ML20207H679
| ML20207H679 | |
| Person / Time | |
|---|---|
| Site: | Fermi  |
| Issue date: | 07/22/1986 |
| From: | Agosti F DETROIT EDISON CO. |
| To: | Adensam E Office of Nuclear Reactor Regulation |
| References | |
| VP-86-097, VP-86-97, NUDOCS 8607250004 | |
| Download: ML20207H679 (9) | |
Text
r- -
1 Frank E. Agoott Vice Prmoent Nucisar Operations Fermi 2 6400 North Ossie Highway e
l Newport, Michigan 48166 Nuclear ouum&4tw oprauon.
July 22, 1986 VP-86-0097 Office of Nuclear Reactor Regulation Ms. Elinor G. Adensam, Director Project Directorate No. 3 Division of EUR Licensing U. S.
Nuclear Regulatory Commission Uashington, D. C.
20555
Dear Ms. Adensan:
Reference:
Fermi 2 NRC Docket ro. 50-341 NRC Licence Uo. UPF-43
Subject:
Service Life of tiain' steam Evnass Lines In September 1905, Detroit Edison discovered through-wall cracks in the two main stean bypass lines.
A subsequent investigation and evaluation resulted in the replacement of the original bypans lines with a thicker-walled, redesigned piping configuration.
As t
part of the redesign effort, Hopper and Associates was contracted to perform a line life prediction.
The results of the Hopper and Associates analysis are
[
discussed in the Attachment.
As discussed with URC personnel subsequent to the l
receipt of the Hopper report, the report predicts a very short line life for the redesigned piping.
These results are considered to reflect a significant level of conservatism due to:
a) the use of the fracture mechanics analytical methodology; and b) the response characteristics of the original bypass line being used l.
in the analysis of the redesigned line.
i.
8607250004 860722 PDR ADOCK 05000341 P
\\
i l
At h
l
Ms. Elinor G. Adensan July 22, 1986 VP-86-0097 Page 2 The enclosed report ic submitted in accordance uith a connitment Detroit Edison made in a July 2, 1986 telephone conversation with MRC-NRR and Region III personnel.
Please direct any additional questions to Mr. R. L.
Hoolley at (313) 586-4211.
Sincerely
~ h E/
cc:
IIr.
M.
D. Lynch fir. U. G. Rogers USHRC Document Control Desk Washington, D.
C.
20555
i 1
I 1
Attachment to VP-86-0097 Detroit Edison's Evaluation of Hopper and Associates Report ITA-5/06-494 i
4 4
i' i
i 1
4 a
k I
1 f
e
I.
IUTRODUCTION In September, 1985, through-wall cracks were discovered in the Fermi 2 main steam bypass lines.
The subsequent investigation and analysis resulted in the replacement of the original bypass lines with a thicker-walled, redesigned piping system.
As part of the redesign of this piping, Detroit Edison contracted Hopper and Associates to review the modified system and perform a liac life prediction.
The Hopper and Associates report (Reference 1) indicated that the redesigned system could have a very short line life.
In a July 2, 1986 telephone conversation between Detroit Edison, URC-NRR and NRC-Region III personnel, Detroit Edison committed to submit a report which details the measures to be implemented to ensure the integrity of the bypass lines recognizing the concerns in the Hopper and Associates report.
Detroit Edison's evaluation of the subject report is provided below.
II.
General II.1 History The main steam bypass system is comprised of two English Electric Company control valves, approximately 300' of API SL Grade B piping and associated pipe hangers and snubbers.
The 300' of piping 10 divided into tuo discrete flow pnths.
There are 30" diameter and 24" diameter portions of piping in each flow path.
The design functions of the system include isolation of the 52" diameter main steam manifold from the turbine condenser steam space during power operation, control of the main steam manifold pressure during start up or low power operation, and steam dumping to the turbine condenser in the event of an accident or turbine trip.
The capacity of the system is 250 of full rated reactor steam flou.
The first indications that the bypass piping had experienced full penetration pipe wall cracking occurred on September 15, 1905.
A plant operator observed steam and unter issuing from one of the pipelines at a snubber location.
The system was shutdown.
After removing the insulation from all of the hanger and snubber locations along the system, cracking was observed at sites of discontinuity along the system.
These sites included
integral lug attachments, small tect connections and the bypass steam flou measurement orifices.
The line failures were discussed alth the URC Resident Incpector and technical personnel from Region III offices.
Periodic discussions were subsequently held with Resident and URC Region personnel to appraise them of the situation at the Fermi II plant.over the cource of the analycia and repairs.
The lines were temporarily repaired and operated to permit test data to be obtained.
The pipe wall stress and line displacement information obtained formed the basis for an engineering assecsment of the root cause of the cracking.
Detroit Edison, in cooperation with Exxon Research Corporation, Stone & Webster, Hopper and Associates, English Electric and others performed a curvey of the industry to determine if this phenomenon had occurred elsewhere.
In parallel with the conduct of the industry survey, both an analytical evaluation of the test data and a literature search were initiated.
The test data evaluation was decigned to determine the root cause of the cracking.
The literature search was conducted to determine the adequacy of the original system design and included the review of technical and system design basis documentation.
The results of these reviews are presented below.
II.2 Initial Analysis The analycis and evaluations revealed the follouing information:
o The root cauce of the high cycle fatigue cracking vac acoustically induced pipe wall vibration.
Further, the source of the acoustic forcing function was high pressure drop acrocc the bypacc valve seat.
This pressure drop reculted in a supersonic unbounded underdeveloped steam jet entering the piping cystem.
o The une of 30" diameter, 3/8" thick unlled piping which had local discontinuities at integral lug attachments, etc., concentrated the high acoustically induced strences.
o Acoustically induced ctresses were not recogniced as a design consideration at the tine the original system was decigned.
