ML16134A679
| ML16134A679 | |
| Person / Time | |
|---|---|
| Site: | Oconee  |
| Issue date: | 09/29/1982 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML16134A678 | List: |
| References | |
| NUDOCS 8210210299 | |
| Download: ML16134A679 (7) | |
Text
~REG UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C. 20555 SAFETY EVALUATION REPORT BY THE OFFICE OF NUCLEAR REACTOR REGULATION STEAM GENERATOR REPAIR AND EMERGENCY FEEDWATER SYSTEM MODIFICATIONS DUKE POWER COMPANY OCONEE NUCLEAR STATION, UNIT NO. 3 DOCKET NO. 50-287 Introduction In April 1982, the steam generator (SG) internal emergency feedwater (EFW) headers at Davis-Besse 1 were discovered to be damaged. Since a similar design was used in the Oconee Unit 3 SGs, Duke Power Company (Duke or the licensee) removed the Unit from operation on April 23, 1982 for inspection of the EFW header and commenced a refueling outage earlier than had been planned.
The damage discovered in the Oconee 3 SGs was similar to that found in the Davis-Besse SGs whe-ein the outer surfaces of the headers were deformed to be concave and uneven, header supports were deformed and header support pins were missing.
Inspections of the Rancho Seco SGs disclosed similar damage. No other operating facilities utilize this particular design. The three affected units have been modified to remove the.damaged internal EFW header from use and securely fasten it in place while incor porating an external EFW header to fulfill the system.injection capability.
Specific details of the damaged internal header and newly installed external headers, in addition to the following discussion, are contained in References 1, 2, 4 and 5.
Evaluation At a meeting on June 24, 1982, Duke and the Babcock and Wilcox (B&W) Company presented their plans to retire-in-place the existing internal EFW header and to install an external EFW header on each of the two steam generators. The new design is a modified design of the external EFW header used at several other B&W plants, including Oconee Units 1 and 2. The proposed design includes a header arrangement with individual riser pipes for the injection of EFW4phrough new penetrations in the SG. There are six riser pipes with flow equalizing orifices at the header. The new EFW injection nozzles pass below the internal EFW header. The existing EFW header will be';secured as an extension of the shroud within the SG and is discussed later in this evaluation. This information was documented by letter dated September 10, 1982 (Reference 5).
8210210299 620929 PDR ADOCK 05000267 P
-2 The licensee stated that the cause of the damage to the internal EFW header was rapid condensation induced high differential pressure.
The Staff concurs with this finding.
Having the external EFW header will facilitate visual inspection of the EFW piping and thereby readily identify any future damage due to rapid condensation induced high differential pressure.
Similar designs have been in service for 22 reactor years without any reported rapid conden sation induced high differential pressure damage.
The proposed design differs from the inservice design in two respects:
higher elevation of AFW injection and use of flow equalizing orifices.
Both of these items will result in a higher EFW flow resistance. Duke stated that an EFW flow test would be performed, but did not specify that the test would verify a minimum EFW flow-in order to assure continued conformance with of our March 10, 1980 letter associated with TMI Action Plan Item II.E.1.1.
The test must verify a minimum AFW flow of 450 gpm at 1065 psia to each SG as indicated in Duke's submittal dated April 3, 1981.
Subsequent discussions with Duke determined that the flow test committed to in Reference 5 would verify the minimum flow requirements and further that TS 4.9.3 requires such a test.
Duke also stated in Reference 5 that no water hammer test was necessary since no water hammer damage has been reported with similar EFW header designs. Although the proposed EFW design is similar to other designs, it is not exactly the same and therefore must be reviewed against the current criteria, namely General Design Criterion 4, "Environmental and Missile Design Bases", and the guidelines of NUREG-0800, Standard Review Plan, Section 10.4.7, "Condensate and Feedwater System" and Branch Technical Position ASB 10-2, "Design Guidelines for Water Hammers in Steam Generators with Top Feedring Design".
All plants with new EFW piping are being reviewed to these criteria and thus are being required to perform a water hammer test acceptable to the staff to assure that there will be no dynamic effects associated with possible fluid flow instabilities during normal initiation of the EFW system.
Therefore, Oconee Unit 3 should perform a water hammer test t6 verify the ability of the EFW configuration to operate free from such flow instabilities as those caused by rapid condensation induced high-differential pressure.
Generally,- the test should be performed with the SG at the normal water level and SG temperature and pressure corresponding to full power without using the EFW system. -The maim feedwater system should be tripped and the water level allowed to drop in the SGs and thereby automatically initiate the EFW system.
Sufficient instrumentation should be placed on the AFW piping, or personnel stationed near the AFW piping, to detect any water hammers.
