SPWLA Thirty-First Annual Logging Symposium, June 24-27, 1990        PAPER R

Paper q

 

DIFFUSION CORRECTIONS TO PULSED NEUTRON CAPTURE LOG: METHODOLOGY

 

WE Murdoch, C.J. Hunter, KR. Randall, and C.W. Towsley

Atlas Wireline Services, Western Atlas International, Inc., Houston, TX

 

 

ABSTRACT

 

Without explicit environmental corrections, a pulsed neutron capture (PNC) lifetime measure­ment provides, at best, a good approximation to intrinsic thermal neutron capture cross section (Sint) Our industry has recognized for years that Slog values are affected by environmental factors other than Sint.  Although logging system designers have taken several measures to reduce or correct for the influences of borehole decay and neutron diffusion, there is still room for improvement.

 

Ideally, we would be able to obtain Sint  independently of borehole geometry and fluid. An optimum correction technique would be robust under all realistic borehole conditions, in­cluding those filled with fresh water and gas. The neutron diffusion problem has not lent itself to useful mathematical treatment over the full range of normal oilfield conditions. Thus, we have chosen to develop empirical corrections to Atlas Wireline Services PDK-100 PNC logs based exclusively on a comprehensive set of laboratory measurements.

 

A unique set of synthetic formations provides the advantages of homogeneity precisely known porosities, and exactly equal Sma values for all configurations. From data acquired in such formations, we have constructed correction functions to transform the  Slog values from long-and short-spaced detectors directly to Sint.  In their present form, the correction functions are specific to particular borehole/casing diameter combinations and require borehole fluid salinity as an input parameter. Corrections for gas-filled boreholes are also available. Formation porosity has proved to have negligible effect on this particular diffusion correction scheme when porosity is moderate to high. Similarly, Monte Carlo studies show little lithology effect. Previously recorded PDK-100 logs are supported, since the corrections are designed to be applied to the standard PDK-100 log outputs (SGMA and SIGL).

 

Field trials in several geographical areas illustrate that borehole fluid dependency can be effectively removed, thus providing more realistic intrinsic cross-sectional values.