Resolution of Generic Safety Issues: Item D-1: Advisability of a Seismic Scram (Rev. 1) ( NUREG-0933, Main Report with Supplements 1–34 )
This issue is described in NUREG-04713 and was raised by the ACRS who recommended that studies be made of the technique for seismic scram and the potential safety advantages and disadvantages of prompt reactor scram, in the event of strong seismic motion.
A seismic scram could be designed to scram the reactor upon the occurrence of a seismic event, before turbine trip or other conditions resulting from the seismic disturbance could cause a scram. The earlier scram could give a lead time between 5 to 20 seconds. The lead time could provide resulting benefits such as reduced loads during the seismic event and, therefore, less burden on the plant systems. It may also reduce the likelihood of a LOCA or severe transient after a seismic event.
An automatic seismic trip system could be designed to utilize existing state-of-the-art seismic instrumentation. The system could be designed to use coincidence logic to reduce the frequency of spurious trips. A "high level" trip could be set based on some percent of the SSE (usually chosen as greater than 60% of the SSE level) and could be designed to minimize spurious trips due to after-shock and low acceleration earthquakes. A "low level" trip would be set to activate on the compressional waves (P waves) when this first arrival caused displacement or acceleration greater than the calculated maximum allowable P wave for an OBE.
The "low level"and "high level" trips assumed were described in UCRL-521561717 and Supplement 2 to NUREG/CR-2800,64 respectively. To select an upper bound on the potential for risk reduction from resolution of the issue, the analysis in NUREG/CR-280064 was chosen because it showed the largest potential advantage and was based on a "high level" trip point.
To identify the possible advantages of a seismic scram system, possible transient and accidents that could lead to core-melt were considered. It was then assumed that the early seismic scram precluded waiting for a later trip and it would, therefore, reduce transient pressure and loads and the heat generation rate in the core. It was then assumed that this would decrease the burden on the reactor's safety systems, e.g., safety/relief valves and turbine-driven pumps. It was also assumed that, in the event of a LOCA, an earlier trip could reduce the fuel rod temperature transient and the containment vessel pressure. Less fluid would be lost during the blowdown phase before the SIS operating pressure is reached. To quantify these benefits, these assumptions were converted to a frequency reduction for certain event initiators.
The assumption was made that the prime benefit of an earlier seismic scram is in reducing the frequency of a seismically-induced transient or LOCA initiator as a result of reduction in the afore-mentioned stresses. Therefore, the core-melt frequency due to an earthquake (where the earthquake's only effect is to induce a transient or LOCA initiator) was estimated. This estimate64 was based on an LLNL Report (UCRL-53037) which provided estimates of earthquake frequency at the Zion site and the conditional probabilities of transient or LOCA initiators, given an earthquake.
Earthquake Frequency at Zion Site
(During Plant Operation)
|Earthquake Level (SSE)||Frequency/RY|
|0.4 - 0.6||8.4 x 10-4|
|0.6 - 0.9||4.5 x 10-4|
|0.9 - 1.8||2.5 x 10-4|
|1.8 - 2.5||1.3 x 10-5|
|>2.5||2.2 x 10-6|
Conditional Probability of Transient or LOCA Initiator
(Given an Earthquake)
|Earthquake Level (SSE)||Conditional Probability|
|LOCA||T1 Transient||T2 Transient|
|0.6 - 0.9||6.0 x 10-5||0.360||0.24|
|0.9 - 1.8||2.5 x 10-3||0.400||0.59|
|1.8 - 2.5||5.2 x 10-2||0.019||0.93|
|>2.5||2.6 x 10-1||0||0.74|
Since no data were provided on the conditional probability of transient or LOCA initiators for 0.4 to 0.6 SSE, no contribution was estimated for this SSE interval. Based on these tables, the following frequencies of seismically-induced transient or LOCA initiators were calculated:
