Continuous wavelet transformation: A novel approach for better detection of mud pulses

Mud pulse telemetry systems use coded pressure pulses (carriers) for transmitting measured downhole information (directional and geological information) to the surface in real time. The detection of the pulses and the extraction of their characteristics (start, duration and frequency) in a noisy pressure signal measured in the mud column at the standpipe are very essential for obtaining correct information and making educated decisions about the drilling operation while drilling. This paper presents a new approach for mud pulse detection at the surface. It uses continuous Morlet wavelet transformations to break the original measured time signal down into distinct localized frequency packets. Experiments were carried out using a flow loop equipped with a mud siren pulser and a special "actuator system". Sinusoidal pressure pulses with various frequencies were produced by the mud siren using different modulations such as frequency shift keying and amplitude shift keying. At the same time, hydraulic noise with various amplitudes and frequencies was induced to the flow loop by the actuator system to simulate typical noise patterns created by the common duplex or triplex mud pumps. The hydraulic noise was overlaid to the mud pulses, resulting in complex time signals for the pressure in the flow loop. For all experimental runs the noise amplitudes induced by the actuator system were greater than the pressure pulse amplitudes generated by the mud siren. The wavelet toolbox in the MATLAB software was then successfully utilized to separate the hydraulic noise from the pressure pulses again and to extract the pressure pulse characterization. The results indicate that the continuous Morlet wavelet transformation can be used to clearly detect the weak pressure pulses coming from the mud siren in a noisy environment. It is possible to identify the carrier periods, frequencies and discontinuity positions in the time domain. The new approach can be used to transmit data at a faster data rate. Furthermore, wavelet transformation offers a smart graphical presentation of the mud pulse signal in two dimensions (time-scale or time-frequency) for visual inspection. (C) 2013 Elsevier B.V. All rights reserved.