US NRC “Resolution of Generic Safety Issues: Issue 193: BWR ECCS Suction Concerns”

Resolution of Generic Safety Issues: Issue 193: BWR ECCS Suction Concerns Description ( NUREG-0933, Main Report with Supplements 1–34 )

(NUREG-0933: Footnote 1814
1814. Memorandum for S. Newberry from J. Grobe, “Potentially Generic Safety Issue”BWR ECCS Suction Concerns,” May 10, 2002. [ML021340802])

This issue was identified1814, 1815 by a Region III inspector and addressed the possible failure of low pressure emergency core cooling systems due to unanticipated, large quantities of entrained gas in the suction piping from suppression pools in BWR Mark I containments.
Three specific concerns were listed in the identifying document:1814

The case where offsite power is not available (the design basis) is somewhat more complicated. The random likelihood of a large LOCA occurring simultaneously with a LOOP is very small, and the probability of a LOOP subsequent to a LOCA is relatively low. The probability value used in the Peach Bottom PRA 1081 was 2 x 10-4 (mean). Combined with the 10-4/RY large LOCA-initiating event frequency in NUREG-1150,1081 the combined LOCA-LOOP event would have a frequency on the order of 10-7/RY or lower. However, a seismic event could cause both a LOOP and a LOCA. (Fire-initiated LOCAs are generally stuck-open SRVs and are not applicable to this issue.)
Such a seismically-induced combined LOCA-LOOP was included in the external events analysis PRA for the Peach Bottom plant.1081

Pump Failure Probability: The parameter of greatest significance for this issue is the probability of pump failure as a function of time after LOCA initiation. This probability can be broken down into two factors: (1) the probability of failure as a function of the volume percent of entrained air in the pump suction; and (2) the fraction of entrained air in the suppression pool volume as a function of time.

Bubble Rise Phenomena: The GE experimental work did not record the parameters of most interest for this issue, and thus it is necessary to use some more general knowledge of such phenomena. According to the BWR Fundamentals training manual, the height of the suppression pool air and liquid space (i.e., minor diameter of the torus) is typically about 29 feet. After being forcibly thrown up “to the ceiling” by the initial blowdown, the suppression pool water will fall back in about one second, held back only by air drag. However, entrained air bubbles in the pool will take much longer to rise to the surface, because of the viscosity of the water.

Consequence Estimate
For this issue, all of the sequences that result in severe core damage include failure of all four RHR pumps. These same pumps are also used for suppression pool cooling and for containment spray. Thus, each of these core damage sequences will also result in containment failure due to overpressure.

Defense-in-depth: The postulated effect of entrained gas bubbles is to defeat a major portion of the low pressure ECCS. Even if the low initiating event frequency results in a low frequency for most of the accident sequences, there is a policy question regarding the wisdom of allowing such a failure, i.e., what is the purpose of maintaining the first group of pumps if there is a high likelihood of failure for this group? This consideration is tempered by the fact that: (a) the estimates used in this screening analysis contain some conservatism, and it is really not known for certain that the first group will fail; and (b) this really applies only to the very large break LOCA, which will violently entrain air in the suppression pool, and the rest of the LOCA break spectrum may not be affected.

The analysis indicated some importance even for the base case, but rises significantly for a BWR with fast sequencing when offsite power is available, and also for a BWR in a high seismic area. However, the various estimates given above include some conservatism, and should be understood as an importance measure, not as a best estimate. That is, if the postulated mechanism is true, these are estimates of what the safety significance would be.
It is suggested that any technical assessment include some effort to address the various points of conservatism within this analysis:
This analysis assumed that non-condensible gas bubbles are uniformly mixed within the suppression pool. The actual situation, including stratification and how deeply bubbles will be driven into the pool, should be investigated.
The number of plants with fast sequencing should be investigated along with the number in high seismic zones.
The efficacy of ring headers and other pump suction piping configurations in isolating the pumps from suppression pool phenomena should be investigated.
The ability of pumps to withstand entrained air, particularly for short periods of time, should be investigated.
Based on the LERF estimates given above, work on the issue continued to the technical assessment stage.1824
Page Last Reviewed/Updated Thursday, March 29, 2012

One Response to “US NRC “Resolution of Generic Safety Issues: Issue 193: BWR ECCS Suction Concerns””

  1. CaptD Says:

    Kudos to the Region III inspector !

    As we learn more about how the Nuclear Industry and Utilities “run” their reactors we learn they are on many occasions “winging it,” which is how Fukushima happened!

Comments are closed.

%d bloggers like this: