ML19322C581
| ML19322C581 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 10/09/1979 |
| From: | Mcdonald W FEDERAL ENERGY REGULATORY COMMISSION |
| To: | Rogovin M NRC - NRC THREE MILE ISLAND TASK FORCE |
| References | |
| TASK-TF, TASK-TMR NTFTN-790829-03, NTFTN-790829-3, NUDOCS 8001170906 | |
| Download: ML19322C581 (6) | |
Text
{{#Wiki_filter:.___ ___________ C ( FEDERAL ENERGY REGULATORY COMMISSloN I w Asm NGTO N 20426 Reference No. NTFTN 790829-03 OCT 9 1979 Mr. Mitchell Rogovin f Director Three Mile Island Special Inquiry Group Nuclear Regulatory Commission Washington, D. C. 20555
Dear Mr. Rogovin:
This supplements my September 11, 1979 response to your letter of August 29, 1979. Enclosed is a revised answer to one of the questions regarding the testing period for TMI-2. Sincerely, William G. Mcdonald Executive Di'ector Enclosure 800;379 g
, REVISED ANSWER TO NRC INQUIRY ON TMI-2 Q. Did the owners of TMI-2 file a report justifying a test period in excess of 120 days? A. A complete report covering the testing period, as required by Electric Plant Instruction No. 9(D) of the Uniform Systems of Accounts, was not filed by the owners.
- However, by letter dated August 18, 1978 (copy attached), the owners advised the Chief Accountant, FERC, of problems encountered during the third day of the initial start-up of the plant.
The owners proposed that the initial start-up date not be used for the purpose of commencing the 120 day testing period referred to in Electric Plant Instruction No. 9(D). No written response was made to the owners' proposal. Instead, on September 29, 1978, Mr. E. J. Holcombe, Vice President and Comptroller of General Public Utilities Service Corporation, was contacted by telephone. He was advised that after the plant was placed in service, the owners should submit the data required by Electric Plant Instruction No. 9(D), irrespective of whether the testing period might end up being less than 120 days if the owners' proposal was accepted. Mr. Holcombe was contacted September 28, 1979, and he stated that the requested data was assembled prior to the accident af March 29, 1979, but was not forwarded to FERC. He promised to mail the data to FERC promptly. Attachment l
rett Service - ' '? GPU Service Corporation e 2= - rF [ p", "q 'f V, ],(. Pa stpam Neo Je se,07054 p 201 263-4933 . _ r.)iiE R c W.$ SiO H August 15, 1975 Federal Energy Regulatorj Cctr.issien 625 N. Capitol Street, N.W. Washington, D.C. 2042E Attention: Mr. L. H. Drennr., Jr., Chief Accountr.:
Subject:
Three Mile Island Nuclea-Geherati-{ 5tation Unit 2 Gentlev.n: In acconiance with Electric Plant Instraution Nc. 9E of your Unifem Syste of Accounts prescribed for Public Utilities and Licenses, as arended by Order 475, wc submit the following infomation conceming the test period for Three Mile Island Nuclear Generating Station Unit 2 in onder that we establish the camencerent of the official test period with the synchronization that is expected to take place in Septer.ber 1975. This request is trade on behalf of our three subsidiarj operati g cmpanies: Jersey Ccntral Pcuer E Light Ccepany, Me ropolitan Edison Ccepag. and Pennsylvania Electric Cmpany. Tnree Mile Island Nuclear Generatir{ Staticr Unit 2 consists of (a) ene reactor vessel utilizing high pressurized water, steam generatcrs, and cther necessa7 ~ auxiliarj equi;nent tc produce steam; and (b) one 900 Kel turbc generating unit. Tne turbo generator was first synchronized to the Ccrpany's system on April 21, 1975 and began test operation for two days until April 23, 1975, when the main steam safety valves failed to function within prescribed limitations (a detailed technical report is attached). Daring these twc days only 1100 ME cf generation was provided. This failure required that the unit be shut dr and the 12 steam safety valves and Irlated piping be replaced by 20 steam safety valves cf a different design. A total of 140 days of additional constructicn tire is expected to be required prior to the nex* attempted synchronization in September. Dae to these extenuating circumstances, we believe that it is appmpriate that the test pericd as outlined in Instraction No. 9D of the Unifem Syster. of Accounts should ccrrence with the synchronizaticn that is expected tc occur in September 1978. Very truly yours, GP SERVICE COPPJPATION / E. C) Holccube&% 14 ~ t EJH:mb Vice President and Cceptroller Attachment hIf ON!b.9ff M h D D J h N!>. > _ooj GPU Servce Corporation is a subsidiary of General Pubic Utilites Corporation
SUMMARY
OF THREE MILE IS1AND NUCLEAR GENERATING STATION UNIT 2 STEAM SAFETY VALVE MALFUNCTION AND RECOVERY PROGRAM A formal start-up testing program was planned and organized for Three Mile Island Nuclear Generating Station Unit 2 similar to the one that was conducted very successfully on Three Mile Island Nuclear Generating Station Unit 1. The program was scheduled for forty (40) weeks. The first event in the test program, functional testing of component parts, but without any power generation, was completed in October 1977, in time to support the May 31, 1978 in-service date scheduled at that time. The test program is planned to fully exercise all of the plant equipment under carefully controlled and monitored conditions so that any deficiencies in' design 6r construction can be identified. The end objective of the test program is to verify that the plant performs in full conformance with all operating and licensing specifica-tions. While it is anticipated that some problems will occur during the test program, for administrative reasons the program schedule makes no explicit provision for delays. Fuel loading into the reactor began during February 1978. The reactor first became critical on March 28, 1978 and was synchronized to the system on April 21, 1978, to begin the testing of actual power generation. By April 23rd, the test program had progressed to the "15-40 per cent power escalation" phase. In so doing the plant has operated at full temperature and pressure at a maximum electrical output of abeut 200 MW and has produced about 1100 MWHrs. of net generation. On April 23, 1978, the reactor tripped while operating at 28 per cent power during the conduct of the start-up program. This type of transient leads to an increase in pressure in both the reactor plant and the steam plant. The pressure increase is controlled by main steam safety valves. The main steam safety valves open as a result of the increase in main steam pressure and relieve this pressure to the atmosphere; however, the main steam safety valves did not reclose when the pressure returned to its normal range. As a result of the safety valves failing to close appropriately, excessive heat was removed from the main steam system, cooling down the steam generators and thereby causing the reactor coolant system to cool down excessively. The rapid cool-down of the reactor coolant system, and the associated decrease in reactor coolant pressure, initiated injection of emergency cooling water in a manner similar to that expected during a loss of coolant accident. During the course of this event, it was noted that liners fro: expansion joints in the discharge piping from the main steam saf ety valves ha d f ailed and were ejected into the air through the main steam saf ety valve discharge stacks. Met-Ed and GPU Service Corporation established a Task Force to review and evaluate the causes and implications of this event and to recommend specific action to be taken to preclude such an occurrence in the future. It was well recognized by the Task Force and others that the main steem safety valves blev down excessively and while corrective action was negessaryl 'u ( \\ L.
