Breaker testing in nuclear installations
Moulded case circuit breakers (MCCBs) are used by the hundreds in nuclear installations, where they play a key role in protecting vital equipment. However, because of the limitations associated with traditional techniques, MCCBs have, in the past, rarely been tested. Now is the time for this situation to change.
Like many devices with safety related functions, it’s often hard to know that an MCCB has developed a defect until it is called upon to operate. Unfortunately by then, it’s too late to do anything about the problem.
In nuclear installations, most of which have hundreds if not thousands of motor drives and power distribution circuits protected by MCCBs, this is a particularly serious issue, as the correct or at least predictable operation of virtually every item of plant has a bearing on overall safety.
The obvious solution is to test MCCBs regularly but, in practice, this is almost never done. The reason is very straightforward – testing MCCBs by traditional methods is costly, inconvenient and time consuming. In addition it often yields results that are of dubious value. Let’s see why.
The standard approach to MCCB testing involves removing the breaker from the equipment in which it is installed, and transferring it to a dedicated test assembly. There, it is subjected to tests designed to establish that its operation is within acceptable limits.
Research by Megger in conjunction with engineers at Pacific Gas & Electric Company’s Diablo Canyon nuclear power plant in the USA shows that this process typically takes a full eight hour shift to test just one breaker. This may sound hard to believe, but the justification is simple.
Taking the motor control centre or distribution board in which the breaker is used out of service requires around an hour. Removing the breaker takes another hour, followed by 15 minutes of radiation screening, and 15 minutes to transport the breaker to the test facility. The test itself takes two hours, with a further hour to return the breaker to the panel, re-install it and put the panel back in service.
Add in time for the pre-job brief, a break and lunch, and it’s easy to see why testing one breaker takes a whole shift. And that’s not the only problem. Testing the breaker outside its normal operating environment often affects the accuracy of the results obtained.
The change in the local magnetic field can for example, alter the characteristics of the magnetic trip function, while the change in ambient temperature does the same for the thermal trip function. Further, the connections used in the test rig are unlikely to match the impedance and thermal resistivity of the connections with which the breaker is normally used.
A much better approach would be to test the breaker without removing it from the MCC or switchboard. This has now been made possible by the introduction of the Oden MCCB test set from Programma.
Testing in place may sound something that would be easy to achieve, but there are many factors that have to be considered. For example, when MCCBs are tested outside the panel, the length and size of wire used to make connections to them for the test is predefined. Clearly, it’s not practical to stick to these requirements when testing in situ.
The Oden test set overcomes this problem by using a handheld probe. This approach has been exhaustively evaluated and shown not to make any difference to the results that would have been obtained had the standard wire been used. Another benefit is that the probe is insulated and rated for the working voltage of the equipment under test, therefore eliminating the need for users to wear special safety gear.
The Oden test set also has an I/30 function that injects one-thirtieth of the normal test current. This avoids pre-conditioning before the test is carried out and also eliminates the 20-minute cooling down period that would otherwise be needed.
The result is that the time needed for each test is greatly reduced, as a practical example at Diablo Canyon clearly shows. Taking the MCC out of service still takes one hour as before, but once this has been done, all of the circuit breakers in the panel can be checked very quickly. In a typical instance, 77 breakers were checked in six hours, and the panel in which they were used was returned to service before the end of the shift.
As a result of these dramatically reduced testing times, it has been possible to implement a greatly improved MCCB testing programme at Diablo Canyon. In the past the testing regime meant that each vital MCCB was tested, on average, once every 20 years. With the new test programme, the average interval between tests has been reduced to just five years.
Overall, the increased attention to MCCBs and improved frequency of testing means that the out-of-tolerance (OOT) rate of breakers on test at the plant has fallen from 25% to just 1%. In addition, the detailed results now available are making it possible to concentrate attention on those breakers that are most likely to give OOT results.
