United States Nuclear Regulatory Commission - Protecting People and the Environment

Resolution of Generic Safety Issues: Issue 132: RHR System Inside Containment (Rev. 1) ( NUREG-0933, Main Report with Supplements 1–34 )

DESCRIPTION

Historical Background

This issue was identified1454 by NRR as a result of concern for the potential loss of the RHR system under a harsh containment environment. The RHR system is normally located outside the containment and is not required to be qualified for a harsh environment. However, in a small group of W plants ( Surry 1 and 2, Beaver Valley 1 and 2, North Anna 1 and 2, and South Texas 1), the RHR system is located inside the containment. Should a harsh environment be created inside the containment, the RHR system might be rendered inoperable for long-term decay heat removal because of environmentally unqualified pumps and valves. A harsh environment inside the containment could result from a LOCA, a steam line break inside the containment, feed-and-bleed operation, or from inadvertent actuation of the containment spray system. Accidents and transients causing potential unavailability of the RHR system were considered in this evaluation.

During the first phase of normal plant cooldown, the temperature of the RCS is reduced by transferring heat from the RCS to the steam and power conversion system. During the second phase of cooldown, the RHR system transfers heat from the RCS to the CCW system, to reduce the temperature of the reactor coolant to the cold shutdown temperature at a controlled rate, and maintains this temperature until the plant is re-started. The RHR system also can provide backup long-term cooling following a LOCA and is used to transfer refueling water between the refueling cavity and the refueling water storage tank (RWST), before and after refueling operations.

Safety Significance

The RHR system is one of the systems available for long-term decay heat removal. For Surry, the systems which provide long-term cooling include the Inside Spray Recirculation System, or Outside Containment Spray Recirculation System coupled with the Low Pressure Recirculation System, and the RHR System. Any one of these systems will provide sufficient capacity for long-term cooling. Loss of the RHR system as a result of a harsh environment inside the containment would result in one less pathway for long-term decay heat removal and, therefore, an increase in CDF and associated public risk.

Possible Solutions

The possible alternative solutions considered for this issue included:

(1) Analyze the fragility of existing RHR equipment to ensure that the equipment will be operable under high temperature and pressure inside the containment resulting from the accidents identified above. Results of the analyses may identify certain equipment needed to be upgraded for environmental qualification;

(2) Install a crosstie to allow flow of the high pressure safety injection (HPI) system from the other unit; or

(3) Replace all existing equipment with environmentally-qualified equipment.

The cost for replacement of all existing equipment with environmentally-qualified equipment was expected to be much higher than that for Alternatives 1 and 2. Therefore, only Alternatives 1 and 2 were analyzed because Alternative 3 would be the least cost-effective approach and would be the most unfavorable option in terms of value/impact assessment.

PRIORITY DETERMINATION

Frequency Estimate

The approach taken to estimate the changes in CDF and public risk associated with the solutions was first to determine the RHR system failure on demand probabilities. Based on the Surry PRA,1318 the accident sequences which are shown to involve credit for the RHR system for long-term cooling include a small LOCA (½" to 2"), a very small LOCA (<½"), loss of main feedwater, and turbine trip. Loss of main feedwater and turbine trip could involve operation of feed-and-bleed cooling or a postulated failure of a pressurizer PORV/SRV to reseat that could lead to a sequence similar to a small LOCA. Based on these sequences and their associated CDFs in the Surry PRA, the CDFs were calculated64 for the base case and adjusted case of Alternatives 1 and 2.

Alternative 1: The base case consists of the scenario in which the RHR system is not environmentally-qualified and is therefore assumed to be unavailable following the sequences identified above except a very small LOCA. This assumption is consistent with the Surry PRA. Using this logic, the base case failure probability (W3) for the RHR system to cool the reactor is 1. The sensitivity of this assumed failure probability of the RHR system is discussed below. For the adjusted case, it was assumed that the RHR system is environmentally-qualified and its failure probability was the same as that given in the Surry PRA: 8.9 x 103.

Based on the failure probabilities of the RHR system, the CDFs for the base case and adjusted case were then calculated by using the accident sequences identified above. The resulting CDFs were 2.6 x 10-6/RY and 2.3 x 10-8/RY for the base case and the adjusted case, respectively.

