Naturally occurring fractures often play a very critical role in defining a shale play, and fracture identification, typically from image logs, has become a key component in a range of situations from efficient drilling design to reservoir development. In this vein, to complement the usage of image logs in fracture detection work, recent applications of multi-component (or triaxial) induction log data were investigated in a four-well US resource play case study.

The multi-component induction tool data were acquired in all four vertical wells in the study. All of these were drilled with oil-based mud systems. Additionally, electrical and acoustic imaging logs were run in three of the wells to compare the fracture indications between the various tools. Whole core was also available to provide visual physical evidence of fracture existence and features.

Inversion of multi-component induction data normally yields bedding dip magnitudes and azimuths. An extension of this inversion technique provides a qualitative indicator of high-angle fracture presence and intensity from the same data. These fracture planes do not have to intersect the borehole, but can be several feet away and still be detected.

The focus of this work was to benchmark induction tool-based fracture characterization in oil-based mud against both slabbed core and traditional wellbore imaging log interpretation. In one of the wells where operational problems precluded image log acquisition, only the induction data were available. In this case, we found an important correlation between the induction log fracture derivation and drilling events and data.

We assert that this new process promises to be an important tool for the recognition and characterization of natural fractures, especially in conditions where pad-contact image logging tools are sensitive to washouts, rugose borehole, deviated wellbores and other adverse borehole conditions, and especially in heavy oil-based mud where acoustic and electrical data quality is often compromised.