ML20205J207

Forwards Addl Info Re Feedwater Nozzle Cracking Analyses (NUREG-0619),in Response to NRC 851120 Request.Heat Transfer Coefficients Dependent on Nozzle & Sleeve Geometry & Leakage Flowrates.Generic Flowrates Conservative

ML20205J207
Person / Time
Site:	FitzPatrick
Issue date:	01/24/1986
From:	Brons J POWER AUTHORITY OF THE STATE OF NEW YORK (NEW YORK
To:	Muller D Office of Nuclear Reactor Regulation
References
RTR-NUREG-0619, RTR-NUREG-619 JPN-86-02, JPN-86-2, NUDOCS 8601300122
Download: ML20205J207 (7)

Text

i e- 123 Main Street Whte Plains. New Wrk 10601 l 914 681.6240 l

1. > NewYorkPower #"1"e"'" e

& Authority aanuary 24, 1986 l JPN-86-02 i

Director of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C. 20555 Attention: Mr. Daniel R. Muller Director BWR Project Directorate No. 2 Division of BWR Licensing

Subject:

James A. FitzPatrick Nuclear Power Plant Docket No. 50-333 NUREG-0619 BWR Feedwater Nozzle and Control Rod Drive Return Line Nozzle Crackina

References:

1. NRC letter D.B. Vassallo to J.C. Brons, dated November 20, 1985.

2. NYPA letter, J.C. Brons to D.B. Vassallo, dated August 20, 1985 (JPN-85-65).

Attachments: 1. Response to Request for Additional Information, BWR Feedwater Nozzle Cracking Analysis.

Dear Mr. Muller:

Attached is the Authority's response to an NRC request for additional information (Reference 1) related to feedwater nozzle cracking analyses.

In Reference 2 the New York Power Authority submitted three reports concerning this issue. The first report, NEDC-30799-P, provided a plant specific fracture mechanics assessment of the FitzPatrick feedwater nozzle with the existing low flow feedwater controller. The analysis demonstrated that the existing low flow feedwater controller complies with NUREG-0619 as amended by NRC Generic Letter 81-11. The second report, SASR 85-09, presented the results of a fatigue analysis of the FitzPatrick feedwater nozzle based on a new seal refurbishment concept and improved analytical methods. Results showed compliance with the ASME code with two seal refurbishments instead of three as reported previously. The third report, SASR 85-10, evaluated the benefits of rerouting the discharge of the reactor water cleanup (RWCU) flow to all feedwater i

nozzles. This analysis indicated that the reduction in feedwater nozzle fatigue usage at the limiting nozzle location is sufficiently small, such that RWCU reroute is not considered justified. l I

$B82RBsERSI85l$h3 P

[\;

g

Should you or your staff have any further questions regarding this matter, please contact Mr. J. A. Gray, Jr. of my staff.

Very truly yours, John C. Brons Senior Vice President

NuclearGeneration cc

Office of the Resident Inspector U.S. Nuclear Regulatory Commission P.O. Box 136 Lycoming, N.Y. 13093 1

1

l l

ATTACHMENT I TO JPN-86-02 dated RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION BWR FEEDWATER NOZZLE CRACKING ANALYSIS l

t l

l l

t NEW YORK POWER AUT!.10RITY JAMES A. FITZPATRICK NUCLEAR POWER PLANT DOCKET NO. 50-333 )

_ _________J

By letter dated August 20, 1985, the New York Power Authority submitted three reports concerning the analysis of feedwater (FW) nozzle cracking for the FitzPatrick plant'. The following responds to a request for additional information (RAI) related to each of these reports. The item numbers are as they appear in the RAI. The reference numbers are the same as in the reports.

NRC Ouestion A. " James A. FitzPatrick Nuclear Power Station Feedwater Nozzle Fracture Mechanics Analysis to Show Compliance with NUREG-0619." General Electric (GE), NEDC-30799-P/

DRFB 13-109-3, December 1984.

A1. Page 3-1 It is unclear how 27 and 28 startup/ shutdown events were obtained from Figure 3-1 and reference 2 respectively. It is also unclear whether the startup data shown in Table 3-1 through 3-3 were_used in the analysis. Provide a more detailed discussicn on the calculation and application of the thermal events / cycles in the analysis.

NYPA Response A1. The 27 and 28 startup/ shutdown events were obtained from i FitzPatrick operating history data (Reference 2). Figure 3-1 is a conservative assumption for the thermal cycle definition during the first five years (i.e., during manual operation of the low flow controller). This thermal cycle definition was

.used in conjunction with the actual.27 startup/ shutdown events which occurred during the first five years.

Extrapolating the total of 28 events which occurred during the first 5.5 years, the startup data shown on Tables 3-1 through 3-3 was used for the thermal cycle definition for the remaining 35 years of plant operation (i.e., during automatic low flow control). For this remaining time period, 177 events were projected.

NRC Ouestion A2. Page 4-1 The report stated that the heat transfer coefficients were taken from the GE analysis (NEDE-21812) (sic). Justify the fact that the generic heat transfer coefficients used in the GE analysis satisfy the plant specific' conditions at FitzPatrick.

NYPA Response A2. The heat transfer ~ coefficients are dependent on nozzle and sleeve geometry, FW temperature and leakage flowrates. The geometries and FW temperatures specific to FitzPatrick are essentially the same as used in the generic analysis. The generic leakage flowrates are conservative for FitzPatrick, as explained below.

As stated in Reference 1 NEDE-21821,.the two estimates of overall heat transfer coefficients for the " improved' interference"ortriplesleevedesignresultedinanoverall thermal conductance of 56 Btu /hr ft OF for the design functioning as intended and 107 Btu /hr ft2 oF with the primary seal grossly failed thus permitting very high leakage.