Alternative 2: The adjusted case for Alternative 2 included the installation of a crosstie between the HPI trains of Units 1 and 2, similar to that installed at Surry. The CDFs for Solution 2 were calculated to be 3 x 10-6/RY and 2.8 x 10-8/RY for the base case and the adjusted case, respectively.

The CDF for the base case of Alternative 2 is different from that for Alternative 1 due to an additional accident sequence (very small LOCA with failure to provide RHR cooling) included in Alternative 2.

Surry: For Surry, the CDFs before the RHR equipment is qualified as operable (base case) and after being qualified as operable by implementing Alternative 1 (adjusted case) were 2.2 x 10-8/RY and 3.9 x 10-10/RY, respectively. The CDFs were much smaller than that resulting from implementing Alternative 1 because Surry had previously installed the crosstie for the HPI system. Such a crosstie was not in place at the other plants considered in this issue.

Consequence Estimate

The next step was to calculate the base case and adjusted case public risk. This was accomplished by multiplying the base case and adjusted case CDFs by appropriate containment failure probabilities and corresponding public dose. For this analysis, the containment failure modes, release categories, and public dose consequences were taken from NUREG/CR-2800.64 Results of this calculation are shown in Tables 3.132-1 and 3.132-2.

The reduction in public risk is the difference between the calculated dose consequence for the base case and the adjusted case.

Alternative Reduction in Public Risk
1 7 man-rem/RY or 220 man-rem/reactor
2 8 man-rem/RY or 250 man-rem/reactor

The total reduction in public risk per reactor was based on the average remaining lifetime of 31 years for the affected plants. The reduction in public risk for Surry was calculated to be 1.5 man-rem/reactor, based on the remaining lifetime of 26 years.

Cost Estimate

The cost estimates for the possible solutions were based on the guidelines provided in NUREG/CR-4627.961 The industry cost for implementation included resources necessary to environmentally qualify RHR systems by analysis and minor equipment modification, or to install the crosstie. The NRC costs for SIR implementation include regulatory review and approval for the licensee proposed resolution. For Alternative 1, the total cost was estimated to be $0.4M/reactor. For plants without a crosstie, the total cost for installation of a crosstie was estimated to be $2.1M/reactor.64

Value/Impact Assessment

Alternative 1: Based on a total potential risk reduction of 220 man-rem/reactor and a cost of $0.4M/reactor, the value/impact score was given by:

Alternative 1 value/impact assessment score expression consisting of: S= 220 man-rem/reactor over $0.4M/reactor equals 550 man-rem/$M

Alternative 2: Based on a total potential risk reduction of 250 man-rem/reactor and a cost of $2.1M/reactor, the value/impact score was given by:

TABLE 3.132-1
Alternative 1 : Perform RHR Fragility Analysis

Accident Sequence Frequency(/RY) Containment Failure Mode Containment Failure Probability Release Category Consequence (man-rem/RY) Public Risk (man-rem/RY)
Base Case
2.6 x 10-6 Greek capital letter gamma
Greek small letter beta
Greek small letter epsilon
0.5
0.0073
0.5
PWR-3
PWR-5
PWR-7
5.4 x 106
1.0 x 106
2.3 x 103
7.1 x 100
1.9 x 10-2
3.0 x 10-3

TOTAL: 7.1 x 100

Adjusted Case
2.3 x 10-8 Greek capital letter gamma
Greek small letter beta
Greek small letter epsilon
0.5
0.0073
0.5
PWR-3
PWR-5
PWR-7
5.4 x 106
1.0 x 106
2.3 x 106
6.2 x 10-2
1.7 x 10-4
2.6 x 10-5

TOTAL: 6.2 x 10-2

TABLE 3.132-2
Alternative 2: Install A Cross-Tie

Accident Sequence Frequency/RY Containment Failure Mode Containment Failure Probability Release Category Consequence (man-rem) Public Risk (man-rem/RY)
Base Case


3.0 x 10-6
Greek capital letter gamma
Greek small letter beta
Greek small letter epsilon
0.5
0.0073
0.5
PWR-3
PWR-5
PWR-7
5.4 x 106
1.0 x 106
2.3 x 103
8.1 x 100
2.2 x 10-2
3.4 x 10-3

