Advanced Spectral Modeling of Noise Log Data: The Key for a Comprehensive Downhole Dynamic Characterization

Abstract This paper discusses the enhanced use of noise logging aimed at characterizing the dynamics of complex reservoirs and addressing wellbore integrity issues. The methodology makes use of a fit-for-purpose quantitative spectral analysis of noise log measurements and can provide direct and fast information about well completion integrity, post-stimulation job efficiency, fluid flow path in the near wellbore region, reservoir porosity characteristics and flow-units identification. The approach is presented by means of a study performed on several wells intercepting different heterogeneous reservoirs and characterized by complex completions and, sometimes, by intensive stimulation jobs. In details, a high-resolution noise pattern modeling in a wide frequency range is performed to discriminate the character of the recorded flow noise in terms of mesopores, macropores, fractures, behind-casing channels and completion elements (including active valves and leaking packers). In favorable scenarios, the noise power amplitude is also used to understand the contribution of active reservoir units. It is proven that providing a quantitative noise pattern classification is fundamental to recognize unusual poor cement placement issues, not detectable by standard sonic and ultrasonic cement logs and to discriminate between leaking and sealing packers. Moreover, in case of acid and/or acid fracturing treatments in carbonate reservoirs, the methodology can identify the generated wormholes/fractures and quantitatively evaluate the efficiency of the stimulation jobs by means of noise power analysis in the related frequency range. In addition, a dedicated spectral noise modeling is also used in order to identify flow-unit contributions in multi-layer scenarios and the type of porosity providing the flow. The reliability of the workflow comes after a successful comparison with the available standard production logging interpretations. The integration of this approach with standard workflows completes the reservoir characterization providing additional dynamic outcomes. The key role played by the enhanced modeling of spectral noise log data demonstrates the versatility of the methodology. Although the added values of this logging technique are already known, the quantitative use of noise power amplitude in selected frequency ranges is relatively new and can shed light on this topic for future advanced applications.