Reference 1 states further, " Based on these heat transfer coefficient estimates, it is concluded that the U-100 Btu /hr ft2 oF crack growth curves of Figure 4-139 & 4-140 are the appropriate curves to. apply to the improved sparger design."

In addition, NEDE 21821 states that " heat transfec coefficient measurements at the Moss Landing test facilitp i)?e confirmed this estimate to be reasonable for the entire range of i secondary seal leakages tested (0 to 1.5 gpm)".

The leakage flowrates in the generic case are related to a FW flowrate of 100% with a failed primary seal, and a worst case secondary seal' leakage of up to 1.5 gpm. In the FitzPatrick case, the FW flowrate at which most fatigue damage occurs is.1%

to 2% of rated feedwater flow. At this flowrate, leakage is much less than in the generic case and the heat transfer coefficients are lower. Therefore, using the heat tra6sfer coefficients from the generic case is' conservative.

NRC Ouestion A3. Page.5-1 i

The report stated that pressure stresses require a scaling factor based on a three-dimensional analysis. Cite the reference for the three-dimensional analysis in the report.

Provide the scaling factor and the justification for using it in the analysis.

NYPA Response A3. The reference for the three-dimensional analysis'is Gilman and Rashid ~(Reference 6) as stated in the.last paragraph on page 5-1. The scaling factor is 1.656 as stated on page 5-4.

The scaling factor accounts for the difference'between the axisymmetric two-dimensional model and the three-dimensional model at the nozzle blend radius. The scaling factor is appropriate to use for the axisymmetric pressure stresses below the material surface, since the axisymmetric pressure stress distribution is similar to the three-dimensional distribution but of different magnitude.

NRC Ouestion B. "Feedwater Nozzle Rapid Cycling Fatigue Analysis James A.

FitzPatrick Nuclear Power Plant, General Electric, MDE-21-0185/DRF813-109-3, January, 1985.

O Bl. Page 9 Cite the reference for the GE feasibility study that was mentioned. This study indicated that seal refurbishment which includes restoration of the initial seal radial gap and axial gap is preferred. This new concept results in a lower number of seal refurbishments, and better feedwater nozzle protection than that offered by the GE analysis.

NYPA Response Bl. The f easibility study cited was an internal GE consensus- which was not formally documented nor submitted to the NRC. In summary, the evaluation concluded the following:

The original pl'an for seal refurbishment was to replace the seals with new seals which would restore the axial gap.to original tolerances. In order to decrease the number of necessary refurbishments by further reducing rapid cycling fatigue, a new plan for refurbishment was established in which the seal and seal housing are replaced. This allows the axial and radial gaps to be restored, and the secondary seal housing interference fit to be re-established. This improved method allows FitzPatrick to refurbish seals twice instead of three times.

NRC Ouestion B2. Page 21 The two. curves in Figure 5-1 show that the new seal refurbish-ment concept has a higher usage factor than that.of the original seal refurbishment concept. Explain why the new seal refurbishment concept is preferred if it results'in a higher usage factor.

NYPA Response B2. The new seal refurbishment concept is preferred because only two refurbishments are required, instead of three. The usage factor is still acceptable-(< l.0), and FitzPatrick will need to do one less'sparger removal an'd refurbishment. Since seal refurbishment is in-vessel work and involves high radiation levels, the new concept will result in significant savings in radia^ ion exposure, plant down-time, and both material and labor costs.

NRC Ouestion C. " Effects of Reactor Water Cleanup Reroute on Feedwater Nozzle Fatigue Usage. James A. FitzPatrick Nuclear Power' Plant,"

General Electric, MDE-22-Ol85/DRFB13-lO9-3, January, 1985.

C1. -Page 8 The report stated that reactor water cleanup (RWCU) flow is a small portion of the total feedwater flow.

Provide the flow rate of RWCU and feedwater system under startup, shutdown, and scram conditions.

NYPA Response C1. As shown in Figure 3-2 of the report, RWCU flow = 72,470 lbm/hr FW flow = 10,444,100 lbm/hr @ 100% flow The feedwater flow varies as shown in Table 5-1 of the report.

The RWCU flow represents 0.69% of-the' rated feedwater flow. At conditions of high feedwater flow, the RWCU flow causes temperature changes of only a few degrees. For example, during Hot Standby or Turbine Roll which are significant system cycling transients, the feedwater flow is 25%, and RWCU mixing has.an effect of about 40F, which is insignificant.

During startup and shutdown', the RWCU flow temperature follows the reactor temperature. The same is true for the feedwater, so the relative flowrates are unimportant because the temperatures are approximately equal.

During scram, the feedwater flow drops as low as 3%, so the RWCU flow of 0.69% can have an effect in improving system cycling fatigue due to scram. However, the scram transient causes little fatigue on the'feedwater nozzle, so improvement due'to RWCU reroute is insignificant.

NRC Ouestion C2. Page 13 The evaluation presented in NEDO-21821 showed that the largest improvement in usage would be achieved by rerouting RWCU (i.e.,

a reduction from 0.70 to 0.46). However, in the FitzPatrick analysis, the reduction is from 0.814 to 0.801. Why is FitzPatrick improvement so low compared to NEDE-21821?

NYPA Response C2. The question appears to refer to Table 4-31 of NEDE-21821. The improvement is from 0.77 to 0.46. However, this is for a single sleeve design with zero leakage. FitzPatrick has a triple sleeve geometry, which shows, for 1.0 gpm leakage, an improvement from 0.54 to 0.48, or a difference of.0.06.' The FitzPatrick difference of 0.013 is Jess than 0.06 because the RWCU flow is less than the typical RWCU flow of 1% rated feedwater.

_ _ ._ , . .