ML20071E269

Forwards Info Re Pressurized Thermal Shock Flux Reduction. NRC in-depth Review of B&W Owners Group Programs Will Reveal Acceptably Low Level of Risk

ML20071E269
Person / Time
Site:	Three Mile Island
Issue date:	03/07/1983
From:	Hukill H GENERAL PUBLIC UTILITIES CORP.
To:	Stolz J Office of Nuclear Reactor Regulation
References
REF-GTECI-A-49, REF-GTECI-RV, TASK-A-49, TASK-OR 5211-83-063, 5211-83-63, NUDOCS 8303100127
Download: ML20071E269 (9)

Text

,

GPU Nuclear Corporstbr.

ENuclear =ngsta 8o Middletown, Pennsylvania 17057-0191 717 944-7621 i

TELEX 84-2386 Writer's Direct Dial Nurnber:

March 7, 1983 5211-83-063 Office of Nuclear Reactor Regulation f.ttn: John F. Stolz, Chief Operating Reactors Branch #4 Division of Licensing U.S. Nuclear Regulatory Commission Washington, D.C. 20555

Dear Sir:

Three Mlle Island Nuclear Station, Unit 1 (TMI-1)

Operating License No. OPR-50 Docket No. 50-289 Pressurized Thermal Shock (PTS)

Flux Reduction As a result of our meeting with your staff on January 14, 1983, we committed to supply you with information available concerning questions raised by the Staff on flux reduction. The enclosed information addresses those specific questions.

As a step in the direction of resolving PTS, we have joined with other B&W Owners to form a PTS Task Force within the B&W Owners Group, to continue the dialogue begun during the week of January 10. During those meetings, we presented the status of several key programs that deal with confirming the continued integrity of B&W vessels not only for PTS events but in accordance with requirements of 10CFR50 Appendices G&H. We believe that once the NRC Staff has a chance to review our programs in depth, they will concur that the risk of PTS to B&W vessels is acceptably low and within the limits applied to the rest of tt.e industry.

Furthermore, we believe that a number of alternatives to reaching the screening criterion should be allowed in considering a proposed rule and we are prepared to technically substantiate these alternatives within the next several months. Among the alternatives are:

o Improvement in chemistry characterization for critical materials, a

Refinement of uncertainties in initial RTndt values.

OM GPU Nuclear Cornoration is a subs. diary of the General Public Utilities Corporation 0303100127 830307 PDR ADOCK 05000289 P PDR

l Mr. John F. Stolz Page 2 1

i l Development of RTnat shift correlation for B&W weld metal.

8 l Refinement of vessel fluence calculations to further reduce l

uncertainty.

At TMI, we are pursuing flux reouction schemes and are committed to resolving l

the PTS issue. We (as a member of the B&W PTS Task Force) will be meeting with the NRC Staff regularly over the next several months. We expect to review a number of other PTS related actions to clarify or mitigate our mutual concerns in addition to flux reduction.

Sincerely, f \

%4Ad

.D. ukill Director - TMI-l HDH/k1k Enclosure cc: J. Van Vliet R. Woods R. Conte l

L .

ENCLOSURE NRC Question 1:

Provide your assessment of the fluence experienced to date by the welds and plates in your pressure vessel, the rate of increase expecteo assuming future fuel cycles to which you are already committed, and a detaileo oescription of the bases for the above (incluaing surveillance capsule date and analysis methods, and generic methoos or correlations useo).

Response 1:

An assessment of fluence experienced to date by the TMI-1 pressure vessel is includeo in Reference 1 (TMI-1 PTS Report) which was submitted in June 1982.

Since then the plant has remainea shut down per NRC order ano no additional fluence has accrued. As stateo in the report (Section 7.3), TMI-1 has completed four cycles of operation corresponding to 3.52 EFPY with a total neutron fluence of 0.18 x 1019 n/cm2 at the peak location. The report also includes a description of the bases for this fluence vslue, including surveillance capsule data, analytical methods, generic evaluations, and correlations used.