Technical papern and literature were not present until the late 1970s and early 1980s, long after the original design had been completed.
n o
The steam bypass system configuration which failed was unique to Fermi 2.
The only other nuclear plant reporting bypass line cracking was Grand Gulf - Unit 1.
The Grand Gulf system employed four valves located near the turbine condenser.
The valve operated sequentially to achieve pressure regulation.
This differs significantly from the two valve, parallel operation of the Fermi 2 system.
II.3 System Modifications As a result of these findings, the follouing modifications were made at Fermi 2.
Detroit Edison has replaced the 30" and 24" bypass system piping installed between the valve outlets and the turbine pedestal wall.
The 24" piping from the pedestal wall to the condenser injection nozsics vill be repltced in a future modification.
The following discussion aelineates the modifications that were made for the piping that was replaced.
The original 30" diameter, 3/8" thick wall pipe was replaced with 30" diameter, 1" thick walled pipe.
The 24" diameter, 3/8" thich wall pipe was replaced uith 24" diameter, 1-1/4" thick walled pipe.
All pipe wall discontinuities were removed.
This included the integral lug attachments, test connections and the flow measure orifices.
The piping support system was redesigned.
The turbine condenser nozcles and steam sparger was inspected as well as the bypass valves.
Dual pressure breakdoun orifices were installed in each flow path.
The entire bypass system uns instrumented with stress gauges.
The instruments were located at previous test cites and all potentially high stress areas, such as pipeline bifurcations at orifice plate weldaents.
III.
Analysis of Through - Uall Cracks III.1 Hopper and Associates Report In parallel with implementing the design modification, a fracture mechanics analysis of the modified system was performed by Hopper and Associates at Detroit Edison's request.
The purpose of this analucis was to predict the service (line) life of the new bypass piping configuration.
The results of this analysis were provided to Detroit Edison via Hopper and Associates Report No. HA-5/86-404 (neference 1).
The report indicated that the line life of the redesigned bypass lines could be very short depending upon the assumptions made.
III.2 Stone and Webster Report The Hopper and Associates report uns subsequently reviewed by Detroit Edison and Stone and Uebster.
As a result of this review, Stone and Uebster uns requested to perform a comparable analysis of the redesigned bypass lines.
This study is presently on-going.
A draft of the Stone and Hebster report has been reviewed by Edison and indicates that an endurance (fatigue) life analysis was performed.
Preliminary results suggest a predicted line life significantly longer than the Hopper and Associates report, yet still shorter than the desired line life.
III.3 Evaluation of Conservatisms Doth the Hopper and Associates report and the draft Stone and Hobster study reflect an appreciable level of conservatism in their analyses.
This conservatism is a result of developing forcing function data and analytical correlations using measurements of the original bypass lines, and applying these characteristics to the redesigned system which has few of the design shortcomings inherent in the original design.
A few of the conservatisms present in the Hopper report that may affect the results of the analysis are presented below, o
The fracture mechanics methodology assumes the procence of discontinuities and flaws in the pipeline material.
These act as stress ricers and promote crack growth.
To assess the adequacy of this assumption, Edison performed an ultrasonic test of the 30-inch diameter pipe's ceam velds between the bypass valvo discharge and the first orifice plate.
No flaus of the cises used in this analysis were identified, o
The Hopper report indicates that for valve openings from 0 to 31%, the maximum stress ranges were measured to be a linear function of the percentage of valve opening.
The extrapolation of this linearity to 100%, as assured in the report, is considered to be excessively conservative.
o The pipe crack correlations were developed from data on 3/8 inch valled pipe.
The applicability of these correlations to pipe with 1-inch to 1.25 inch wall thichness is uncerte.in.
o The Hopper report does not acknowledge the presence of the two flow orifices installed in the redesigned line.
Therefore, the ability of Detroit Edison to determine the service life of the bypass lines via the analytical methods described above is indeterminate.
In order to more definitively predict the service life, Edison will implement a test program to measure characteristic data of the redesigned bypass lines at low power operations.
This program is discussed briefly belou.
IV.
Testing Progran A test program will be implemented by Edison that will provide measured pipe vall stress data over the design flow ranges of the bypass lines.
The test program will consist of two tests at each of three low reactor power levels (tentatively set at 5%, 8% and 12% of' rated power).
The results of these tests will provide:
a) actual stress measurements for the redesigned bypass line; b) test data at various flow rates through this system; and c) support an evaluation of the ability to linearly extrapolate stress data to 100%.
The test data will be provided to both Hopper and Associates and Stone and Uebster for incorporation into their analyses.
The results of the revised analyses will be reviewed by Detroit Edison.
V.
Interim Goeration Administrative guidelines have been issued which are designed to minimize the potential for cracking in the present bypass lines for the period of time betueen restart and final resolution of this issue.
These guidelines direct Fermi 2 operations personnel to utilize the bypass line flou path at less than 30% of its capacity to the extent possible.
VI.
Conclusion As discussed above, Detroit Edison has reviewed the Hopper and Associates report with both Hopper and Stone and Uebster.
It is the joint technical opinion of Edison, Hopper and Stone and Ucbater that this report reflects a significant level of conservatism and, therein, predicts an unrealistically short service life.
Detroit Edison, therefore, believes that the present bypass lines are suitable for service at the present time.
The lines will perform their design basis function without posing an undue threat to the health, safety, and uc1 fare of plant personnel or the general public.
References 1.
IIopper and Associates Report No. IIA-5/86-494,
" Summary of Procedure for Fatigue Crack Growth Analysis of Steam Bypann Pipeline", dated May 15, 1986.
2 s