A Novel Approach to Borehole-Breathing Investigation in Naturally Fractured Formations

Abstract

Summary The occurrence of reversible mud losses and gains while drilling in naturally fractured formations (NFFs) is of primary concern. Borehole breathing can complicate the already difficult practice of fingerprinting the changes in the return-flow profile, hence undermining the reliability of kick detection. Issues can also derive from misdiagnosing a kick and attempting to kill a breathing well. The objective of this work is to correctly address the phenomenon and increase insights regarding its physical characterization. The fluid progressively flows in and out of fractures as a consequence of three mechanisms: bulk volume deformation, fluid compressibility, and fracture-aperture variation. To represent this complex scenario, a model involving a continuously distributed fracture network is developed. A time-dependent, 1D dual-poroelastic approach is coupled with a variable fracture aperture and a passive porous phase. Finite fracture network length is considered, and no limitation on the number of fractures is posed. The latter permits us to analyze long openhole sections intersecting several fissures, which is a more realistic approach than the available single-fracture models. The proposed model is able to quantify pressure distribution in fractures and pores, together with the flow rate entering or exiting the fractures. Furthermore, a useful application of the model is proposed by suggesting its application as a breathing discriminator during kick diagnosis. The shut-in drillpipe pressure (SIDPP), recorded from a real kick, has been compared with one caused by a simulated breathing case. Although the two SIDPPs show significant similarities, the correct modeling of breathing can help the identification of the major differences between a kick and breathing.