ML20236X100
| ML20236X100 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 07/29/1998 |
| From: | Rone A GENERAL PUBLIC UTILITIES CORP. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| 1940-98-20418, GL-88-01, GL-88-1, NUDOCS 9808070123 | |
| Download: ML20236X100 (23) | |
Text
..
a
[ '
GPU Nuclear. inc.
\\
Route 441 South NUCLEAR Post Office Box 480 Middletown, PA 17057-0480 Tel 717-944 7621 July 29, 1998 1940-98-20418 U. S. Nuclear Regulatory Commission Attn: Document Control Desk Washington,DC 20555 Gentlemen:
Subject:
Oyster Creek Nuclear Generating Station (OCNGS)
Docket No. 50-219 Facility Operating License No. DPR-16 Request for Approval of Altemate 17R Outage Inspections Related to Generic Letter 88-01 Intergranular Stress Corrosion Cracking (IGSCC) Commitments On July 8,1998, GPU Muclear announced that it had identified no buyer for the OCNGS and that n final decision on the future of Oyster Creek will be made sometime after the New Jersey Board of Public Utilities makes a decision on GPU's restructuring filing. Based upon these developments, GPU Nuclear has determined that some projects whose primary purpose was to keep the plant running through its licensed life will be deferred, some 17R Outage (September 1998) projects will be streamlined, and more effort will be focused on the eventual decommissioning of Oyster Creek.
The deferred work would include the planning for a refueling outage in the fall,2000 and certain modifications, inspections and testing not equired to support safe operation of the plant through the
)/
fall,2000.
The review of work which could safely be deferred one operating cycle or whose scope could safely be reduced has identified some specific 17R Outage IGSCC inspection activities committed to as part of NRC approved activities related to Generic Letter 88-01 (OCNGS Technical l
Specification 4.3.I).
i
\\
By this letter, GPU Nuclear requests the following summarized attemative initial scope for the h0 upcoming 17R Outage. Attachment I to this letter provides the detailed supporting basis for this request.
9800070123 980729*
I PDR ADOCK 05000219 P
PDR; L
u
.n
- 1. Reactor flead Cooling:
Inspect all seven welds, since they represent the highest risk of exceeding Code allowable depth if an indication ofIGSCC was missed in the last inspection. This is primarily due to the small diameter, wall thickness and environment.
- 2. Core Spray:
Inspect the two Category D welds, since they represent a higher risk of exceeding Code allowable depth if an indication ofIGSCC was
- missed in the last inspection and the fact that they are not stress improved. This is primarily due to the small diameter, wall thickness and environment.
- 3. Isolation Condenser:
Inspect no welds, because they were stress improved and have no recordable indications. A missed flaw in the last inspection is unlikely to exceed Code allowable depth at the end of Cycle 17.
- 4. Reactor Recirculation:
Inspect the two welds that contain IGSCC in the as-stress improved condition. The purpose is to verify that no change, as would be expected in the stress-improved condition and with Hydrogen Water
- Chemistry (HWC), in the flaw size has occurred.
- 5. Shutdown Cooling:
Inspect the one weld that is not stress improved and contains indications not interpreted to be IGSCC. Even though GPU Nuclear considers that the weld is protected by HWC, it does represent a
- slight risk that IGSCC may exist due to the ID indications.
- 6. ReactorWaterCleanup (RWCU):
Inside Second Containment Isolation Valve (CIV)
. Inspect three welds that represent one-half,i the welds that contain indications not interpreted to be IGSCC. Even though GPU Nuclear considers the welds protected by HWC, they do represent a slight risk that IGSCC may exist due to the ID indications.