Cycles 1 to 4 utilized an annual out-in-in fuel management strategy. The calculatea maximum fluence accumulation rate using this scheme for future cycles would be 0.055 x 10 19 n/cm2 per EFPY. Cycle 5 is also designed as an out-in-in cycle. The projected cycle capacity factor for Cycle 5 is about 73%xand 0.22 1019 the expected total fluence accumulation at the end of cycle is about n/cm2, Current fuel management planning is for conversion to an in-out-in fuel loading strategy beginning with Cycle 6. Vendor (B&W) estimates to date, which are based on similar extenoeo-cycle designs for other 177-fuel assembly plants, show that the in-out-in scheme will provide a reouction in peak fluence of approximately 30%. This will result in an incremental fluence accumulation rate of about 0.039 x 1019 n/cm2 per EFPY for TMI-1. Based on current projected generation scheoules and using a cumulative capacity factor of 73% for Cycles 6 and beyono, the total peak fluence at end-of-license n/cm2, (EOL), which occurs in 2008, is estimateo as 0.89 x 1019 NRC QuestionJ :

Using the above fluence information, provide your assessment of the RT NDT presently existing in your pressure vessel welds and plates utilizing the methocology outlined in Appendix E to Enclosure A of SECY-82-465, and the expecteo future rates of increase, ano the expected dates when the applicable proposeo screening criterion will be exceeded.

4

Enclosure Page~2 Response 2:

The table below summarizes the requested information. A cumulative capacity factor of .73 for Cycle 5 and beyond was used.

RTNDT ( F) l A I I I I l F'uence Peak l R. G. 1.99 l Guthrie +20 l 110L9 n/cm2l Fluence l I circ l axial I cire l axial l Date per EFPY l1019 n/cm2 WF 70, WF 25 WF 8 ISA 1526lWF 70 WG 25,WG 8 lSA 1526l l .

l & i l l 07/15/83 1 0.051 1 0.18 l 122 l 140 l 119 l 131 l 192 l 209 l 182 l 202 l (Current)l l l l l l l l l l l 07/15/84 l 0.055 l 0.22 -l 134 l 155 l 131 l 142 l 199 1 217 l 189 l 210 l (E0C-5) l l l l l l l l l l l October / l 0.039 l 0.76 l 254 l 268 l 244 l 259 l 255 l 280 l 240 l 270 l 2003 l- l l l l 1 l 1 i l l (SA-1526 l l l l l l l l l l l reaches l l l l l l l l l l l 2700F) l l l l l l l l l l l 05/18/08 l 0.039 l 0.89 l 262 l 277 l 264 l 267 l 263 l 290 l 248 l 279 l (Eno of l l l l l l l l l l l License)' l l l l l l l l l 1 I i l l Screening l l l 300oF l 2700F l 300of l 2700F l Criteria l l l l l l l Allowable fluence at the controlling weld (SA-1526) using the Guthrie correlation is 0.64 x 10 19 n/cm2 to meet the 270oF RTNDT axial welo screening criteria.

NRC Question 3:

Provide a description of the flux reouction options that you have considered for your plant. Include for each option:

a '. Description of fuel management and/or fuel removal and/or fuel replacement with dummy elements, showing core maps for future cycles;

b. Quantitative assessment of resulting flux reduction to critical welds and plates;

c. Parametric stuoy showing future RTNDT values resulting from both the earliest practicable implementation of the option, ano from the latest possible implementation of the plan that will still avoid exceeding the RTNOT screening criterion at end-of-life;

Erclosure Page 3

d. Discussion of advantages ano disadvantages of the option, particularly emphasizing power reouctions caused by the option.

With respect to power reouction, discuss the magnituoe of the reduction and the particular limit (e.g., hot channel factor, DNBR, etc.) causing the power reouction. Also analyze how much relief would be necessary (with respect to the particular limit) to allow full power operation, and assess whether such relief woula be an improvement to overall plant safety (considering LOCA, PTS, transients, etc.).

Response 3:

3.1 In-Out-In Strateay

a. Present planning includes conversion from the current annual out-in-in fuel management strategy to an extended-cycle in-out-in loading scheme which will reduce the neutron fluence to the reactor vessel. This change will be implemented with Cycle 6. Figure 1 (attached) shows a 1/8-symmetric core map for Cycle 5, an annual rycle, indicating the locations of fresh, once, twice, and thrice-i burned fuel assemblies. Figure 2 (attached) shows similar preliminary core maps for transition Cycles 6 and 7 using a 76 fresh fuel feeo batch and once and twice-burned assemblies; burnable poison rods will be usa . Figure 3 (attached) shows preliminary core maps for a 72 feed equilibrium 18 month cycle which will also utilize burnable poison rods.

b.

Vendor (B&W) estimates based on similar extended-cycle designs for l

other 177-FA plants indicate that the in-out-in fuel management scheme will provide a reduction in peak fluence of approximately 30% below that for the current out-in-in scheme.

l Fluence reduction factors will be known more accurately when analysis of. the final design of Cycle 6 is completed later in 1983 and specific power and flux distributions are available, however, estimated table: end-of-license fluence values are shown in the following l

- - - - End-of-License Fluence -- 1019n/cm2 _ _ _ _

With Unmodified With Modified To meet Weld Weld Fuel Management Designation Fuel Management Screening Factor Plan Plan Criteria WF 70 0.76 .90 0.68 1.25 WF 25 1.00 1.18 0.89 WF 8 1.00 1.05 1.18 0.89 1.34 SA 1526 0.84 0.99 0.75 0.64

_. m ---

Enclosure Page 4

c. The earliest practicable implementation of the in-out-in fuel management scheme is Cycle 6, presently scheduled to start in October 1984.

As inoicateo in the chart of Response #2, this plan does not result in all welas meeting the RTNDT screening criteria-by the end-of-license using the Gurhrie equation. The SA-1526 axial weld RTNDT is calculated to exceed the 2700F screening criterion in 2003 after an additional 14.6 EFPY, based upon current TMI-l generation forecasts..

At end-of-license, in 2008, the Guthrie correlation predicts an RTNDT of less than 100F above the 270oF criterion. All other welds of concern ~(WF 70,'WF 25, WF 8) remain below the screening criteria at EOL using the Guthrie correlation.

It may be noted that all welos, including SA 1526, remain Delow the screening criteria when the, correlations of Regulatory Guide 1.99 are used.

d. Baseo on B&W extenced-cycle designs for other 177-FA plants the in-out-in strategy for TMI-l is not now expected to result in any core power reouctions or oecreases in operating and maneuvering limits.

3.2 In-In-Out Strategy

a. GPUN is investigating other fuel cycle design options that could be used to decrease reactor vessel accumulated fluence. These options are based on an in-in-out, or very low leakage (VLL), extenced cycle fuel management scheme.

i The VLL scheme is an extension of the in-out-in (LBP) design in '

that it further reduces RV fluence by reducing the relative power of the ;%ripheral fuel assemblies. Most of the fast fluence seen by the RV is due to the peripheral fuel assemblies. Fluence reduction is accomplished with the LBP design by placing once-burned assemblies on the core periphery with fresh fuel loaaed in the interior core regions. The VLL scheme places twice-burned assemblies on the periphery thus further lowering the power of the peripheral assemblies. Burnable poison rods are also used in the VLL oesign. A schematic of the various fuel management schemes is shown in Figure 4.

b.

A mooerate estimate of fluence reduction for the VLL scheme is 20%

over that of the in-out-in scheme. This would give an incremental peak fluence value of 0.031 x 1019n/cm2 per EFPY for TMI-1.

Using a 73% cumulative capacity factor for Cycle 6 ano beyond, and assuming the VLL scheme is implemented for Cycle 6, an EOL peak fluence of 0.75 x 1019n/cm2 results.

Enclosure Page 5

c. This fluence results in an RTNDT at the controlling weld (SA-1526) of 269oF which meets the 270cF axial wela screening criterion.

