The Antonio J. Bermudez basin in Southern Mexico is a low porosity massive Jurassic and Cretaceous carbonate reservoir that is extensively faulted and fractured due to post depositional salt intrusions. The natural fractures create many drilling challenges and obstacles. Underbalanced drilling with foam fluid systems has been used to minimize mud losses in these fracture systems. The underbalance drilling, and drilling with casing greatly improves well construction, but these drilling techniques also create many formation evaluation challenges. For example, open hole sonic logs require a liquid filled borehole. Also, formation resistivity is such that lateral logs would be the preferred resistivity device, but they require a conductive borehole fluid.

Artificial Neural Network (ANN) methods have traditionally been used for reconstruction of petrophysical data due to tool pulls and poor borehole conditions. Drilling obstacles, in this field, sometimes prohibit the running of conventional open hole logs. Thus, an ANN technique has been developed that uses cased hole pulsed neutron log (PNL) data to synthetically generate a conventional, open hole, triple combo log. These logs are used for well-to-well correlations and for petrophysical evaluation. Still, due to the synthetic smoothing of an ANN result based on casing formation evaluation data (i.e., PNL), detection of natural fractures remains problematic using ANN-based synthetic density data. On the other hand, cased hole dipole sonic anisotropy analysis is routinely used successfully to identify natural fracture systems and far field stress orientation for geomechanical applications. Intervals to be perforated are selected by combining through-casing sonic information with petrophysical analysis of synthetic open hole logs.

Data and production results from several wells will be presented and discussed in the paper.

You can access this article if you purchase or spend a download.