Outside Second ClV Inspect no welds. GPU Nuclear considers them fully protected by HWC, and the fact that we have detected no IGSCC in any RWCU welds over the last 5 outages. Additionally, the four CIVs are being modified during 17R to resolve the concern about the ability to close the valves under blowdown conditions (i.e., a pipe break outside the second CIV).
r l
b i
i L
as 1
Should GPU Nuclear detect IGSCC in previously IGSCC-free welds, GPU Nuclear proposes to expand the inspection scope. The general criteria GPU Nuclear would use includes increasing the sample among welds of the same category, weld condition, and environment. For example, should GPU Nuclear find IGSCC in the Reactor Head Cooling,, Core Spray, Shutdown Cooling, and Recirculation system initial samples, GPU Nuclear would not perform any additional inspections since GPU Nuclear is doing 100% of the welds in the respective Category in those systems. Should GPU Nuclear detect IGSCC in the initial sample in the RWCU system, GPU Nuclear would inspect the remaining Category D welds containing indications other than IGSCC. If GPU Nuclear detects 2
IGSCC in the second sample, GPU Nuclear proposes to inspect the remaining Category D welds inside the second CIV.
Additionally, should GPU Nuclear detect significant IGSCC, in terms of either number of new welds containing or growth of existing flaws, GPU Nuclear will evaluate what additional actions are required and discuss them with the NRC.
GPU Nuclear recognizes the short timeframe for review that has been brought about by the recert Oyster Creek announcement and the burden this may place on the NRC Staff. Ilowever, GPU Nuclear does request an expedited review of this proposed alternative scope in order to finalize planning ofIGSCC inspections for the upcoming 17R Outage scheduled to begin in September, 1998.
l If you have any questions or comments on this matter, please contact Ron Zak, Corporate l
Regulatory Affairs at (973) 316-7035.
'trulyy G d l* ^'&
lv A. H. Rone Vice President and Director l
Engineering l
\\
l c: Administrator, Region 1 Senior Resident inspector Oyster Creek NRC Project Manager
- e I
1940-98-20418 Page 1 of 19 l
l ATTACHMENT 1 Supporting Basis for Alternate 17R Outage IGSCC Inspections i
t l
l 1
t
I
- e I940-98-20418 Attachment i Page 2 of 19
" L Introduction The Oyster Creek Nuclear Generating Station (OCNGS) is committed to performing inspections ofits stainless steel piping exposed to reactor coolant in accordance with its commitment to NRC's Generic Letter 88-01, "NRC Position On IGSCC In BWR AusteniCc Stainless Steel Piping." GPU Nuclear has been performing the inspections since the 12R Outage (1988).
Additionally, GPU Nuclear has taken action to minimize the potential for developing new IGSCC and growth of existing IGSCC. These actions include pipe replacement with resistant material, stress improvement (both induction heating (IllSI) and mechanical (MSIP)), weld overlay repairs, and hydrogen water chemistry (llWC). Also, GPU Nuclear has implemented j
water chemistry control to comply with the EPRI BW R Water Chemistry Guidelines.
The current GL 88-01 inspection scope for the upcoming 17R Outage,in September 1998,is 94 welds. The current estimated personnel exposure for this scope of work is about 15-20 Rem. A large portion of this exposure is associated with inspecting 10% of the reactor water cleanup (RWCU) welds outside the second containment isolation valve. These welds have never been inspected, are wrapped with asbestos insulation, and require substantial amounts of surface preparation for the inspection.
In April 1997, GPU announced that it was considering an early shutdown of the OCNGS in the year 2000 at the end of Cycle 17. On July 8,1998, GPU Nuclear announced that it had identified no buyer for the OCNGS and that a final decision or, the future of Oyster Creek will be made sometime after the New Jersey Board of Public Utilities makes a decision on GPU's restructuring filing. Based upon these developments, GPU Nuclear has determined that some projects whose pril ary purpose was to keep the plant running through its licensed life will be deferred, some 17R Out. : (September 1998) projects will be streamlined, and more effort will be focused on the eventual decommissioning of Oyster Creek. The deferred work would include the planning for a refueling outage in the fall,2000 and certain modifications, inspections and testing not required to support safe operation of the plant through the fall,2000.
l Because of the potential for early shutdown, GPU Nuclear has reviewed the scope of several projects to determine if all or a portion of the scope could be deferred to the next outage (18R) with little or no impact on the safe operation of the plant. GPU Nuclear has reviewed the scope of the GL 88-01 program for 17R, the inspection history (i.e., a backward look) and the mitigating actions GPU Nuclear has taken (i.e., looking forward). GPU Nuclear has concluded that the scope for 17R could be reduced with minimal risk and achieve a dose savings of about 10-15 Rem.
This evaluation reviews the OCNGS IGSCC inspection history, the current 17R inspection scope, water chemistry performance (including IlWC), and evaluates the potential for crack growth from the last inspection of the welds using both the GL 88-01 crack growth rule and new crack growth irodels developed by the BWRVIP which incorporate the impact of water chemistry on crack
1940-98-20418 Page 3 of 19 growth. Based on this evaluation, GPU Nuclear proposes an alternative scope for GL 88-01 for the 17R Outage. The 17R Outage is currently scheduled to begin on September 26,1998. If the NRC approves the scope reduction and, afler the outage, the decision for early closure is reversed, the j
inspections that would have normally been performed in 17R will be performed in 18R (in other l
words GPU Nuclear would consider this an approval for a deferral of the inspections).
Additionally, GPU Nuclear will continue to comply with the EPRI BWR Water Chemistry Guidelines and continue IIWC.
I 1
II.
Background
In 1988, the NRC issued Generic Letter (GL) 88-01, "NRC Position on IGSCC in BWR Austenitic Stainless Steel Piping." As the title implies, GL 88-01 provided the NRC position on IGSCC in BWR piping, which consisted of detailed positions on 13 separate issues covering IGSCC in BWR piping, including materials, water chemistry, inspections, repairs, and leakage monitoring. The GL also required the licensees to incorporate a change in the Technical Specifications (TS), in the section addressing insersice inspection (ISI), committing to the staff positions identified in the GL.
The OCNGS TS Amendment 154 incorporated the GPU Nuclear commitment to follow GL 88-01 l
or in accordance with attemate methods approved by the NRC stafr(section 4.3.1). Reactor coolant leakage monitoring is addressed in TS section 3.3.D.
The last fomial update to the OCNGS GL 88-01 program is TR-050 Rev. 3. This report was submitted to NRC before the 13R Outage. GPU Nuclear identified the plans for the 13R Outage and beyond, which included:
Replacement of the Isolation Condenser piping outside the drywell with Type 316NG stainless steel in 13R. All the welds were stress improved using MSIP in 14R.
l Replacement of the drywell penetration piping, which contained inaccessible welds in the scope of GL 88-01, in the Isolation Condenser and Reactor Water Cleanup systems, with piping that contained no welds.
i I
Complete inspections of all Category G welds inside containment, except eight Recirculation e
l safe ends were completed in 14R after stress improvement.
Inspect 10% of the RWCU welds outside the second containment isolation valves (CIVs) that e
are in the scope of GL 88-01.
There has been additional correspondence between GPU Nuclear and the NRC regarding the GL 88-01 program. This additional correspondence includes NRC's agreeing to our using Code Case 1
i
en 1940-98-20418 l
Page 4 of 19 N-481 for the Recirculation pump casing and suction elbow welds, which are statically-cast Type 316 stainless steel, and discussion regarding the requirement to inspect 10% of the RWCU l
Category G welds outside the second CIVs. The latest letter from the NRC was on GPU Nuclear's request for exempting these RWCU welds from future inspections. The NRC declined the request because, at the time, GPU Nuclear had not satisfactorily resolved the question of the ability of the gate valves to isolate under blowdown conditions (i.e., in the event of a pipe break outside the second CIV). By GPU Nuclear letter 6730-97-2198, dated August 5,1997, GPU Nuclear l
committed to modifying the four valves in question during the 17R Outage to satisfactorily resolve the CIV concern. These four CIVs remain scheduled to be modified during the 17R Outage.
