ML021050447
| ML021050447 | |
| Person / Time | |
|---|---|
| Site: | Prairie Island  |
| Issue date: | 02/12/2002 |
| From: | NRC/NRR/DLPM |
| To: | Nuclear Management Co |
| References | |
| TAC MB4012 | |
| Download: ML021050447 (6) | |
Text
Followup Questions From February 12, 2002 Phone Call With Prairie Island Unit 2 "Information provided is reasonably accurate, but has not been validated using the normal NMC procedure for correspondence to the NRC"
- 1)
Provide more details (SG location, plug type, damage mechanism, etc.) on the tube(s) with leaking explosive plugs.
GEN LEG ROW COL RC VOLTS DEGREE MecCode %Value PERCENT CHANNEL LOCATION EOC Date 2
C 33 79 R33C79 CLTSP 46 46 M
01C
+ 0.0 4
8001 2
C 43 37 R43C37 CLTSP 47 47 M
02C
+ 0.0 4
8001
- 2)
Provide the results of the indications related to AVB wear, e.g. number of indications, estimation of the deepest wear scar, etc.
SG ROW COL LOCATION PCT-2002 PCT 2002 Growth 21 19 6
NV1 23 22 1
21 34 28 NV3 23 23 0
21 34 31 NV1 21 13 8
21 34 31 NV2 27 21 6
21 34 31 NV3 32 25 7
21 34 31 NV4 23 14 9
21 25 32 NV2 25 28
-3 21 34 32 NV2 21 22
-1 21 17 38 NV2 11 13
-2 21 28 45 NV2 32 33
-1 21 34 56 NV3 16 16 0
21 21 61 NV4 15 16
-1 21 23 61 NV4 24 24 0
SG ROW COL PCT - 2002 PCT - 2000 Growth 22 11 56 9
7 2
22 16 71 17 15 2
22 19 34 21 20 1
22 25 49 17 11 6
22 32 67 17 16 1
22 32 67 27 22 5
22 32 67 30 25 5
22 32 67 18 17 1
22 32 72 11 9
2 22 33 41 13 11 2
22 36 33 31 31 0
22 36 33 41 39 2
22 36 33 34 31 3
22 36 33 25 25 0
22 36 69 28 25 3
22 38 39 9
8 1
22 38 39 9
8 1
22 38 48 15 8
7 22 38 50 20 15 5
22 38 50 27 20 7
22 39 45 17 11 6
22 40 41 18 19
-1 22 40 41 23 25
-2 22 40 41 16 15 1
22 41 29 22 22 0
22 41 48 20 20 0
22 42 38 11 14
-3 22 42 38 17 19
-2 22 43 57 24 26
-2 22 43 58 18 17 1
- 3) Provide the results of the indications related to loose parts, e.g., location of indication, visual confirmation, final number of confirmed loose parts, etc.
Map Location Location Tube I
Label SG Column Row Description Wear (%) Retrieved Remarks Resolution A
21H 38-39 1 38-39 sludge rock 0
no smooth edges, light weight allowed I 2H 01approximately 1 inch long metal 21H 50-51 31-32 Istnp 1
0 no adhered to tube and tubesheet not a loose part C
22H 20 23-24 sludge rock 1
0 no smooth edges, light weight
'allowed either grnding shaving or weld seal D
22H 22 28-29 2.5 inch by 3/16 inch metal strip 0
nes n
removed E
22H 22-23 23 none 0
no couldn't access due to sludge history F
22H 45-46 30-31 sludge rock 0
no smooth edges, light weight allowed G
22H 47 32-33 none 0
no couldn't access due to sludge history 3/8 inch dia. X 1/2 inch long 22H 55 31 cylindrical metal object 0
yes on H25 list - AVB retainer bar end removed 3/8 inch dia. X 1 inch long machining remnant, crumbled on I
22H 54 31 machine cud 0
yes retneval removed J
22H 50-51 34 none 0
no nothing seen at this site inspected clear fhard sludge or object captured in K
22H 38 9
hard sludge 0enot a loose part
- 4)
Discuss the previous history of SG tube inspection results, including any look backs performed.
