Friday, July 29, 2016

Strange trip at Susquehanna nuclear power plant

Attached are pages from the Susquehanna Unit 2 technical specifications, two of the single-line electrical drawings, and Chapter 13 of the FSAR. I could not find drawing E-9 Sheet 71 which has the breakers within Motor Control Center 2B246. But drawing E-9 Sheet 20 is attached which shows the breakers for MCC 1B246. Assuming that Unit 1 is configured similar to Unit 2 in this regard, the components in that MCC should be identical or close. The event notification stated that the fault on MCC 2B246 disabled a drywell unit cooler, causing drywell pressure to increase. MCC 1B246 powers drywell unit coolers on Unit 1.

The attached pages address some of the questions posed:
1) Shift Technical Advisors are required to be on-shift to provide advise to the operators. I worked several years as an STA at Browns Ferry and the STAs at Grand Gulf reported to me when I was the Reactor Engineering Supervisor. The STAs worked 24/7, but they did not have to be in the control room at all times. Basically, STAs have to be summonable by the operators. That's typically defined as being available within 10 minutes of being called. The STAs don't have to be in the control room. But the STAs are not supposed to become "hands on." They are supposed to remain detached so as to be free to look around, gather dots, and connect them. If the STA was out in the plant trying to restore power or take steps to cope with the loss of power, the STA erred by straying from assigned responsibilities.

2) The tech specs say that the drywell pressure causes an automatic trip when pressure rises to 1.88 psig or less. The event report says that operators manually tripped the reactor as pressure neared 1.3 psig on the rise. I suspect that this is not the signal that any one bypassed because it's wicked hard to bypass this one.

3) The event report states that operators turned HPCI off after it had automatically initiated. The TMI operators turned off an automatically stated makeup system, which contributed much to that meltdown. NRC may be irked with the haste at which HPCI was turned off, but turning if off would be standard BWR procedure for such a situation.

4) Bad deed to avoid increased NRC oversight doesn't make sense. The unplanned scrams performance indicator shows some margin for another scram or two:
The NRC recently sent a Chilled Work Environment letter to TVA due to about half the operators feeling that management would retaliate against them for raising safety concerns. Seems that in November 2015, TVA sanctioned several operators. TVA contends that the sanctions were for misdeeds occurring about five weeks earlier. The operators contend the sanctions were in response to concerns the operators raised a day or two before the sanctions were levied. Since the NRC issued its CWE letter in March 2016 after asking around, it seems that NRC believed the operators' story instead of managements.

I relay this history because it may apply to the morale of the work force at Susquehanna. If the work force believes that the 10 terminations were for the wrong reason, it won't matter if they were for the right reason.

Besides 10 seems like way too many managers to mess up a single event. With 10 people you can play 5 on 5 basketball. It typically only takes 1 to 3 managers to mess up an event. They may have been wanting to trim 10 positions and found 7-9 draftees to complement the 1-3 volunteers.

No comments: