ML19256D731
| ML19256D731 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 04/26/1977 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML19256D730 | List: |
| References | |
| NUDOCS 7910220654 | |
| Download: ML19256D731 (13) | |
Text
.
SAFETY EVALUTION BY THE ENGINEERING BRANCH OF OPERATIONAL TECHNOLOGY SUPPORTING AMENDMENT NO.
TO LICENSE NO. rPR-50 METROPOLITAN EDISON COMPANY THREE-MILE ISLAND NUCLEAR STATION - UNIT NO. 1 DOCKET N0. 50-289 Introduction The original Three-Mile Island, Unit 1 (TMI-1) design included three reactor vessel surveillance specimen holder tubes (SSHTs) located near the reactor inside vessel wall. Each of these SSHTs housed two capsules containing reactor vessel surveillance specimens.
Failures of the SSHTs at Babcock & Wilcox (B&W) designed plants were first discovered when TMI-1 pulled its first surveillance capsule during the refueling in March 1976.
Inspection revealed that the SSHTs had suffered severe damage. To prevent further damage, all surveillance capsules and all parts of the SSHTs that had failed or were deemed lik.aly to fail were removed from the vessel.
Proposed Proaram Since the discovery of the damage to the SSHTs, B&W has undertaken the design, manufacture and testing of an improved SSHT. SSHTs of this improved design are presently installed in Davis-Besse 1, Crystal 1
River-3 and Three Mile Island-2 (_TMI-2). All three of these plants have reactors supplied by B&W and all are in the process of beginning initial operation at the present time or within 1453 161 g0g20
. the next few months.
In addition, all of these reactors have the same basic B&W 177 fuel assembly design as TMI-l.
The acceptability of the redesigned SSHis has been demonstrated by a test program reviewed and approved by the staff and conducted in conjunction with the Hot Functional Test performed at Davis-Besse 1.
Installation of the redesigned SSHTs in the Davis Besse 1 and Crystal River-3 and T!il-2 reactor vessels did not present any unus.ual difficulties because it was performed prior to neutron activation of the ~ reac:cr internals.
Studies of methods to install the redesigned SSHTs in the irradiated B&W react:rs indicate that substantial difficulties will be experienced - primarily because precision machining, alignment and inspection must be performed remotely and under water. Although such problems do not in themselves justify relief from a requirement to re-install the SSHTs in TMI-1, they would cause significant radiation to personnel.
Based on their experience in removing the SSHTs at Three Mile Island-l and Rancho Seco-1, B&W estimated that installing SSHTs in irradiated reac::rs wculd result in personnel exposure of about 100 man-rem per reac cr.
In the interest of main:sining the radiation exposure of 71 ant personnel as 10w as reescnably achievable, the licensee, in c:oceration with S&W and the owners of other S&W 177 fuel assemoly plants, has procosed an alternative program that does not require reinstalling the SSHTs in TMI-l and the other irradiated B&W plants.
1453 162
. This program is very complex, as it includes provisions to provide additional information, if required under Appendix G 10 CFR 50 Paragraph V.c., in addition to the normal requirements of Appendix H.
The proposed plan involved integrating the interrupted surveillance programs into the programs for new plants in a manner generally similar to that covered in Appendix H,10 CFR 50, paragraph II,C.4.
In the case of TMI-l and TMI-2, the reactors are located at a single site.
In other cases, the plants are at different sites. There are three distinct features of these proposed programs:
l.
A host-reactor feature, in which the original surveillance materials frcm one or more reactors that have been in service will now be irradiated in a new host reac:cr, that can be fitted with the newly-designed capsule holders on the thermal shield in less time and with-out radiaticn exposure of the workmen, and,
2.
An augmented surveillance feature in wnicn enere will be Icre weld metal specimens and scme larger fracture mechanics (c:mpact tensten or CT) specimens placed in the capsules, and 3.
A data-sharing feature in which all availaole irradiation da:a for ali cf the bcitline welds of a given reactor will be c:nsidered in predicting its adjusted reference tamperature and in making any fracture analyses for that reac: r.
Typically, several of tne weics in any one vessel were made with the same weld wire and flux as tnose 1453 163
. used on some other reactors. The data sharing feature is required because the welds in these reactors have high radiation sensitivity due to high copper content, large and random variation of copper from point to point in the weld, and low initial upper shelf energy.
The specific program proposed for TMI-l involves installing the original TMI-l surveillance capsules in extra locations provided in the TMI-2 vessel. This plan will accomplish the original purpose of obtaining information on the effect of radiation on material that is identical to one of the controlling materials in the TMI-l reactor vessel on a schedule that provides an appropriate lead time over the vessel irradiation rate.
The evarall fr.tes.ted orogram also will provide information from surveillance programs in Crystal River 3 Three Mile Island 2, and Davis Besse 1 on material considered to be representative of the welds in the TMI-l vessel.
It is also important to note that still more information relevant to the TMI-l ves:el materials will be obtained from the NRC funded HSST irradiation programs underway.
Details are provided below.
There are three weld materials of primary interest for the TMI-l vessel.
Procedure Qualification (P.Q.) numbers
- WF 70 and WF 25 are used in the top and center circumferential welds. The eid of life (EOL) fluence for I9 both of these welds is estimated to be 1.2 x 10 nyt, and both have compositions that are expected to make them relatively sensitive to
- Weld materials are specifically identified by the ASME Code by *he procedure Qualification Test number. A procedure qualification test is required on each combination of heat of weld wire and batch of flux.
1f8f lcA
- radiation damage.
'W 'd P.Q. No. SA-1526, used for two of the longitudinal welds, also has high copper, and the EOL fluence at the azimuthal locations I8 of these longitudinal welds is 9 x 10 so they may become limiting during the service life. Another shell weld, the lower circumferential, is made of a material that is expected to be radiation sensitive (P.Q. No. WF 67),
but the E0L fluence at this location is estimated to be at least an order of magnitude lower than that of the other circumferential welds, so it will never be limiting.
The original surveillance material, WF 25, was the same as the center circumferential weld, so information on its radiation behavior is already available from the test results from one capsule irradiated in TMI-1, and two more capsules will be withdrawn later from the TMI-2 reactor.
In addition to this ir.tegrated program, "research" capsules containing tensile, Cv, and two size.: of CT specimens of B&W archive materials will be included in the overall B&W power reactor surveillance program.
Weld materials included in the "research'* capsule that are likely to be limiting for the Three Mile Island, Unit 1 reactor vessel are: SA 1526, WF 25 and WF 70.
In addition, specimens of WF 67 material are also included in the "research" program. These capsules will be irradiated in Davis Besse 1, Crystal River 3 and Three Mile Island 2 reactor vessels.
The presently planned withdrawal schedule calls for first capsule withdrawals 1453 165
- in 1981-2 at which time the upper shelf energy of specimens is pre-Jicted to be about 50 ft-lbs, and again in 1989 when the specimens will have received a #1uence approximately equal to that at the vessel ID at end of life.
Also, research programs being funded by NRC will provide continual information on the effect of radiation on these specific weld materials and on several additional B&W weld materials expected to respond to radiation in a similar manner. These programs, HSST-2 and HSST-3, consist of many tensile, Cv, and CT specimens irradiated in a test reactor.
Although information on shift in RT will be obtained, the main emphasis NDT of the HSST programs is to develop methods that can be used to better evaluate low shelf toughness using the rather small specimens used in the power reactor programs.
The following table shows where samples of the pertinent weld materials will be irradiated in the proposed tr.tegrated and "research" programs, what kinds of specimens will be used, and when information will be available under the present plan.
1453 166
. RADIATION DATA FOR TMI-l REACTOR Removal Specimen Weld Capsule Reactor Date Tyoes WF-70 R-1 Davis-Besse 1 1981 Cv, CT (Upper R-2 Davis-Besse 1 1929 Cv, CT Circumferential R-1 Crystal River 3 1982 Cv, CT Weld and 0.0. of R-2 Crystal River 3 1989 Cv, CT Lower Circumfer-HSST-3 Test Reactor 1978 Cv, CT ential Weld) to 1.6T WF-25 TMI-l E TMI-l 1976 Cv, tensiles (Center TMI-l A TMI-2 1982 Cv, tensiles Circumferential TMI-l C TMI-2 1990 Cv, tensiles Weld)
R-1 TMI-2 198i Cv, CT R-2 TMI-2 1989 Cv, CT nSST-2 Test Reactor 1977 Cv, CT to 4.0T HSST-3 Test Reactor 1978 Cv, CT to 4.0T NRL Test Reactor NRL Test Reactor SA-1526 R-1 TMI-2 1982 Cv, CT (Longitudinal R-2 TMI-2 1989 Cv, CT Weld)
Following welds are not controlling WF-67 R-1 Davis-Besse 1 1981 Cv, CT (Lower R-2 Davis-Besse 1 1989 Cv, CT Circumrerential R-1 Crystal River 3 1982 Cv, CT Weld)
R-2 Crystal River 3 1989 Cv, CT NRL Test Reactor WF-8 SA-1494 1453 167
e
.a.
Staff Evaluation The staff has evaluated the effectiveness of this overall pregram plan, and has concluded that the information to be developed that is directly and indirectly relevant to the TMI-l reactor vessel will be sufficient to provide assurance of safety margins against vessel failure that ccmply with Appendix G,10 CFR 50. Further, it is the staff's opinion that even without additional irradiation turveillance programs in the TMI-l vessel, the proposed program will provide more useful information tha n would have been obtained from the original surveillance program.
Until data bec:me available frem the surveillance program, a censervative prediction of radiation damage can be made by using R.G. 1.g9*, which is based en the staff's analysis of all data available at the time :he Guide was written. New data, in particular the results of the augmented integrated surveillance program described abcve, will be used to upcate the Guide periodically. Predictions of the adjustment of reference temperature and the drop in upper shelf energy are given grapnically as functicns of cooper and phospncrus contant and of fluence.
In addition there is an "Upcor Limit" line en eacn graph, which is to be used when informatic' about the capcer and phosphorus c:ntents is inadequa:e.
Because *.1e chemical analyses of the BW welds have snewn considerable P
variation, the staff intends to use the Up:er Limit lines as the basis f r any. prediction requ'. red a: tnis time.
- Regulatory Guice 1.99, " Effects of Residual Elements on Predicted Radiation Damage to Reactor Vessel Materials," July 1975.
Revisien 1 is to be published in Acril 1977.
1453 168
.g.
The staff also has considered the uncertainties involved in applying radiation effects information obtained in other reactors to the TMI-l vessel. The major uncertainties involved are:
1.
Accuracy of neutron fluence calculaticns 2.
Magnitude and effect of variatien in neutron spectra between reac:ces 3.
Magnitude and effect of variations in irradiation temperature between reactors.
4 Magnitude and effect of variations in rate of irradiation en material properties.
The effects of these variables h:ve been studied for at least 20 years.
Although scme uncertainties still remain, the effects are
'airly well established and understecd as discussed belcw.
1.
Neutrcn flux calculations for the react:r vessel wall and irradiation capsule lccations have been developed over many years.
The dosimetry used in irradiation caosules has furnished informa:icn that was used to check out and refine the calculational metheds.
It is generally believed that the fast neutron flux and fluence in these lccations can be calculated :o an accuracy of - 2C%,
particularly if scme dosimetry checks are available.
Ocsimeters frem the original TMI-l surveillance program were removed and tested, so the fluence calculations for o'-* iussel have been verified.
'453 1e9
. In tn s connection it should be emphasized that the effect of neutron radiation on reactor vessel steel varies as the square root of the fluence, so uncertainties of 20 to 50% in fluence are not highly significant.
The staff has also considered the fact that the design of the TMI-l vessel, internals, and core is almost identical to that of the other reactors that will be used to obtain radiation effects information.
These considerations are the basis for the staff's conclusion that uncertainties in the calculation of neutron fluence will be small, and the effect of such uncertainties on the assessment of the radiation effects on the vessel ma:erial will also be small.
2.
Although differences in neutron energy spectra can cause uncertainties in the effects of radiation on material when this is evaluated without consioering spectrum effects, only very large differences in spectra are significant. The variations frem one B&W reactor to anotner are claimed to be relatively minor, because they have similar gecce:ry.
The staff considered the possible differences in neutron spectra tha:
could cccur between the B&W pcwer reactors involvec in the integra:ac Such effects can be dealt with, if necessary, througn :he program.
use of neutren damage functicns that are being develoced for :nat 1453 170
s
. purpose.
Mcwever, the worst expected differences are judged inc:n-secuential based on present kncwledge of irradiation effects.
If additional developments (theoretical or experimental) suggest that the neutron spectra effects might be significant under scme conditions, appropriate actions will be taken.
3.
The effect of the temperature of irradiation has also been the subject of considerable research.
It is well known that radiation damage is less severe at 500*F than at 500*F (the temperature range of cencern).. The differences in effect on the steel appear to be noticeable and shculd be taken into acc unt if the irradtation temperature difference is over about 25'F.
Encugh information is known to permit conservative evaluations of the effect of temperature differences of at least 50*F, and probably even ICO*F cr more.
The differences in the temperature of the surveillance capsules and vessel walls between the 5&W power react:rs involved in the integrated program are expected to be less than 50*F, and can be conservatively evaluated.
4.
The effect of trradiation rate has also been evaluated by research programs at NRL and other laboratories. Althougn the c:nsensus of experts en this sucject is tha: :nere will be ne major differences in material property changes by irradia:icn rates varying over 2 to 3 orders of magnitude, more cats fr m surveillance pr: grams are needed to provide verificatien.
Mcwever, the differences in 1, 4 s s*i G T l1
~
. the rates of irradiation of specimen in the integrated program and the limiting material in the walls of the affected vessel will be less than one order of magnitude, therefore the staff has con-cluded that there will be no significant uncertainties in this program associated with differences in rate of irradiation.
Staff Position The Staff has evaluated the adequacy of the proposed integrated, augmented reactor vessel material irradiation program for TMI-1 as an alternative to the original program that was interrupted by failure of the associated hardware. We conclude that the proposed program will provide the infor-mation required to provide safety margins that ccmply with Appendix G, 10 CFR 50, and that the uncertainties involved in using data obtained from surveillance specimens irradiated in various other B&W power reactors to establish TMI-l vessel operating limitations are small and can be accounted for both easily and conservatively.
Of equal, if not greater, importance is the staff's assessment of the proposed integrated, augmented program (with possible minor modification yet to be finalized), to wit:
It will provice more useful infor :ation
- nan could have been extracted fr:m the criginal surveillance program.
The procesed pr gram will also give results of the kind required to mee:
Paragraph V.C of Appendix G, 10 CFR 50.
1453 172
. until the results of the proposed surveillance program become available, the staff's predictions of radiation damage in the B&W power reactors will be based on the current revision of Regulatory Guide 1.99.
At present, this is Revision 1.
1453 173
-.