III. OCNGS GL 88-01 Inspection Results GPU Nuclear has been inspecting the OCNGS GL 88-01 welds since the 12R Outage. Actually, GPU Nuclear began programmatic inspection of this piping in the 11R Outage in 1986. The 11R Cutage was the first outage during which inspections used personnel and procedures qualified at the l
EPRI NDE Center under the three-party agreement (in accordance with Generic Letter 84-11). Of the about 380 welds in the GL 88-01 scope, including 85 in the RWCU system outside the second CIV, Oyster Creek currently has 11 welds in service with indications ofIGSCC, Nine were repaired with full structural weld overlays (four in Core Spray, four in Recirculation, and one in Shutdown Cooling) and two are in service without repair (Recirculation system welds NG-C-9A and NG-D-18; both were stress improved before inspection found IGSCC). All of the cracked welds contained indications that required further evaluation during the inspection. That is, the initial inspection detected indications associated with geometry, counterbore, root, etc.. The additional inspections required as a result of these indications led to the conclusion that these welds contained IGSCC. All of these welds contained recordable indications in addition to the IGSCC.
All the repairs and use-as-is dispositions have been reviewed and accepted by NRC as required in GL 88-01. In the 12R Outage, during the pre-startup pressure test, there was one weld found leaking in a piping component that connects the Reactor llead Cooling line to the closure head; it was removed and replaced. Before the 13R Outage, about 30 of the 129 welds in the isolation Condenser system outside the drywell had indications ofIGSCC; these were all removed from service during the pipe replacement project during 13R mentioned above.
Therefore, of the 298 original welds that are currently in the scope of GL 88-01, IGSCC has been detected in only 12 (about 4%). No new indications of cracking were detected in the last three utages (14R (1992,60 welds),15R (1994,160 welds), or 16R (1996,79 welds)). No indications l
ofIGSCC has been found in RWCU, which is not stress improved and including the welds outside i
the second CIV, nor in the Isolation Condenser system piping inside the drywell.
1940-98-20418 Page 5 of 19 Ill.
Current 17R Outage GL 88-01 Scope The current scope, following the requirements of GL 88-01 for the OCNGS 17R Outage is:
Recirculation:
3 welds, including one weld overlay repaired (WOR) and one use-as-is (stress improved)
Shutdown Cooling:
3 welds, including two welds stress improved (Sla) in 15R and one i
Category D (inadequate access for SI)
RWCU:
47 welds, including 9 Category G welds outside the seemd clV, and l
38 Category D inside the second CIV i
Core Spray:
8 welds, including two Category Ds (inadequate access for SI); the remainder are Category C (Sid and inspected twice)
Isolation Condenser: 26 welds, inside the drywell, all were SId in 15R and were inspected in 15R: this is the second post-Si inspection that would result in these welds' being classified as Category C Closure Head:
7 welds, all Category D last inspected in 15R.
There are no creviced welds (e.g., thermal sleeves) in the 17R scope.
The specific welds, whether stress improved, and the inspection history since the 11R Outage are listed in Table 1. As can be seen, except for the RWCU welds outside the second CIV, all the welds in the 17R scope have been inspected at least twice using qualified personnel and procedures.
Additionally, except for the RWCU welds outside the second CIV, all but 7 of the 84 welds were last inspected in the 15R Outage (1994). These 7 welds are all Category C welds (stress improved and inspected twice after the SI).
The currently estimated radiation exposure to perform these inspections is about 15 Rem. A large part of this dose is attributable to inspecting the RWCU welds outside the second CIVs. Most of these inspections are on welds located near the regenerative heat exchangers, which are relatively high dose rate areas, and require the removal ofasbestos insulation and significant surface preparation for performing meaningful inspections.
I l
V.
Water Chemistry
1940-98-20418 I
l Page 6 of 19 OCNGS Water Chemistry Performance GPU Nuclear has reviewed the water chemistry history of OCNGS since 1984. This was the first cycle after an extended outage. The first outage after that was 11R, during which GPU Nuclear performed weld inspections to GL 84-11. Recall that all welds inside the second isolation valves were inspected at least once by the end of the 13R Outage, except for some of the recirculation inlet and outlet nozzle safe end welds, which were completed in the 14R Outage. The average conductivity, while at power (>0%), for this period was 0.09 S/cm.
GPU Nuclear also evaluated the water chemistry history from the 15R Outage to present. This represents the operating period from the last outage in which most of the welds were inspected.