TOTAL: 8.1 x 10-0

Adjusted Case
2.8 x 10-8 Greek capital letter gamma
Greek small letter beta
Greek small letter epsilon
0.5
0.0073
0.5
PWR-3
PWR-5
PWR-7
5.4 x 106
1.0 x 106
2.3 x 103
7.4 x 10-2
2.9 x 10-4
3.2 x 10-5

TOTAL: 7.4 x 10-2

Alternative 2 value/impact assessment score expression consisting of: S= 250 man-rem/reactor over $2.1M/reactor equals 120 man-rem/$M

Surry: Based on a potential risk reduction of 1.5 man-rem and a cost of $0.3M, the value/impact score was given by:

Surry value/impact assessment score expression consisting of: S= 1.5 man-rem over $0.3M equals 5 man-rem/$M

Other Considerations

(1) The uncertainties in the analysis involve whether the Surry PRA can be extrapolated for the group of plants under consideration. Surry is a 3-loop W plant and Beaver Valley 1 and 2 and North Anna 1 and 2 are also 3-loop W plants; however, South Texas is a 4-loop W plant. Differences in the primary system were not expected to have any significant impact on this evaluation.

(2) A sensitivity analysis for the failure probability of non-qualified RHR equipment was performed, assuming that the failure probability was reduced from 1 to 0.1. Results of the analysis shows the following:

Alternative Public Risk Reduction Value/Impact Score
1 18 man-rem/reactor 45 man-rem/$M
2 49 man-rem/reactor 23 man-rem/$M

These value/impact scores, combined with the reduction in public risk, would place this issue well in the low priority category.

(3) The failure probability of 1 for the RHR system represented an upper bound value and was considered unrealistic for the following reasons:

(a) The cut sets were evaluated for the RHR system of Surry by the use of the IRRAS Code.1455 Results of the evaluation indicated that loss of RHR due to failure of either one of the two motor-operated RHR suction isolation valves contributed about 68% of the failure probability of the RHR system. These two MOVs are considered primary system isolation valves and designed for an operating temperature of 650F and a pressure of 2,485 psi. The control circuitry for these valves is located outside containment in the auxiliary building. The design temperature and pressure and the location of the valve control circuitry make the MOVs less susceptible to fail under the environmental conditions following the accidents involved.

(b) Following these accident sequences, the auxiliary feedwater system and containment fan coolers will be removing heat from the containment prior to operation of the RHR system. Consequently, the temperature and pressure inside the containment will be lowered substantially and the RHR system would not be subject to an environment such that failure would be a certainty.

(c) The RHR system is located in the lower elevation and in the periphery of the containment. Therefore, the RHR equipment is not expected to face direct impingement of high pressure water/steam released from the primary system. Further, thermal stratification inside the containment will result in lower temperature in the low elevations. The RHR system will be operating under less severe environmental conditions. Again, inevitable failure of the RHR system is not a certainty.

Based on the considerations above, the failure probability of 0.1, which is an order of magnitude difference between the absolute failure and the normal failure probability for the RHR system, represented a best estimate value.

CONCLUSION

The value/impact scores for both Alternatives 1 and 2, combined with the reduction in public risk, would place this issue in the medium priority category (See Appendix C). For Surry, the value/impact score combined with the reduction in public risk would result in a drop ranking. Further, the calculated cost/benefit for plants without an HPI cross-tie was well above the guideline of $1,000/man-rem ($22,000/man-rem and $43,000/man-rem for Alternatives 1 and 2, respectively).

Even if the upper bound value of RHR failure probability were assumed (thereby producing a medium priority), the cost/benefit would be above the $l,000/man-rem ($1,800/man-rem and $8,400/man-rem for Alternatives 1 and 2, respectively). The estimate of CDF reduction for plants without an HPI crosstie was on the order of about 3 x 10-6/RY, using the upper bound value for all accident sequences involved. This estimate was not considered a substantial reduction as required by the 10 CFR 50.109 backfit rule. Thus, this issue was DROPPED from further pursuit (See Appendix C). In an RES evaluation,1564 it was concluded that consideration of a 20-year license renewal period did not change the priority of the issue.

Page Last Reviewed/Updated Thursday, March 29, 2012