Previous history look backs are pertormed in the following instances:
A bobbin coil DSI indication at TSP's greater than 2.0 volts, MRPC roll transition I code indication greater than 1.5 volts, MRPC tubesheet crevice I code indication greater than 0.5 volts, any MRPC I code freespan indication greater than 0.3" long and all MRPC TSP I code indications not reported in the previous examination. In addition to these requirements May 2000 data was reviewed on all the in situ pressure test tubes along with any requests from the Steam Generator Integrity engineers to supplement information for the condition monitoring and operational assessment.
A look back on the PLP in both steam generators was performed. The following table summarizes the results of that review.
SG Row Col 2001 2000 1998 1997 1995 21 31 50 PLP PLP N/A PLP NDD 21 32 50 PLP PLP N/A NDD P
21 32 51 PLP PLP N/A NDD P
21 38 3
PLP NDD P
21 38 39 PLP NDD N
21 39 38 PLP NONDD N
22 23 20 PLP PLP PLP PLP 22 23 22 PLP PLP PLP PLP 22 23 23 PLP PLP PLP PLP 22 24 20 PLP PLP N/A PLP 22 28 54 PLP NDD NDD PLP 22 29 22 PLP NDD NDD NDD 22 30 45 PLP PLP PLP PLP 22 30 46 PLP PLP PLP PLP 22 31 46 PLP PLP PLP PLP 22 31 54 PLP PLP PLP NDD 22 31 55 PLP NDD NDD NDD 22 32 47 PLP PLP PLP PLP 22 33 47 PLP PLP PLP PLP 22 34 50 PLP PLP PLP NDD 22 34 51 PLP PLP PLP NDD N/A = Data not available
- = Data wasn't reviewed
- 4) Provide the results of the in-situ pressure tests in SG 22.
Eddy Length Width Leakage Max Current SG Row Col Indication Location Voltage Reason Inch Degree Result Pressure Change 21 3
17 SAI TRH+0.O7to+0.13 0.38 RTZ PWSCC 0.06 0
2816 No 21 9
18 MAI TRH+0.14to+0.24 0.46 RTZ PWSCC 0,10 0
2816 No 21 14 28 SAI 1BH+16.67to+16.86 0.35 Crevice ODSCC 0.19 0
2816 Yes 21 23 31 MAI 1BH+16.79to+17.29 0.66 Crevice ODSCC 0.50 0
2816 Yes 22 13 29 SAI TSH+0.25to+0.75 0.09 Free Span ODSCC 0.50 0
5256 Yes 22 2
31 SAI TRH+5.86to+6.24 0.34 Crevice PWSCC 0.38 0
2816 No 22 14 411 MAI 1BH+17.29to+17.75 0.08 Crevice ODSCC 0.46 0
2816 Yes 22 19 47 MAI TRH+17.82to+18.38 0.14 Crevice ODSCC 0.56 0
2816 Yes
- 6)
What are the plans, if any, for implementing the voltage-based criteria for Unit 2?
The implementation of the voltage based repair criteria is not required this outage. The results of bobbin coil inspections and rotating coil inspections at the tube support plates, for other than cold leg tube support plate thinning are:
SG Number of Distorted Maximum Voltage of Distorted Number/Type of Confirmed RPC Indications Support Plate Indications Support Plate Indications (All DSI were examined by RPC) 21 27 1.46 0
22 52 1.3 0
- 7)
During a phone call with the licensee for Prairie Island Unit 2 on February 12, 2002, the NRC staff inquired about the actions being taken to ensure that any significant flaws in the U-bends of low row tubes, including the apex and tangent point locations, will be detected. Significant flaws are considered those that fail or may fail to satisfy the NEI 97-06 tube integrity performance criteria prior to the next scheduled inspection. Specific questions aimed at addressing this issue are:
- a. What procedures are being implemented to ensure adequate data quality in the U-bend region of low row tubes, including apex and tangent point locations?