LOCA = (4.5 x 10-4/RY)(6 x 10-5) + (2.5 x 10-4/RY)(2.5 x 10-3) + (1.3 x 10-5/RY)(5.2 x 10-2) + (2.2 x 10-6/RY)(2.6 x 10-1)
= 1.9 x 10-6/RY
T1 = (4.5 x 10-4/RY)(0.36) + (2.5 x 10-3/RY)(0.40)+ (1.3 x 10-5/RY)(0.019)
= 2.7 x 10-4/RY
T2 = (4.5 x 10-4/RY)(0.24) + (2.5 x 10-4/RY)(0.59) + (1.3 x 10-5/RY)(0.93) + (2.2 x 10-6/RY)(0.74)
= 2.6 x 10-4/RY
The accident sequences were defined based on the Oconee-3 and Grand Gulf-1 risk equations in NUREG/CR-2800,64 and the affected release category and core-melt frequencies were:
Oconee-3 Grand Gulf-1
(PWR-1)e = 1.4 x 10-9/RY (BWR-1)e = 8.5 x 10-11/RY
(PWR-2)e = 3.3 x 10-9/RY (BWR-2)e = 1.7 x 10-8/RY
(PWR-3)e = 4.3 x 10-8/RY (BWR-3)e = 1.4 x 10-9/RY
(PWR-4)e = 9.3 x 10-11/RY (BWR-4)e = 1.4 x 10-9/RY
(PWR-5)e = 1.2 x 10-9/RY
(PWR-6)e = 8.2 x 10-9/RY
(PWR-7)e = 1.3 x 10-7/RY
It was then assumed that a high-level scram could reduce the frequencies of all these sequences, with the exception of those for which failure to scram is an inherent part of the sequence, by reducing the frequencies of their seismically-induced initiators. Removing the contribution from the two failure-to-scram sequences from the totals for all these sequences, the following maximum frequency reductions were calculated based on total elimination of seismically-induced transients and LOCAs.
Oconee-3: F = 1.8 x 10-7/RY
Grand Gulf-1: F = 2.0 x 10-8/RY
To calculate the risk reduction (W), the dose factors from Appendix D of NUREG/CR-280064 were used with the following results:
Oconee-3: W = 0.26 man-rem/RY
Grand Gulf-1: W = 0.13 man-rem/RY
The total public risk reduction was then calculated as follows:
W = (0.26 man-rem/RY)(85 PWRs) (28.8 years) + (0.13 man-rem/RY)(44 BWRs)(27.4 years)
= 790 man-rem
Industry Cost: It was assumed that 10 man-weeks of labor would be required to install and test the seismic scram system and that it could be performed during scheduled outages. The labor was assumed to cost $2,270/man-week and equipment was assumed to cost $150,000/plant. Therefore, the total implementation cost for 129 plants was estimated to be about $22M. The total cost for operation and maintenance was estimated to be $33M, based on 4 man-weeks/RY. Thus, the total industry cost was estimated to be $(22 + 33)M or $55M.
NRC Cost: NRC costs were assumed to be negligible compared to industry costs.
Total Cost: The total industry and NRC cost associated with the possible solution was estimated to be $55M.
Based on an estimated public risk reduction of 790 man-rem and a cost of $55M for a possible solution, the value/impact score was given by:
(1) At the time of the initial evaluation of this issue in November 1983, San Onofre and Diablo Canyon had seismic scram systems installed.
(2) Another advantage suggested in Supplement 2 to NUREG/CR-280064 was a slight addition of time for some core-melt scenarios. This was not included in the above analysis because it was not believed to be a significant amount of time in relation to preventing core-melt.
(3) Potential disadvantages of a seismic scram are spurious trips due to maintenance or test, and possible trips when a trip would not be necessary, i.e., the plant would ride through an event. Through careful design, these could be minimized.
(4) A number of critical comments486, 487 were made regarding the LLNL study (UCRL-53037) that was used in the NUREG/CR-280064 analysis. These comments indicated that the report was too optimistic regarding the benefits of a seismic scram.
This issue was given a low priority ranking (see Appendix C) in November 1983. In NUREG/CR-5382,1563 it was concluded that consideration of a 20-year license renewal period did not change the priority of the issue. Further prioritization, using the conversion factor of $2,000/man-rem approved1689 by the Commission in September 1995, resulted in an impact/value ratio (R) of $71,428/man-rem, which placed the issue in the DROP category.