NRC-88-0081, Revises Response to Generic Ltr 87-05 Re Identification of Potential Degradation of Mark I Drywells.Response to Item 2 Revised to Correctly Reflect Util Commitments Concerning Minimizing Possibility of Leakage from Refueling Cavity
| ML20151Q966 | |
| Person / Time | |
|---|---|
| Site: | Fermi  |
| Issue date: | 04/20/1988 |
| From: | Sylvia B DETROIT EDISON CO. |
| To: | NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM) |
| References | |
| CON-NRC-88-0081, CON-NRC-88-81 GL-87-05, GL-87-5, NUDOCS 8804270191 | |
| Download: ML20151Q966 (5) | |
Text
a C. Ralph Syms Group Vice Presdent 6400 Nr th Dme Hi Ec ison==
ghway o
April 20, -1988 NIC-88-0081 U. S. tbclear Regulatory Comnission Attn: Document Control Desk Washington, D. C.
20555
References:
1)
Fermi 2 NIC Docket No. 50-341 NIC License No. NPF-43 2)
NIC Generic Letter 87-05, "Request for Additional Information-assessent of Licensee Measures to Mitigate and/or Identify Potential Degradation of Mark I Drywells," dated March 12, 1987 3)
Detroit Mison Letter to NIC, "Potential Degradation of Mark I Drywells," NIC-87-0075, dated June 15, 1987
Subject:
Revised Response to Generic Ietter 87-05 Reference 3 submitted Detroit Mison's response to the Generic Ietter 87-05. This letter is to revise our response to item No. 2 of Generic Letter 87-05 to correctly reflect our comitunts regarding minimizing the possibility of leakage from the refueling cavity. The revised response indicates that the sight glass attached to the manifold of the drywell seal bellows drain will be monitored daily for the first three days when the rea: tor cavity is flooded and weekly thereafter for the remaining period while the reactor cavity is flooded, not when the refueling pool is flooded as indicated in the earlier response.
Additionally, the sand cushion drains will be monitoral monthly fo-signs of moisture.
The conplete response to Generic Letter 87-05 is enclosed herewith.
Drawings, figures, and UFSAR subsections referenced as attachments in the enclosure have alrealy been submittal with the original response arx1 are therefore not included with this letter.
I g
e804270191ADOCKgj8o$$41
(
PDR ppg P
s.
s-USNIC April 20, 1988 NIC-88-0081 Page 2 If you have any questions, please contet Mr. Girija Shukla at (313) 586-4270.
Sincerely,
- Enclosure cc: Mr. A. B. Davis Mr. R. C. Knop i
Mr. T. R. Quay Mr. W. G. Rogers i
i 8
Enclosure to NIC-88-0081 Page 1 NIC Item 1 Provide a discussion of your current program and any future plans for determining if the drain lines that were provided at your fmility for renoving any leakage that may result from refueling or from spillage of water into the gap between the drywell and the surrounding concrete or from the sand cushion itself are unplugged and functioning as designed.
Response to Item 1 Fermi 2 has four drain lines that can be used to remove noisture from the sand cushion located between the Drywell wall and the surrounding concrete.
(See attmhed drawing 7M721-2219). These four drain lines are 11/2 inch schedule 40 galvanized steel. There is a P-trap on the end of the pipe. The drains are filled with sand up to the P-trip. A sanple of the sand will be taken during the first refueling outage.
This would be the appropriate tine to perform this test since the fuel pool areas will be flooded. If any leakage were to occur this would be the nost likely time. Chce the sanple is taken, it will be testa 3 for noisture and the presence of any contaminants that would result from corrosion of the drywell steel. In addition, the ems of the drain lines will be visually examined to determine any evidence of past leakage from the sard cushion. Further reconneMations pertaining to any additional or future mtions will be based upon the results of these exninations.
The drain lines located in the gap between the Drywell and surrounding concrete have been visually inspected and found not to oe plugged and functioning as designed. A nore detailed discussion of Detroit Edison's future plans for these drain lines is discussed in the response to h7C Mditional Item section contained in this letter.
NIC Item 2 Provide a discussion of preventive maintenance and inspection activities that are currently performed or are planned to minimize the possibility of leakage from the refueling cavity past the various seals and gaskets that might be present.
