Resolution of Generic Safety Issues: Issue 63: Use of Equipment Not Classified as Essential to Safety in BWR Transient Analysis (Rev. 1) ( NUREG-0933, Main Report with Supplements 1–34 )
BWRs are required to be operated within set thermal limits to maintain the integrity of the fuel cladding during postulated events. One of the established thermal limits is the minimum critical power ratio (MCPR). The critical power ratio (CPR) is the ratio of the fuel bundle power at which boiling transition begins to the actual bundle power. By maintaining the CPR above a predetermined safety limit everywhere in the core, boiling transition and possible fuel failure can be precluded. Several postulated transients result in a reduction of CPR. In order to assure that the CPR safety limit is not violated, an operating limit must be set. The operating limit is based upon the CPR safety limit and the change in CPR from the most limiting or severe transient. By operating above the operating limit, a plant will not violate the CPR safety limit for any of the postulated transients. Since the CPR operating limit is related to core thermal output, a higher operating limit may result in restricted output. Therefore, it is advantageous to show the least change in CPR for all transients analyzed.
In the past, applicants have been required to assume failure of certain equipment and only take credit for the operation of other equipment in the analyses of transients.386 The combination of these assumptions has often dictated the most severe transient with respect to CPR. In 1981, the Reactor Systems Branch (RSB) of NRR expressed concerns with the credit given for equipment which is not classified as safety-related.385
The solution proposed for this issue would be to require that the analysis of transients only rely on equipment classified as safety-related. It is likely that this would result in increased costs to industry either due to penalties in derating plants or in new equipment to meet the safety requirements.
Based on the concerns385 expressed by RSB, PNL did an assessment64 of this issue and concentrated on those transient events which involve the high-water level trip and the turbine bypass system which are not classified as safety-related. Two approaches were used in analyzing this issue.
The first approach was to examine three major transients and evaluate changes in CPR due to failure of the high-water level trip (L8) and/or failure of the bypass valves (TBP) to open. The transients analyzed included: (1) turbine trip without bypass (100% power); (2) loss of one feedwater string (100% power); and (3) feedwater control failure (high - 50% power). Based on examining these transients with the Browns Ferry Simulator,64 it was concluded that: (1) the CPR did not exceed the fuel cladding integrity limit; and (2) in each case, other Reactor Protection System (RPS) signals provided a scram and in no case examined was the reactor vessel coolant inventory compromised due to failure of the non-safety grade equipment. Therefore, the potential for fuel damage and core-melt was considered negligible.
The second approach was a quantitative analysis to estimate an upper bound on potential risk reduction assuming an upgrading (to safety system requirements) of the high-water level trip and the turbine bypass system. Grand Gulf was used as the representative BWR.
The rationale for the use of the second approach to evaluate public risk was based on the premise that this fault combination comprises the majority of the public risk. While the probable failure result is a departure from nucleate boiling (DNB) which may result in rupture of the fuel cladding, the consequences are constrained by the limited amount and form of radioactivity which is released. Conversely, while the probability of failing to achieve subcriticality is much less than the probability of DNB, the consequences are so much greater that the risk to the public is dominated by this failure sequence.
T23C is the only affected Grand Gulf dominant accident sequence for this issue; T23 are those transients other than loss of offsite power (LOOP) that require emergency reactor shutdown and C is the failure to achieve subcriticality. T23 was redefined to include only those transients associated with failure of L8 trip and TBP. Redefining T23 involved identification of transient initiators associated with L8 trip and TBP. These initiators and associated mean total frequencies of occurrence taken from the EPRI ATWS reappraisal307 are as follows:
|Turbine trip with TBP valve failure||0.01/RY|
|Feedwater--increasing flow at power||0.16/RY|
|Loss of feedwater heater||0.04/RY|
The percentage contribution of these non-LOOP transients (0.21/RY) to the mean total frequency of all BWR non-LOOP transients (8.78/RY) taken from the EPRI report307 is [(0.21/RY)/(8.78/RY)] x 100% = 2.4% where:
|Total BWR transients||8.90/RY|
The frequency of the above transients normalized to Grand Gulf (7/RY)64 produces the redefined value of T23= (0.024)(7/RY) = 0.168/RY. The failure to achieve subcriticality (C) is defined by C = (RPLS + CR)(MANSD) where:
|RPLS||= 1.9 x 10-6, the failure rate of the Reactor Protection Logic System|
|CR||= 5.8 x 10-6, the failure rate of the Control Rod Drive System not inserting given a trip signal|
|MANSD||= the failure rate of the recirculation pumps failingto trip, or the operator fails to manually insertthe control rods, or the operator fails to initiatethe Standby Liquid Control System (SLCS). The MANSD failure is dominated by the SLCS failure rate, 0.1.|
Hence, for Grand Gulf,
Thus, for Grand Gulf, the non-LOOP transient accident sequence is given by:
The base case, affected core-melt frequency is given by F = 1.29 x 10-7/RY.
For the adjusted case, it was assumed that improving the high water level trip and the turbine bypass system would increase the availability of the RPLS by a factor of 10 (at most); this would produce extremely conservative results.
The adjusted-case, affected core-melt frequency is given by F* = 1.01 x 10-7/RY. The reduction in core-melt frequency (F) is (F - F*) = (1.29 x 10-7/RY) -(1.01 x 10-7/RY) = 2.8 x 10-8/RY.
The base case, affected public risk is given by W = (1.29 x 10-7/RY)(7.1 x 106 man-rem) = 0.92 man-rem/RY, where the dose in man-rem is that for a BWR-2 type release, as defined in Section 3.2 and Appendix D of NUREG/CR-2800.64
The adjusted-case, affected public risk is given W* = (1.01 x 10-7/RY)(7.1 x 106 man-rem) = 0.72 man-rem/RY. The reduction in public risk per plant (W) is (W -W*). Therefore, W = (0.92 - 0.72) man-rem/RY = 0.20 man-rem/RY.
Therefore, the total public risk reduction is estimated to be (44 BWRs)(27.4 yr) (0.20 man-rem/RY) = 240 man-rem.
Industry Cost: A rough estimate of costs was made based on an upgrade (to safety system requirements) trip and the turbine bypass system. It was estimated that equipment costs would be on the order of $500,000 for each system. The engineering and installation costs would be double this cost. Therefore, industry costs would be on the order of $3M/plant. Annual operation and maintenance was estimated at about 2 man-weeks/year; this amounts to ($2,270/man-week)(2 man-weeks/year)(27 years/plant) or $150,000/plant. The total industry cost was therefore estimated at $3.1M/plant or (44 plants) x ($3.1M/plant) = $136.4M for the affected plants.
NRC Cost: NRC costs were considered negligible when compared to industry costs and were not estimated.
Based on an estimated risk reduction of 240 man-rem and a cost of $136.4M, the value/impact score is given by:
The following items were also considered to have an effect on the reliability of non-safety grade systems and, depending on their outcome, could impact this issue.
(1) The high-water level trip was investigated as part of USI A-47, "Safety Implications of Control Systems."
(2) Existing licensing practice has been to require TS for the surveillance of the high-water level trip and the turbine bypass system.
(3) On future BWRs (BWR-6), the high-water level trip will be part of the RPS.
(4) In the RES Accident Management research program, existing and alternate strategies for arresting core damage and/or radioactivity release during a severe accident are to be evaluated. Credit is to be considered for all equipment: safety-grade; non-safety-grade; and even offsite equipment (i.e., fire trucks).
Based on the calculated risk reduction, value/impact score, and the other considerations, this issue was DROPPED from further consideration.