a. Contacts open
b. When closing the circuit breaker
c. When closing a gate and operating normally
d. When breaking and opening normal currents
e. When opening and breaking a fault.
Cases a, b and c are relatively simple. Normally, in this case, the system is not affected by vacuum loss.
However, cases d and e require further discussion.
Assuming a loss of vacuum at a three-phase feeder vacuum breaker, the switching operation will not result in a fault if the load served by the faulty breaker is a delta-connected (ungrounded) load. Essentially, nothing happens. The two good phases (in this case phase 1 and phase 2) will be able to clear the circuit and the current in the faulting breaker (phase 3) will stop.
A different case of grounded loads, where the opening of the two good phases (phase 1 and phase 2) will not cause the current to stop flowing in phase 3 and the arcing will continue in phase 3. Nothing can stop it, and this current will continue until some backup protection tripping operation is performed, with the result, of course, of circuit breaker destruction.
Since circuit breakers in the 3-15 kV range are primarily used in grounded circuit systems, we studied the effects of faulty circuit breakers in our test laboratory some years ago. By opening the tubes to atmosphere, we deliberately deprived the interrupter of vacuum. We then performed a complete short circuit opening of the circuit breaker.
As predicted, the “flat” interrupter did not successfully clear the affected phase and the “flat” interrupter was destroyed. The laboratory back-up circuit breaker cleared the fault.
After testing, the circuit breaker was removed from the switchgear unit. It was dark, but mechanically sound. Soot was removed from the breaker and switchgear unit, the faulty breaker was replaced, and the breaker was reinserted into the switchgear compartment. Further short circuit opening tests were performed on the circuit breaker the same day. Years of field experience since that test was performed supported the information obtained in the laboratory experiments.
One of our customers, a large chemical company, experienced separate failures on specific circuit configurations (one with an air magnetic circuit breaker and one with a vacuum circuit breaker). Two different installations in different countries were involved. They shared a common circuit configuration and failure mode. The circuit configuration was a contact circuit in which the power supply on each side of the circuit breaker was out of sync, applying approximately twice the rated voltage across the contact gaps, which caused the circuit breaker to fail. Because these faults result from applications that violate the guidelines of the ANSI/IEEE standards and greatly exceed the design ratings of the circuit breaker, they do not indicate a design problem with the equipment. However, the damage caused by the failures is out of interest. In the case of the air-magnetic circuit breaker, the unit enclosure of the faulted breaker was destroyed and the adjacent switchgear units on either side were extensively damaged, requiring extensive reconstruction. The air magnetic circuit breaker was a total loss. In the case of the vacuum breaker, the failure was much less violent.
The vacuum breaker was replaced and arc by-products (soot) were removed from the breaker and compartment. The unit was put back into service. Our experience with testing in the laboratory, where we often explore the limits of interrupter performance, supports these results.
Recently, several tests were performed in our high power test lab to compare the results of attempted interruptions with “leaking” vacuum interrupters. A small hole (approx. 3mm diameter) was drilled in the interrupter housing to simulate a vacuum interrupter that had lost vacuum. The results of these tests were very interesting:
1. The vacuum circuit breaker attempted to break 1310 A on one pole (rated continuous current = 1,250 A). Current was allowed to flow in the “faulty” breaker for 2.06 seconds, at which time the laboratory breaker opened. No parts flew out of the “failed” breaker or interrupter and the breaker did not explode. The paint on the outside of the interrupter flaked off and the rest of the breaker was undamaged. 2.
2. an attempted opening of the second pole of the same vacuum circuit breaker at 25 kA (rated opening current = 25 kA) was made with an arc duration of 0.60 seconds, at the time of the laboratory circuit breaker opening current. The arc burned a hole in the side of the interrupter chamber. The circuit breaker did not explode and no parts of the circuit breaker flew out. Luminous particles were ejected from the hole in the arc chamber. No mechanical parts or other circuit breakers were damaged. Essentially, all damage was confined to the arc chamber where the failure occurred. Our experience has shown that the impact of a vacuum interrupter failure on equipment is very small compared to the impact of failures using other alternative opening techniques.
But the real question is not what the outcome of a failure might be, but how likely is it to fail? The failure rate of vacuum interrupters is so low that vacuum loss is no longer a significant problem. In the early vacuum interrupters of the early 1960s, this was a big problem. Vacuum interrupters are constructed by brazing or welding all the connections between different materials. No organic materials were used. In the early years, a lot of hand fabrication techniques were used, especially when borosilicate glass was used for the insulating casing because it could not withstand the high temperatures. Today, machine welding and intermittent induction furnace brazing use extremely tight process controls. The only moving part inside the interrupter is the copper contact, which is connected to the interrupter end plate by a welded stainless steel bellows. Because the bellows is welded to the contacts and the interrupter end plate, this moving connection has an extremely low failure rate, which is indicative of the extremely high reliability of today’s vacuum circuit breakers.
In fact, the MTTF (Mean Time To Failure) for vacuum interrupters has now reached 57,000 years.
Customer questions about vacuum loss were a legitimate concern in the 1960s, when the use of vacuum circuit breakers for power applications was in its infancy. At that time, vacuum interrupters leaked frequently and surges were a problem. Only one company offered vacuum circuit breakers at the time, and reports indicated they had many problems.
Introduced in Europe in the mid-1970s, the main conceptual differences between the modern Siemens vacuum interrupters and the early 1960s designs are the non-contact materials and process controls. The use of bismuth-copper contacts makes it more difficult to deal with surge phenomena than the use of today’s chrome-copper contacts. Similarly, it is more difficult to control leakage with vacuum interrupters that are primarily handmade than with today’s devices. Today, there is a great deal of emphasis on process control and elimination of the human factor (variability) in the manufacturing process.
The result is that today’s vacuum interrupters can be expected to have a long service life and to apply dielectric stresses to the load equipment that are not significantly different from those associated with conventional air magnetic or oil circuit breakers.