Enclosure to FE 0081 Page 2 Besponse to Item 2 The Drywell seal bellows drain enpties into a manifold which is equipped with a sight glass. This sight glass will be monitord daily for the first three days when the reactor cavity is flooded and weekly thereafter for the remaining period while the reactor cavity is flooded, in order to detect any potential leakage of water. If any leakage is detected, Operations will evaluate and determine if the water is going to the space around the Drywell shell. In a3dition, the sand cushion drains will be mnitored enthly for signs of moisture.
Procedures will be developed or revised to incorporate the subject actions prior to flooding the reactor cavity.
hT Item 3 Confirm the information listed in Table 1 of Generic Letter 87-05 is correct with regard to your facility.
I Response to Item _3, Detroit D31 son has reviewed the information in Table 1 of Generic letter 87-05 with respect to Fermi 2.
In the colunn title "Gap Material" it states "foam."
To be nore accurate, the gap material used at Fermi 2 is Polyurethane foam. The other information listed for Fermi 2 is correct.
Mditional Information Recuested Mark I owners whose desigt. is such that the sand cushion is open to the gap between the drywell and surrounding concrete were required to submit aiditional information. This additional information is supplied in the nost part in the response to the IE Mditional Item below. Fermi design does have the open gap between the drywell and surrounding concrete. Drawing 6C721-2358 and 7M721-2219 (attached) shows the design etployed by Detroit a31 son for the subject area. In a3dition, Figures 1 arxi 2 (attached) show the current Fermi design and the Oyster Creek design of the drain system. It can be seen that the Fermi Design contains nore drain lines which provide for better
- drainage, bTC Mditlpnal Item _
Provide any plans for performing ultrasonic thickness measurements of the drywell shell plates and a3jacent to the sand cushion or any other
i s
Enclosure to NBC-88-0081 Page 3 proposed mtions to ascertain if plate degrMation has occurred.
Since the degradation that has occurred at Oyster Creek is localized, sufficient details should be included to show that the sanpling basis for ultrasonic thickness neasurements is adequate in terms of size and test location.
Easr>onse to NIC Additional Item Drawing 6C721-2358 (attached) shows the subject area. In order to ultrasonic test (UT) the drywell plates adjzcent to the sand cushion, concrete would have to be cut, removed and then rcplzced orce the Ur was over. This process is neither feasible nor would it be advantageous for the reasons indicated below:
First, the design of the refueling bellows assenbly minimizes the possibility of leakage into the drywell gap. The leakage of the drywell to cavity seal bellows during refueling was considered to be a source of leakage into the sand cushion at Oyster Creek. The refueling bellow assenbly forms a seal between the drywell arx1 the refueling pool to permit flooding of the refueling pool. The drywell seal bellows at Fermi 2 extends from the drywell shell out to the refueling pool and consists of a cylirdrical, one piece stainless steel bellows (drawing 91721-2169 attached). All seal boundaries consist of welded steel and contain no active conponents.
Secondly, a different type of coating is utilized at Fermi 2 for the exterior of the drywell. At Fermi 2, the full length of the exterior surface of the drywell is coated with carboline Carbo Zinc 11 (drawing B2-93 69-5562. This is a self-curing zinc-filled inorganic two-part basic zinc silicate conplex supplying galvanic corrosion protection to steel surfzces in marine environments. The coating is insoluble in water and resistant to aggressive water and solvents (UFSAR subsection 6.2.1.6 attzched). The polyurethane foam sheets locatcd within the drywell gap are coated with an epoxy resin binder (on both sides), to prevent water leakage into the foam (Specification 3071-48). In zddition, the Fermi 2 drywell, in the vicinity of the sand cushion, is 1.5 inches thick (drawing SC-721-2701 attached).
Third, a walkdown found no recent indications of noisture in the sand cushion drain lines.
Finally, the area directly below the drywell seal bellows contains a 2" drywell to reactor well seal drain, a 2" liner drain, and a 4" drywell seal rupture drain (Figures 1 and 2 attschcd). Should the refueling bellow assenbly fail, this design minimizes the possibility of water leaking into the sand cushion. Instead, the leakage would enter these drain lines.
i