A scoping study to evaluate the VLL scheme was recently completeo by B&W. These results are not yet available for use by GPUN.

However, preliminary information inaicates that reduction in the incremental fluence factor may be greater than the 20% estimate.

d. Based on the B&W scoping study, use of the VLL scheme is expected to increase core power peaking by several percent. Plant-specific analyses would be required to establish new operating limits and to verify the feasibility of the VLL scheme for TMI-1, 3.3 Further Studies The estimates on. cussed above indicate that TMI-l would meet the present screening criteria with the in-in-out (VLL) scheme. The criteria may also oe met with the currently plannea in-out-in fuel loading scheme, when cetailed core design analyses are performed for TMI-1.

Nevertheless, GPUN is considering funoing acoltional analyses to more precisely determine the neutron flux contribution from each peripheral fuel assembly location which could be useo to refine fuel loaoing patterns.

GPUN is also participating in the " Fluence Analysis and Dosimetry" tasks of the RV Materials Evaluation Program. These tasks are expected to provide more accurate RV fluence aeterminations with minimal uncertainties and are discussed further in Response 4 below.

NRC Question 4:

Discuss any alternatives you may be considering to flux reduction that will result in delaying or avoiding exceeding the RTNDT screening criterion. These would include topics such as archival materials research, plans to sample ano analyze as-built materials, etc.

Respc~;,e 4:

GPU im iclear, as a member of the B&W Owners Group Materials Subcommittee, is tea in several programs that may provide acceptable alternatives to flux rv . tion. They include:

)

Enclosure Page 6 4.1 The B&W Owners Group has previously funoed a program to more fully characterize the chemistry of reactor vessel weld metal. This effort consisted of chemical analysis of both weld metal surveillance program samples and actual reactor vessel archive weld metal. The results of this study, documented in Babcock & Wilcox leport BAW-1500, September 1978, will aid in reducing uncertainties in RTNOT calculation oue to variations in weld metal chemistry.

4.2 Current studies by B&W are aimed at characterizing the initial RTNDT of B&W welo metals. A statistically significant data base for these materials exists. The results of this analysis will serve to better define it.

both the initial RTNDT value and the variation associated with 4.3 The B&W Owners Group, under the RV Integrity Program, has obtained a substantial body of data on the RTNDT of irradiated B&W beltline weld metals. A B&W correlation has been developed that predicts significantly lower shifts of RTNDT as well as a significantly reduced band of uncertainty.

Current plans of the Owners Group include the submittal of a topical report during April of 1983. We believe that the use of this refined correlation as an alternative to the methodology specified in SECY 82-465, will demonstrate that the TMI-l RV welds will not exceed the propo. sed screening criterion, using planned fuel management strategies, during the life of the plant. The use of this methodology is consistent with 10CFR Appendices G&H as well as the Commission's position of allowing alternate means of not exceeding the screening criteria.

4.4 The Owners Group is currently working with B&W on the feasibility of benchmarking calculated vs. actual pressure vessel fluence.

Conceptually, this effort would consist of installing dosimeters both insioe and outside the vessel at the same location, ano adoitional dosimeters at the same elevation over part of the vessel outsioe circumference.

The results of this effort woulo enable B&W Owners to more precisely define the inner vessel wall fluence and its variatinn with circumferential location. Since the critical TMI-l walo is not at the calculatea location of highest fluence, a more precise definition of the fluence at this location would reduce the uncertainty associateo with  ;

the current RTNDT calculations.

l l

\

1 I

j l

l

i Enclosure Page 7 l

4.5 To further improve the accuracy of RV fluence determinations GPUN is

- participating in Phase XA (Fluence Calculation Benchmarking) of the Reactor Vessel Materials Evaluation Program. This task will benchmark analytical techniques by comparing calculated fluence values to a known flux source in a simulated 177-FA configuration with dummy capsules.

Expected benefits are a reduction of fluence error. bands due to uncertainty ano a corresponding reduction in predicted material degradation.

r