Upon completion of resolution on all low row u-bend examinations (both mid range and high frequency) the resolution analyst(s) shall calibrate, perform and record noise measurements in accordance with Reference 2.9 at 300 kHz and 400 kHz for the mid range probe and 600 kHz and 800 kHz for the high frequency probe. If any measurement on the mid range data exceeds the average of the EPRI qualification, the resolution analyst will report the run as RHF with a Quality code of QEN. If any measurement on the high frequency data exceeds the average of the EPRI qualification, the resolution analyst will measure the noise of the mid range data with a band pass filter (BPF) set at 23 Coefs., low cut at.0433 x RPM and high cut at.267 x RPM applied to the 300 kHz raw channel. If the results are less than the average of the EPRI qualification, both the primary and secondary resolution analysts will re-evaluate the mid range data using the BPF on the 300 kHz raw and process channels. If the results of the re evaluation is NDD, report the high frequency data as NDD with a Quality code of QEN. If the results are in excess of the EPRI qualification, report the high frequency data as TBP with a Quality code of QEN.
What are the acceptance standards for data quality in these areas?
Prairie Island Unit 2 2002 U Bend Noise Criteria EPRI ETSs # 99997.1 800 kHz Average 600 kHz Average EPRI ETSS # 96511.2 400 kHz Average 300 kHz Average 300 kHz Band Pass Filter Average Apex Vp-p 1.56 1.27 Apex Vp-p 1.22*
1.09*
1.09 Apex V v-m 0.54 0.37 Apex V v-m 0.41 0.40 0.40 Tangent Vp-p 2.50 2.04 Tangent Vp-p 1.62 1.49 1.49 Tangent V v-m 0.88 0.62 Tangent V v-m 0.62 0.65 0.65 Based on the ability to find a structural significant flaw in noise with a Vp-p of greater than 2.0 volts, the criteria used at Prairie Island for this outage is 2.0 volts
- b. What are the noise levels present in the U-bend region of low row tubes, including the apex and tangent point locations? Please provide average and
maximum values as well as a measure of the variability of the noise level about these averages.
The attached cumulative distribution plots (only fax copies for 21 SG are available for the phone call) show the average and spread in noise levels in Row 1 and Row 2 U-Bends at Prairie Island Unit 2. The population of noise levels at Prairie Island Unit 2 is equal to or less than the noise population in the EPRI data set in industry use for qualification of detection of degradation in low row U-Bends. Use of multiple probes and frequencies essentially limited the maximum noise level in each tube to less than (or equal to) the average noise level in the EPRI qualification data set. As discussed in 7a, the maximum peak to peak noise at 300khz and 400 khz was relaxed to 2 volts based on updated "must detect" signal requirements, noise phase angle studies and noise injection evaluations
- c. What signal to noise ratio is necessary to ensure a flaw in the U-bend region of a low row tube will be reliably detected and how has this been determined?
Flaw detection is not just a simple matter of signal to noise ratio. The type of noise must be considered.
High frequency random noise may obscure a flaw signal of equal or greater amplitude and in this case a 2 to 1 signal to noise ratio may be required for reliable flaw detection. The noise in Prairie Island U-Bends is low frequency cyclic noise that is very repeatable from one scan line to the next. It is primarily associated with tube ovality. Heat treatment increased noise levels somewhat but a comparison of pre and post heat treatment noise shows that the fundamental shape of the noise signals are unchanged. The amplitude of the low frequency noise signal was somewhat increased by heat treatment. In the case of flaw signals in the Prairie Island noise, a signal to noise ratio of 0.5 or less still remains detectable
- d. Figure 5 of the licensee's letter dated February 28 indicates that for crack lengths of 0.8 and 1.5 inches, the "should detect" voltages are approximately 1.6 and 1.4 volts, respectively, with the plus point at 400 KHz (using the "measured burst pressures" curve). Assess the detectability of such hypothetical voltage responses for the range of noise levels present at the apex and tangent point locations of the low row u-bends at Prairie Island Unit
- 2.
Two methods were used to verify that the "must detect" flaw voltage was easily detected in the worst case low row U-Bend noise at P12. The EPRI noise widget was used to increase the noise background to levels in excess of the worst tube at P12. Setttings were adjusted to match P12 noise characteristics and then the nosie amplitude was increased. The must detect flaw remained detectable beyond the noise level in the worst case tube. This approach was verified by adding a "must detect" flaw signal to the noise backgrounds of P12 U-Bends. The must detect flaw was easily detected in the most severe noise present at P12.