Improved CBM of top drives using advanced sensors and novel analysis techniques

Although technology in the oil and gas industry is always advancing, there are still numerous tools, sensors, and analysis techniques used regularly in the field that have remained unchanged for many years. This is especially true when observing critical drilling components such as the top drive. Top drive monitoring techniques today are only effective after a severe problem occurs on the top drive. The consequence of this is that when the top drive fails, it is always unexpected, resulting in unnecessarily high repair costs (an average of $60,000 - $120,000 per failure) and, even more importantly to operators, downtime to the operations (which ranges from $100,000 / day for land operations to larger than $1,000,000 /day for large-scale offshore operations). These costs arise from the fact that when the top drive fails, drilling is effectively shut down until a repair is made. A solution to this problem is to change the fundamental way in which top drives are monitored. Rather than using traditional time-based preventative maintenance, monitoring strategies can be changed to encourage condition based maintenance (CBM). CBM varies from scheduled or preventative maintenance in that it fundamentally relies on data analyzed from the machinery to ascertain whether maintenance is required. CBM has been applied in many industries throughout the past half century and opens up a new dimension in performance by quantifying the ongoing health of a monitored machine. Three separate analysis techniques are recommended for implementation onto top drives: vibration analysis, oil analysis, and thermal analysis. Vibration analysis and oil analysis have proven successful in other industries to detect mechanical faults in components such as shafts, gears, bearings, and motors. For thermal analysis, a novel methodology is introduced which relies solely on temperature sensors already installed on most operational top drives. When model-based expected temperatures deviate sufficiently far from measured temperatures it is assumed that the system has thermally failed, triggering an alarm to the operator. The software code to implement such a procedure has been developed for a programmable logic controller (PLC) and is currently being tested in the field to obtain real-time analysis results.