The average conductivity during the period from the 15R Outage was 0.09 pS/cm; 97% of this time, the conductivity was 0.3 S/cm or less,90% of this time, the conductivity was 0.1 S/cm or less. The short periods where conductivity exceeded 0.3 S/cm were associated with startups and shutdowns or the shutting down of the hydrogen feed, which creates a chromate spike. These periods were of short duration.
OCNGS Hydrocen Water Chemistry Perfonnance Based on the results of the Hydrogen Water Chemistry (llWC) Mini Test performed in January 1986, a threshold level of hydrogen injection necessary to produce mitigating conditions (ECP less than or equal to -230 mV SHE and conductivity less than 0.3 S/cm)in the Recirculation and RWCU Systems at Oyster Creek was determined. Based on that value, a hydrogen injection system was designed and installed at Oyster Creek. Although injection first began in Cycle 12, the first full cycle of HWC was Cycle 13.
I During Cycle 13, hydrogen was injected a high percentage (estimated to be 70 to 80%) of the time that the plant was operating, but equipment difficulties, especially with flow contre elements, limited the operating time under mitigating conditions to a losv percentage (ava lability estimated to be 10 to 15%) of the 18-month cycle. Despite the equipment difficulties and the low HWC Availability in Cycle 13,14R inspections results for Recirculation and RWCU Systems showed no new defects and no changes from the previous inspection in 13R.
With the resolution of the flow element problems in Cycle 13, HWC performance durir.g Cycles 14 and 15 was very good with calculated availability very close to 90%. Inspection results in i
15R and 16R again showed no changes in Recirculation or RWCU system welds compared to l
previous inspections. Monitoring of HWC during these two cycles was accomplished by ECP
{
measurements in an external autoclave fed by flow from the "A" Recirculation loop. During this l
three and a half year operating period, correlations were established between measured ECP, 1
indicated hydrogen injection flowrate, and continuously monitored main steam line radiation l
levels and Recirculation dissolved oxygen concentrations. These correlations pennitted the i
l i
t
1940-98-20418 Page 7 of 19 establishment of secondary parameters that were used to control hydrogen injection and IlWC performance during periods when ECP measurement was not possible.
Near the end of Cycle 15, prior to 16R, because of a desire to limit personnel radiation exposure resulting from maintenance of the ECP measurement system, GPU Nuclear elected to abandon the ECP autoclave and to control HWC perfonnance by adjusting hydrogen injection to limit Recirculation dissolved oxygen to a maximum value of 2 ppb. This dissolved oxygen control value is based on the correlations established in the previous two operating cycles. During Cycle 16 to date, HWC availability, calculated from continuously measured oxygen concentrations, has been greater than 95%. GPU Nuclear will continue hydrogen through the end of Cycle 17.
VL Cr ack Growth This section provides a qualitative discussion on the existence of flaws and the initiation and growth of flaws in the welds in the current GL 88-01 scope for the 17R Outage since their last inspection in 15R Outage. It also provides a discussion of the uninspected RWCU welds outside the second CIV.
Crack Growth Rates GL 88-01 provides an equation used by the staff to evaluate crack growth ofIGSCC in BWR piping welds. Generally, this equation has been used to disposition flaws found in larger-diameter piping (12-inch and greater nominal pipe diameter). Flaws detected in smaller-diameter piping have generally required repair or replacement of the affected piping. This is because the residual stresses from welding in small diameter piping tend to be more linear from ID to OD (highly tensile on the ID to highly compressive on the OD). And, even though the stress contribution to crack growth becomes less tensile as the crack grows through the wall, the contribution from the j
operational stresses results in a positive stress contribution to stress intensity at the crack tip. This I
results in a tendency for cracks in smaller-diameter piping to be driven through-wall.
