algorithms were developed by molecular biologists for coping with the irregular patterns of amino-acids in the human gene. They were developed to handle explicitly duplication, stretching, and missing sequences, all of which are common features in geological well data. The problem with regular application of this approach in geology is that the lithological terms normally used are rather imprecise (unlike amino acids) and lead to ambiguities in the correlation. The present paper shows how the use of a numerical lithology derived from a sub-set of petrophysical data is combined with gene-typing algorithms to give detailed correlations. Successful correlation should also use knowledge of the probable lateral extent of each unit being matched, and this information has been incorporated in the present approach. Examples are shown from the Norwegian North Sea.

This content is only available via PDF.
You can access this article if you purchase or spend a download.