This paper presents a medical approach to classify shock waveforms acquired at 31,250 hertz downhole. The shock signals are treated as drilling electrocardiogram (D-ECG). The D-ECGs are processed using clustering algorithms and merged with drilling incidents to identify an arrhythmic signature pattern that can lead to catastrophic failures.
In medicine, the analysis of heartbeat cycles in an electrocardiogram signal is very important for monitoring heart patients. In the drilling industry, downhole shocks are present most of the time. They are present so often that the authors introduce the concept of drilling electrocardiogram (D-ECG) based on shock waveforms acquired at high frequency. The shock module was implemented in hardware using a field programmable gate array (FPGA) and run inside the control unit of an RSS to complement the navigation systems composed. The shock acquisition and processing are performed at 31,250 Hz, providing enough bandwidth to fully reconstruct high-frequency events.
A novel methodology combining field incidents with machine learning clustering algorithms is proposed to identify arrhythmic shocks signatures and whirl and bit bounce in real time, preventing failures to the BHA.