Faulty SCR- Full Description and Solution

Downtime Narrowly Averted with Faulty SCR Drilling DC Drive: In December 2002, on a rig in the GOM, an unusual SCR drilling drive issue occurred. A small amount of AC ripple on a weak power supply caused rapid and uneven wear in the DC motor brushes. Fortunately, the situation did not result in downtime but could have quickly progressed to damage a mud pump DC motor sufficiently to require its replacement.

Equipment Configuration: The equipment in the mud system consisted of DC drilling traction motor, shunt wound, two motors mounted on each of three mud pumps. Dual BAYLOR 6 pulse 750VDC SCR drives are used to drive the mud pump motors. One SCR bay is assigned to each DC motor: two SCR bays and two motors per mud pump. There is load sharing between the SCR drives. Multiple assignments allow the mud pumps to be driven from different SCR drives (Fig 1).

Fig 1

Event: Both motors had been running fine for days at about 75% RPM and 75% amps. There were no alarms on the SCR bay or fault conditions of any kind. Both motors were running without atypical noise, vibration, or heat. During a break in drilling operations, the electrician opened the motors to do a routine inspection. He noticed that all of the brushes closest to the risers (winding end) of the commutator on motor A were extremely short (see Fig 2 and Fig 3). The commutator and brushgear were satisfactory. Since the pump had been running fairly hard for quite some time without problems or alarms, the electrician was not overly concerned, but decided to watch the running motor after drilling operations resumed. The short brushes were replaced and the pump restarted.

Fig 2 - Typical DC Drilling Motor

Fig 3 - Detail of Commutator Brushgear

After restarting, the pump again ran well without obvious symptoms of a problem. The electrician looked briefly at the brushes, motor, and SCR drive during operation. The SCR bay had no alarms. There was no sparking nor other visible commutation problems. The motor with short brushes was running approximately 20° F warmer than the other motor on the same pump. After 10 hours the electrician inspected the motors again. The same brushes were once again very short and close to failure.

SCR bay assignments were changed so that the mud pumps were driven by different SCR bays (Fig 4). After the change, there still were no apparent problems. After 10 more hours the suspect motor, now driven by a different SCR bay, was examined. This time brush wear was normal. Motors on the other Mud Pump 2 were examined. The brushes on motor A (the motor driven by the suspect SCR bay) were dangerously short. Again only one set of brushes was short, but this time it was the brushes at the other end of the commutator, the end closest to the bearing. The SCR drive was taken out of service and examined in more detail. Eventually a small amount of AC ripple on a weak power supply was discovered. The power supply was replaced and the pumps started up again. Brush wear was normal on all motors, as was motor, pump, and SCR bay performance. The pump motors have subsequently experienced no problems.

Fig 4

Analysis: In this instance a relatively minor AC ripple, not even enough to generate an SCR alarm, on a power supply was the root cause of the brush failure. A brush failure could have occurred with no warning. Properly installed brushes on DC motors normally wear (or burn) at the same rate. Drive faults normally cause vibration, sparking, or burn all brushes, not just one set. In this case, it would appear that an amperage overload was only experienced at one end of the commutator. SCR design engineers were consulted. This problem has been seen before but is very rare. Only the diligence of the electricians routinely checking all the DC motors prevented catastrophic motor failure. The major operational problem is that even though motor failure would occur in less than 24 hours, the fault was not bad enough to cause a system alarm or noticeable vibration.

Solution: Regular equipment monitoring by the rig electricians identifies and prevents failures, as proven in this case. If the rig has a Vessel Monitoring System (VMS), it can be adjusted to alarm on a small temperature differential between two motors on the same pump, alerting the electrician to check the motors. If there is no VMS, it would not be difficult to add differential temperature monitoring and alarms to the mud pump motors.

return to top