Cyclic Borehole Effects in Deviated Wells.
Highly deviated wellbores sometimes suffer from a cyclic variation in borehole size. Even though the caliper oscillations may be relatively small, a salty mud can combine with the periodic hole size variation to produce wireline data that has been severely compromised. Interestingly, it may be the deepest reading tool (resistivity) which suffers the largest degradation.
A straight-forward solution, calibrated to specific wellbore conditions, has been developed which facilitates a much more representative formation evaluation.
Upon occasion, the bottom-hole drilling assembly in a deviated well can yield a cyclic, or corkscrew, hole which has small (+/− 1/4") oscillations about the mean wellbore radius. Figure 1 is the CBIL image of the borehole in one such instance, and it's seen that the borehole radius has a distinct, periodic character, with an approximate 4' wavelength. In this image, the wireline tool is rotating while being pulled out of the hole, giving a corkscrew appearance. The actual hole size, however, is oscillatory in that the hole dia meter systematically changes, along the wellbore (coaxial) direction.
Typically, the problem arises when changing from sliding to rotating drilling modes. While sliding, the bent housing steerable mud motor produces an in-gauge hole, but when one switches to rotating (in order to drill ahead Lister), the bent assembly will often interact with the stabilizer, to produce alternating hole sizes. The period of the oscillation is determined by the bit-to-stabilizer length.
While the volumetric variation is extremely small in comparison to the total volume seen by the deep resistivity tool (or even the porosity logs), the periodic nature of the caliper oscillation call yield a log anomaly (noise) which nearly overwhelms the basic formation response.
By working in the frequency domain, however, it is possible to characterize the "noise" via a Fourier Transform of the caliper log, which is the physical manifestation of the problem. And to then design a filter which may be applied to the remainder of the logs. In effect, one removes from the resistivity and porosity data, those spatial frequencies which correspond to the "noise" found in the caliper log, leaving the basic formation response.
We have typically encountered the problem in highly deviated wells. Since the formations of interest are usually thick and nearly horizontal, true formation variations in the historical sense (along the wellbore) are gradual. As a result, the high cut filter designed by examination of the caliper Fourier spectra, does not seriously compromise the resulting evaluation.
The approach developed here is general, and may be applied in other deviated wells as necessary. The Fourier Transform concept can further be used to quantify bed boundary resolution, at various logging speeds, thereby allowing one to optimize (particularly uphole passes, for correlation purposes) wireline operations (minimizing rig time).
The Fourier Transform amplitudes are key to proper filter design, so that one must be familiar with basic Fourier principles.