A calibration-free inversion algorithm for evaluating cement quality behind highly contrasting steel pipe

In the context of subterrain well inspection, it is important to efficiently and accurately evaluate the quality of the cement sheath that fills the annular region between the rock formation and the casing, which is a string of steel tubes. A continuously increasing number of such wells are being drilled in extreme environments requiring heavy, attenuative fluid filling the borehole for well stabilization, and larger and thicker casings for mechanical stability. Under those conditions, examining and ensuring the quality of the cement for safety reasons is of paramount importance. We have studied the limits of the most common ultrasonic pulse-echo-based technique for cement evaluation and corrosion detection. Its principle is based on the analysis of the resonance frequency and decay rate of an ultrasonic pulse reverberating in the casing. By analyzing the echo it extracts information about casing geometry and the acoustic impedance of the outer medium. Due to the specific geometrical requirements of the technique, the inhomogeneous, attenunative fluids and the thicker-walled casings pose increased demands on the processing approach. To overcome this problem we have developed a robust, calibration-free inversion algorithm based on a full 3D-model of the ultrasonic pulse-echo transducer which honors the physics of all interactions and the actual geometry. Unlike other methods, the here presented technique does not require the a priori knowledge of environmental parameters such as fluid properties, which are instead deduced from the measurement itself. The exponential decay of the received echo envelope depends on two parameters: the acoustic impedance of the fluid and that of the annulus both of which we find simultaneously by fitting the slope and the offset of the tail. In addition, the algorithm's optimization of performance time by several orders in magnitude allows usage of this processing in real time on a standard computer. Extensive laboratory studies and field testing have been carried out to support and validate the algorithm over an extended range of logging conditions, fluid types and a comprehensive set of casings with diameters up to 22 in. and wall thicknesses up to 1.25 in.