Resolution of Generic Safety Issues: Item B-5: Ductility of Two-Way Slabs and Shells and Buckling Behavior of Steel Containments (Rev. 1) ( NUREG-0933, Main Report with Supplements 1–34 )
This item has been divided into two parts which have been evaluated separately.
PART I - Ductility of Two-Way Slabs and Shells
This issue was identified in NUREG-04713 and involved concern over the lack of information related to the behavior of two-way reinforced concrete slabs loaded dynamically in biaxial membrane tension (resulting from in-plane loads), flexure,
and shear. A task is defined which involves developing a more dependable and realistic procedure for evaluating the design adequacy of Category 1 reinforced concrete slabs subject to a postulated LOCA or high-energy line break (HELB).
If structures (concrete slabs) were to fail (floor collapse or wall collapse) due to loading caused by a LOCA or HELB, there would be a possibility that other portions of the reactor coolant system or safety-related systems could be damaged. Such loads would be caused by very concentrated high-energy sources causing direct impact on the structures of concern. The damage could lead to an accident sequence resulting in the release of radioactivity to the environment.
A task was defined to determine with sufficient accuracy the influence of biaxial membrane tension in the plane of the slab on the resistance function and the permissible ductility ratio of two-way slabs loaded in flexure and shear. The end product of the task was to be the development of a simplified practical method which could be used for design and analysis of a slab subjected to the above loading.
SEB has concluded215 that there is sufficient information pertaining to the design of two-way slabs subjected to dynamic loads and biaxial tension to enable a reasonably accurate analysis. Based on this information, we conclude that a solution has been identified for this part of the overall issue.
PART II - Buckling Behavior of Steel Containments
This issue is identified in NUREG-04713 and involves concern over the lack ofa uniform, well-defined approach for design evaluation of steel containments. The structural design of a steel containment vessel subjected to unsymmetrical dynamic loadings may be governed by the instability of the shell. For this type of loading, the current design verification methods, analytical techniques, and the acceptance criteria may not be as comprehensive as they should be. Section III of the ASME Code does not provide detailed guidance on the treatment of buckling of steel containment vessels for such loading conditions. Moreover, this Code does not address the asymmetrical nature of the containment shell due to the presence of equipment hatch openings and other penetrations. Regulatory Guide 1.57478 recommends a minimum factor of safety of two against buckling for the worst loading condition provided a detailed rigorous analysis, considering inelastic behavior, is performed. On the other hand, the 1977 Summer Addendum of the ASME Code permits three alternate methods but requires a factor of safety between 2 and 3 against buckling, depending upon the applicable service limits.
If steel containment shells were to fail due to loading which may cause buckling, one of the plant's levels of defense would be lost and could result in release of radioactivity to the environment. The loading would have to be due to a high-energy source. A large LOCA or HELB near the containment wall could possibly provide such a load.
At present, SEB has developed and is using a set of interim criteria145 for evaluating steel containment buckling for plants undergoing operating license review. A longer term project with RES is also underway.
The events for which containment may be impacted by a high enough energy source would be a large LOCA or a HELB. A small fraction of these would occur close enough to the containment wall to have the potential to rupture the barrier. If the containment is not adequately designed, a failure could occur.
The total event frequency is a combination of the above events. The frequency of a large LOCA or HELB can be compared to the frequency of a PWR-9 event or a BWR-5 event.16 Therefore, the frequencies are assumed to be 4 x 10-4/RY for a PWR-9 and 1 x 10-5/RY for a BWR-5.
Given the LOCA or HELB, then the probability of the large LOCA or HELB occurring close enough to containment wall and the break causing buckling is 0.1, based on conservative analysis (for priority determination) that assumes there would be around 10% of all LOCAs with enough energy and close enough to the wall to cause buckling. Combining event frequency and conditional probabilities,
For PWRs, F = (4 x 10-4/RY)(0.1) = 4 x 10-5/RY
For BWRs, F = (1 x 10-4/RY)(0.1) = 1 x 10-5/RY.
We assume that the containment fails and that the ECCS and other levels of defense in depth would be available. It would be possible that recirculation in a PWR may be affected (i.e., boil off to atmosphere); however, it would be expected that sufficient sump water would be collected for recirculation.
Therefore, we chose the consequences of these events to be comparable to a PWR-8 event and a BWR-4 event which involve significant damage and containment leakage.
Assuming midwest typical meteorology and a uniform population density of 340 people per square mile,
PWR-8 Consequence = 75,000 man-rem
BWR-4 Consequence = 610,000 man-rem.
Assume that structural reinforcing would be required for some containments and all steel containments are affected at a cost of $300,000 per plant. There are about 33 steel containments (existing and proposed). Assuming 2 staff-years of NRC effort to develop crtieria and 1 staff-week/plant to review modification and prepare SERs, the NRC manpower cost is $300,000. Assuming $300,000 for technical assistance, the total NRC cost is $0.6M. Therefore, the total cost is $[(33)(0.3) + 0.6]M = $10.5M.
Based on a risk reduction of 810 man-rem for 9 PWRs and a risk reduction of 4,392 man-rem for 24 BWRs, the total risk reduction for this issue is 5,200 man-rem. Therefore, the value/impact score is given by
(1) The severity of the safety issue, if one exists, is unknown and therefore the cost to industry of fixing the "problem" is also an unknown because no fix is presently proposed.
(2) Present staff requirements are contained in an interim position145 which is being used for NTOLs.
The scope and significance of this issue is not well-defined because the issue deals more with a potential problem than with a problem that is occurring or is even predicted to occur. Since steel containments are designed to available standards requirements, this issue is not as significant a problem as if no criteria existed for the design.
Because the safety significance of this issue was unclear with the current state of knowledge, it was recommended that the NRC continue to investigate, better define, and possibly resolve this issue. Since this issue is directed towards assessing the design adequacy of safety-related structures (i.e., containment), it was believed to have an inherent importance that justified further consideration. The results of the investigation could then be used to assess the potential safety benefits and costs that would be obtained if new design requirements for steel containments were implemented. Thus, this issue was determined to be of medium priority with a large uncertainty.
RES efforts in resolving this issue resulted in a proposed revision to SRP11 Section 3.8.2 that would be applicable to CP and OL applications filed after the effective date of the SRP Section revision. Operating plants were not affected by the proposed SRP revision because there was a general staff consensus that existing steel containments had adequate design conservatism regarding buckling. The proposed SRP revision was a formal promulgation of the changed staff review practices since the first SRP had been published in 1975 and added guidance for the review of asymmetric containment designs. It included an interim set of criteria for evaluating steel containment buckling that had been developed several years earlier by the former Structural Engineering Branch of NRR and had been applied to plants undergoing operating license reviews.
During the review process, the Structural and Geosciences Branch of NRR identified two concerns with the proposed resolution: (1) although the proposed revision to SRP11 Section 3.8.2 reflected NRR practice on the most recent licensing reviews, NRR expressed the concern that it contained technical requirements that were overly conservative; and (2) there was a general consensus that existing plants with steel shell containments had acceptable margins regarding buckling.1106 However, there was no readily available documentation to show this.
In March 1988, the Structural and Geosciences Branch of NRR issued a memorandum that: (1) summarized NRR's concern with the proposed revision to the SRP Section; (2) provided an evaluation that concluded that existing steel containments had adequate margins against buckling; and (3) stated that it was NRR's judgment that the issue of steel containment buckling had very little safety impact and was not worth pursuing further, considering the staff resource constraints. Thus, the issue was RESOLVED and no new requirements were established.1107