The test can be considered completed when:the water level in the SGs starts to rise and no flow induced instabilities have been detected. Following discussions, Duke agreed to our position and committed to perform an acceptable water hammer test prior to the resumption of full power operation.
The extensive inspection program carried out for Davis-Besse 1 showed more or less the same type of damage for other plants. Namely, the vertical wall of the header was distorted inward towards the center of the SG, the support brackets were bent or damaged and the dowel pins were either out of position or missing.
As a result of the header distortion, the
-3 distance between the peripheral tube and the unstabilized header was reduced, resulting in potential intermittent contact between the tube and the internal header assembly.
It should be noted that no cracks of any length were observed in the welds of the header of Davis-Besse Nuclear Plant.
For the Rancho Seco Nuclear P1ant, in addition to the typical damage described above, the inspection program revealed two cracks in the corner welds of the header.
The first crack is approximately 20 inches long and the other is 15 inches long. Both these cracks were located in the lower inside weld which joins the inner plate to the bottom of the internal header of the SG A and B respectively.
For the A once through steam generator (OTSG) of the Oconee 3 plant, in
.addition to the previous typical damage, a hole is observed in the internal header top horizontal plate and a hole and crack in the bottom horizontal plate. While making the bulkhead repair, a third small hole was found on the inner vertical wall at the elevation of and between two of the sixty 1 1/2 inch flow holes. A localized area of corrosion was noted at each hole location on the inside surface of the header. Moreover, during the preparation to rotate the A header to allow the hole in the bottom plate to be cut for analysis, many cracks were observed in the lower inner corner weld of the header. Further inspections were made on most of the upper inner weld. However, because of accessibility limitations, the upper outer weld was inspected in only a few areas thus resulting in less than full inspection. Because of this and because of the many observed weld cracks, a very conservative approach of assuming all corner welds were fully cracked was taken in the stabilization repair program and related stress analysis and thus the additional radiation exposure which would have been required for full inspection of the welds was avoided. Analysis of the causes of these cracks, by cutting two welds samples from the A header, reveals a lack of fusion, lack of full penetration and lack of lower horizontal plate weld prep. While the construction drawing required a "V" shaped weld prep at the interface of the vertical and horizontal walls, the analysis shows that only the vertical walls were prepped which resulted in an inconsistent weld pattern with areas of lack of fusion and lack of full penetration. Thus, the areasi of these welds were simply inadequate to withstand the loads generated when the header was deformed during the EFW initiations.
For the Oconee B OTSG in addition to the typical damage observed on other plants, one small hole was found in-the top horizontal plate of the header.
Further inspection of the weld has resulted in discovering only two cracks
-approximately two inches long and three inches apart at the upper inner weld.
In addition to the visual and UT inspections, a section of the lower plate was cut out for analysis.
The analysis showed a weld of higher quality than the same weld in OTSG A, which was apparently strong enough to withstand the forces on the weld when the header deformed.
Moreover, the analysis shows that the hole was the result of a corrosive attack in a very localized area.
The repair program performed for the B OTSG included installation of a new external header, and strengthening and stabilizing the internal header by addition of welds and gusset plates to insure the stability of the internal header.
-4 In order to demonstrate the structural adequacy of the B OTSG stabilized header, the licensee performed a three dimensional analysis where the stabilized header was modeled using the ANSYS Finite Element Code. The shroud was also modeled along with the alignment pins. The analysis was performed for the appropriate load combinations of deadweight, flow induced vibration', operating basis earthquake, thermal transients, safe shutdown earthquake, LOCA, and main steam line break. It was shown from the results of the analysis that the load combination of Level D loads which includes the main steam line break and safe shutdown earthquake is the limiting case for the header weld analysis. Even in this limiting case, a factor of safety of 2.2 over ASME Section III Code allowable stresses is obtained. This could be interpreted to mean that only 45.5%
of the weld would be required to insure the structural integrity of the header provided that the damage in the weld is evenly distributed in the circumferential direction.
Another analysis has been performed using the load Level D combination with an assumed 28 inch crack in the inner corner of the weld to determine the effect of such a crack on the header stress pattern.
The results of the analysis shows that the crack does cause a slight increase on the local stresses but'has no impact on the safety function of the header. Moreover, the largest loading on the corner weld is a moment about the tangential axis due to differential thermal expansion. Thus, the potential for crack propagation will be in the radial direction (i.e., through the wall) and not in the circumferential direction.
For the A OTSG, in addition to the typical repair program performed for the B OTSG, the decision was made to reinforce the header to such an extent that the corner weld would not be required for structural integrity.
