ML20207J779

Safety Evaluation Supporting Util 840813 Response Re Dynamic Loading of Cracked Fuel Element

ML20207J779
Person / Time
Site:	Fort Saint Vrain
Issue date:	12/30/1986
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20207J773	List: ML20207J769 ML20207J779 ML20207J793
References
TAC-49055, NUDOCS 8701080587
Download: ML20207J779 (3)

Text

.

[ ~go UNITED STATES 8 n NUCLEAR REGULATORY COMMISSION

$ $ WASHINGTON, D. C. 20555 ty

.....f Enclosure 1 SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATING TO CRACKED FUEL ELEMENT DYNAMIC LOADING PUBLIC SERVICE COMPANY OF COLORADO FORT ST. VRAIN NUCLEAR GENERATING STATION DOCKET N0. 50-267

1.0 INTRODUCTION

In 1982, during fuel reloading at Fort St. Vrain Nuclear Generating Station, two of the Segment 2 fuel elements (1-2415 and 1-0172) were found to have vertical cracks in their "B" faces. Public Service Company of Colorado (PSC), the licensee, performed studies to determine the cause and significance of these cracked fuel elements. Los Alamos National Laboratory (LANL) was under contract to the NRC staff in the review of this issue. During the review, LANL raised a concern regarding the effects on the structural integrity of a cracked fuel element due to dynamic loading from an earthquake. The safety significance of this issue is the possibility of a failed fuel element interfering with the flow of coolant through a fuel element column.

PSC provided a response to the LANL concern in PSC letter dated August 13, 1984 (P-84275) from 0. R. Lee to E. H. Johnson. LANL provided a review in a Technical Evaluation Report, " Fort St. Vrain Fuel Element Dynamic Response." The staff has reviewed the licensee's submittals and LANL's report. This safety evaluation provides the staff's conclusions on the issue of the dynamic loading of cracked fuel elements.

2.0 EVALUATION As part of its response to this issue, PSC presented the results of calculations in which the analyses of the cracked fuel element under thermal and irradiation-induced stresses were perturbed by analytically imposing a lateral load on the fuel element at the time of peak calculated stress. In these calculations, from zero to five cracked webs were assumed to exist in the fuel element. The pattern of these assumed cracks was similar to that observed in the actual fuel elements.

8701080587 861230 PDR ADOCK 05000267 3 PDR

. In order to determine the dynamic load to be imposed on the fuel elements, the design basis earthquake (DBE) was used. The DBE for Fort St. Vrain has a peak horizontal ground acceleration of 0.1 g. PSC states that the maximum seismic load on a fuel element during a DBE has been determined to be 1,500 pounds. This load was determined from the weights of fuel and reflector elements that can act on a single element during an earthquake, and from the 0.26 g lateral acceleration expected in the core as a result of the DBE.

The results of calculations, which assume a 1,500 pound lateral load, show that the largest calculated increase in the peak in-plane stress /

strength ratio of the cracked fuel element is 0.02. This increment appears to be a minor perturbation of the stresses caused by the thermal and irradiation-induced strains, and should not cause structural failure of the fuel element.

The calculations involved the imposition of static loads on the cracked fuel element analytical model to simulate seismic loads. PSC states that experimental evaluations of fuel ele.nent seismic strengh have shown that fuel element performance under both static and dynamic loading conditions is essentially the same for relative impact velocities up to 120 inches per second. Relative impact velocities in the core from a DBE are less than this value. PSC also states that the magnitudes of static and dynamic loads required to produce failure are the same, about 70,000 pounds.

This is much larger than the 1500 pound DBE load for Fort St. Vrain.

As part of this review, LANL performed a numerical model study of a graphite fuel element / reflector element stack on a core support system.

The numerical model was excited at the core support block with the north-south horizontal component of the 1940 El Centro earthquake record scaled to a peak acceleration of 0.2 g. As a result of this study, LANL estimated about 10 impacts with a peak velocity of about 8 inches per second on a fuel element from a DBE. They refer to the results of other studies that show that 210 impacts at relative velocities of 58 inches per second are required to initiate failure in an element being hit by an adjacent element. They conclude that the conditions necessary for fuel element failure are far in excess of that calculated for fuel elements under a DBE. LANL also concluded that calculations show that there is a large margin on uncracked fuel element strength during a DBE and this margin is not significantly reduced even when a crack is present in the fuel block. The staff is in agreement with these conclusions.

The primary safety importance of the structural failure of a fuel element is the possible disruption of core coolant flow. Among the materials the staff has reviewed on the consequences of coolant flow blockage are PSC letter P-78146 from J. K. Fuller to William Gammill and PSC letter P-84104 from Don W. Waremburg to John T. Collins.

The postulated failure of a fuel element could cause blockage of coolant .

flow in the fuel element column. Complete blockage of the coolant flow to a column during operation would result in a rapid increase in column

. temperature over several minutes. PSC estimates indicate that sufficiently high temperatures would be reached to fail the fuel particle coatings and release fission products contained in the column to the primary system. The problem would be detected by the increase in coolant activity. For a bounding condition, maximum gas temperature is estimated to be on the order of 6,000 degrees Fahrenheit and a complete sublimation

, of the graphite, fuel and fission products in the overheated column could occur. Under conditions of increased coolant activity the operators would trip the reactor and proceed with heat removal. The vaporized mixture would recondense primarily on the core support floor, the steam generators and the reactor vessel interior cavity as the temperature decreased. Even assuming the instantaneous release of the fission products in the blocked column to the primary coolant, PSC calculated that the 180-day doses at the boundary of the low population zone would be much less than tho~se calculated for the Design Bases Accident No..I during the operating license stage of review. This is primarily because the fission products contained in one fuel column are about one half of one percent of the total core inventory, a small fraction of the total reactor core.

3.0 CONCLUSION

S The staff concludes that there is a very low likelihood of an earthquake resulting in the failure of a cracked fuel element and the subsequent loss of coolant flow through the column. In addition, the staff has reasonable assurance that even if it were to assume a loss of coolant to a fuel element column, this would not result in a significant risk to the public health and safety.

Principal Contributor: R. Rothman Dated: DEC 3 o ,ggg 1

l l

,. - -- ., - - - . ._. _ . . - _ - - - .