| ABSTRACT Drilling vibration is concerned with the dynamic behavior of drillstrings and their components while drilling. It is becoming an increasing important subject due to growing awareness that many drillstring failures and drilling problems can be prevented by monitoring the dynamic behavior of the string. As drillstring fatigue and bit wear can be linked to the stick-slip motion, limiting it by simple surface vibration monitoring is very attractive concept. Vibrations generated while drilling are frequently regarded as being detrimental to the drilling process, causing problem such as bit damage or drill string wear. An alternative view of vibrations is to consider them as information, promising interpretation related to bit wear, bit failure, stick-slip, even as formation hardness of the rock. The roller cone bit fulfils its task of destroying rock by means of successive percussions. Part of the these percussions destroys the rock; another part is reflected into the drillstring and is partly conveyed to the surface. This dissertation considers that the useful information can break down into three intervals: a"low frequency" (LF) interval covering the DC-25Hz domain corresponding to many RPM dependent excitation mechanisms including 1xRPM, 2xRPM and 3xRPM harmonics of the rotation speed because of tricone bit, a"midst frequency" (MF) interval covering the 25Hz-200Hz corresponding to the spectum of tricone teeth cutting the formation and a "high frequency" (HF) interval covering the 200Hz-10kHz corresponding to sonic wave"noise"reflecting variations of rock hardness. Summarizing 3 years of study work and field experience, the chief contribution of this dissertation is: 1. To establish and develop the drillstring vibration wave model and formulate the resonant frequencies of the string. 2. To design a data acquisition system, the surface measurement system of drillstring vibration, that receives, measures, and analyzes sonic waves. 3. To propose ways of analyzing drillstring vibration, i. e. Fourier analysis FFT, power spectral analysis BURG algorithm, cepstrum analysis and wavelet analysis of vibration. 4. To present a reliable process methods, signal frequency modulation (FM) and amplitude modulation (AM) , aiming at monitoring drillstring working status and predicting bit failure. 5. To prove the ability of identification in detecting rock hardness variance using vibration signal and its"noise spectum. " |