To provide this reinforcement, two ribs were welded inside the header at each of the header to shroud attachment points, one at each of the attach ment welds and also a pair of reinforcing ribs at mid span between the attachment points.
The reinforced header is subject to loads which cannot be simulated using axisymetric models. To provide the required accuracy, the header, 30 reinforcement ribs, eight attachment points and an attenuation length of the shroud were modeled as a three-dimensional structure using the ANSYS Computer Code.
The header and reinforcement ribs were modeled using quadrilateral plate elements.
Because of the uncertainty about the header corner welds, the model contains no moment or translation connection between the header wall plates.
The only exception to this is when pressure forces the corners of the plates together.
In this case simple support is modeled.
Since fillet welds were used between the reinforcing -plates and the header walls, no credit was taken for the weld's ability'to carry moments.
The 8 gusset plates which provide the remainder of the attachment was modeled by two triangular plate elements.
The modeling of the attach ments was a refinement of the previous model used for the unreinforced header.
The shroud is also modeled using quadrilateral plate elements and one dimensional gap elements to simulate the alignment pins and their inter action with the SG shell.
The analysis was performed f6r the appropriate load combinations of dead weight, flow induced vibration, operating basis earthquake, thermal transient, safe shutdown earthquake, and main steam line break.
The loads were combined according to the ASME Code Criteria and the resulting stresses compared with allowable values also in accordance with the ASME Code.
The conclusion
-5 is that the reinforced header is adequate for all anticipated loads and that the header structure, as reinforced, has sufficient margin such that none of the-corner welds of the header are required. In addition, the corroded area around the header hole.of the A OTSG was reinforced such that, with the reinforcement, there is 108% of the cross-sectional area that would exist in an undamaged inner and top wall.
Therefore, it is concluded that this area of the header will be at least as strong as an uncorroded section.
The licensee has also performed an analysis to demonstrate that the dis tance between the stabilized internal header and steam tube bundle is suf ficient to preclude damage to the adjacent SG tubes. It was shown that for the limiting case, substantial margin still exists-in the inconel tubes to preclude any rupture. Furthermore, the licensee has committed to inspect the stabilized internal header, the attachment welds and external header sleeves through selected openings during the next two refueling outages and at the ten year inservice inspection intervals.
With the exception of two dowel pins (3/4 inch diameter, 2 11/16 inches long), all loose parts have been located and removed from the SGs.
The two unrecovered dowel pins, from SG 3B were not located through extensive visual inspections of the steam annulus area and the 15 tube support plate (on the inside.of the shroud).
The inspections disclosed no damage or spreading of the SG tubes, indicating that the dowl pins could not have entered the tube bundle.. Therefore, the pins must be either in the steam annulus (unlikely following extensive cleaning) or the steam line. The dowel pins pose no safety problem in the steam line, since even if steam flow pushed the pins along the steam line, they would be captured by screens upstream of the maii steam stop valves.
Based on the proven performance of the external header design used for other similar plants, the analyses which were performed for both the internal and external header, the margin of safety associated with all loading combinations postulated to occur including the potential crack of 28 inches for the B OTSG together with the other conservative analysis assumptions used.for the A OTSG, the flow and water hammer tests which will be performed to verify proper system performance prior to resumption of normal operations and the future inservice inspection planned for the retrofitted and the secured headers, the staff has concluded that the repair program, together with the new external header and associated a auxiliary feedwater piping system modifications, is adequate and sufficient to permit the safe.continued operation of the plant.
-6 Conclusion We have concluded, based on the considerations discussed above, that:
(1) because the facility modification does not involve a-significant increase in the probability or consequences of an accident previously evaluated, does not create the possibility of an accident of a type different from any evaluated previously, and does not involve a significant reduction in a margin of safety, the facility modifications do not involve a significant hazards consideration, (2) there is reasonable assurance that the health and safety of the public will not be endangered by operation in the pro posed manner, and (3) such activities will be conducted in compliance with the Commission'sregulations and will not be inimical to the common defense and security or to the health and safety of the public.
Dated:
September 29, 1932 The following NRC personnel have contributed to this Safety Evaluation:
P. C. Wagner, A. Hafiz, J. Ridgely.
-7
REFERENCES:
- 1. DPC Reportable Occurrence Report RO-287/82-06 dated May 14, 1982.
- 2. DPC Reportable Occurrence Report RO-287/82-06, Revision 1, dated June 14, 1982.
- 3. NRC Letter to DPC dated June 23, 1982.
- 4. DPC Reportable Occurrence Report RO-287/82-06, Revision 2, dated August 27, 1982.
- 5. DPC Letter to NRC dated September 10, 1982,