SUMMARY
2_ it was believed at that time that normal adjustments to the valves would remedy that problem. The major thrust of the initial action was to correct the deficiencies in the discharge piping expansion joints, and such correc-tive action was completed about the middle of May. Concurrent with the repairs to the safety valve discharge piping, the plant was cooled down for cleanup of the chemicals added to the reactor coolant system in conjunction with injection of emergency cooling water and correction of other minor problems identified by the test program. Upon return of the plant to normal operating temperature and pressure without nuclear power, main steam safety valve testing was initiated on iby 15, 1976, with the expectation of adjusting the reclosure pressure of the valves to correct the excessive blowdcen. It becare apparent about May 20th that the allowable adjustments were not correcting the reclosur( problem with the main steam safety valves. On May 23, 1978, a meeting was held with engineering executives of the Lonergan Company (designers and manufacturers of the safety valves), and GPU Service Corporation, to determine the course of action that would be taken to correct the deficient valve operation. The Lonergan Company stated that two specific changes would result in acceptable valve performance. One change was a reduction of the back pressure caused by the design of the valve discharge piping and the second involved internal modifications to the valve. These changes were made on two valves and testing was resumed on May 26, 1975, with the modified discharge piping, and on Fby 31, 1975, with the modified valves. Concurrent with the joint effort with Lonergan, Burns and Roc (the Architect-Engineer for TMI 2) was directed to start engineering work to design modifications to the plant which would be necessary if the Lonergan valves had to be replaced. GPUSC personnel began immediately to canvass valve suppliers to identify the availability of replacement valves. In the first part of June, a testing facility in Huntsville, Alabama became available for modification to permit off-site testing of the valves, and arrangemants were made to test both modified and unmodified valves at that facility, in the meantime, testing at the plant continued through the 4th of June. Fifty-one (51) valve tests were accomplished in the period from May 18 through June 4, sixteen (16) of which were with modified valves. One hundred eleven (111) tests were conducted at Huntsville through June 22, 1976. ~ None of the tests at Huntsville or at the plant site resulted in acceptable valve performance. I l During the period of evaluation and testing from mid-May through June 2 3, 1978, many alternatives were considered for resolution of the problem. The search made throughout the United States for available replace-ment valves resulted in no valves being found that were available immediately that would fit the TMI 2 steam line configuration. The Forked River Nuclear Generating Station is being supplied valves of the same size as the Lonergan ( valves but which are made by another valve manufacturer but those valves will not be available until the end of November 1979. Smaller valves were located that were available immediately and which are similar to the valves used for TM1 1.
. 500';*- - On Jun( 22, 1978, it was apparent that the valve testing at Huntsville was non-produc tive and that further testing of the Lonergan valves would not achieve satisfactcry results. It was decided, therefore, to purchase twenty (22) stalle r valves of the TMI 1 design to replace the twelve (12) Lonergan valves which did not perform adequately. By that time, the necessary modifi-c a t i e n.* to the main stca: lines had been identified and the necessary material to acco:plish these modifications had been located, and procurement had conctnced. The stear line modifications involve welding into the four (4) main stea: lines, a total of twenty (20) new nozzles. k'elding the nozzles into place re; ire'. tting holes in the main steam pipes, precision fit up of the new nec:les with th+ holes and completion of the welding process which cust includc in-Trecess a 4 post-weld inspecticns and pcst-weld heat treattent. In additien,
- - d a f e t:. vv.ve discharge piping insice the building has to be removed and rc;1ared with different piping with a different configuration.
All of this verk rust be acco=plished in an elevated an'd congested area of the plant. Fellowin; cerpletion of the piping modifications, the valves cust be installed, tht sten lint insulated, the systet hydrostatically tested, and the saf et:. valves lift and reseat pressures tested. The main steam safety valve modification was initiated June 23, 1975, and is expected to be cocplete by the end of August. Completion of the main stear safety valve modification in August will permit a return to power in mid-Feptembct, a delay of about one hundred forty (140) days in the project. 3 j}}