1 l
In larger-diameter piping, the residual stress from welding is still highly tensile on the ID; however, it becomes less tensile fairly quickly as you progress from the ID to the OD and, in fact, becomes compressive at around 20% through-wall and reaches maximum compression at about 40%
through-wall (see Figure 3 of NUREG-0313, Rev. 2, Appendix A). The operational stresses are generally not large enough to overcome the compressive residual stress at this point, and one would l
expect crack growth to stop. This has been the experience to date in non-creviced welds, where cracks generally have not exceeded 50% through-wall. In fact, at Oyster Creek, no crack in the Recirculation system piping has exceeded 50% through-wall.
l l
_, _ _ _ _ - - - - - - - - - - ~ - - - - - - - _ _ _ - -
1940-98-20418 Page 8 of 19 The NRC equation is da/dt = 3.59E-8 x K " inches per hour, where da/dt is the rate of crack l
2 growth (through-wali) and K is the stress intensity at the through-wall location of the crack tip.
This equation was established by reviewing crack growth data available at the time and was i
somewhat bounding for expected " normal" water chemistry effects. However, it clearly is conservative with u.;pect to the effects of the much-improved water chemistry controls implemented since the GL was issued and the implementation of hydrogen water chemistry at many BWRs, including Oyster Creek.
1 Since the discovery of cracking in the BWR shrouds, a substantial effort has been expended to more precisely understand crack growth. This resulted in the issuance of a report by the BWRVlP (Evaluation of Crack Growth in BWR Stainless Steel Internals (BWRVIP-14)). While the title says i
it is for internals, it does have applicability to BWR stainless steel piping, since the specimens evaluated are essentially independent of the location of the components with fluences less than 10 '
2 2
n/cm (E>l MeV). The equation denloped af ter reviewing all the data (equation 3-2) takes into account stress intensity (as the NRC equation does); however, it also accounts for conductivity, electro-chemical potential, and the temperature. Testing has shown that conductivity and ECP have a significant impact on crack growth. BWRVIP-14 also established a factor of 10 increase in the calculated value to reflect the 95* percentile model.
For the purposes of these discussions, the following comments apply:
- 1. For pipe diameter 12" and greater, the residual stress distribution is as discussed in NUREG-0313 Rev. 2, Appendix A. For less than 12", the residual stress is considered linear from the ID to the OD (i.e.,30 ksi on the ID, zero at mid-wall, and -30 ksi on the OD).
- 2. The operating stress is 7.5 ksi (typical pressure, deadweight and thermal stress).
- 3. Stress intensity is the sum of the stress intensity contribution from residual and operating stresses at the crack tip.
- 4. The discussion encompass two starting points for crack growth:
A. A 10% flaw exists, which would represent our missing a flaw requiring further analysis for flaw growth as required by the Code during the last inspection (actually, j
the minimum allowable flaw size allowed for surface flaws found during ISI per l-Table IWB-3514-2 is 10.6% for a 1-inch thick weld). Missing a 10% through-wall flaw is considered highly unlikely, and B. a 0.001-inch flaw exists, which represents a crack " initiating" at the time of startup following the last inspection.
w_________:___________.
l l
1 1940 98-20418 l
Page 9 of 19
- 5. The flaws are 360 of the pipe circumference.
l
- 6. No credit is taken for stress improvement. This is consen'ative because stress improvement has been shown to be extremely effective in mitigating both the initiation of new flaws and further propagation of existing flaws (less than 30% through-wall).
J
- 7. The allowable flaw depth at the end of the operating period (i.e., September 2000) is 42% through-wall as allowed by Section XI, IWB-3640 for shielded metal arc welds.
l The 42% value is from Table IWB-3641-6 for a weld made with shielded metal are j
weld material under emergency and faulted conditions. This is the maximum through-l wall flaw allowed for a stress ratio s 1.2, which represents an operating stress s 29 ksi.
l
- 8. The operating time is considered as 365 days per year for the operating period evaluated (i.e.,6 years). This adds a level of conservatism to the evaluation in that Oyster Creek's operating cycle capacity factor has been around 90% and the refueling outages in the period have lasted about 45 days.
l No crack growth analysis was performed for discussion purposes for the Recirculation system l
piping welds. This is because the two welds with IGSCC have been monitored closely since detection and there has been no apparent growth in either. An expanded discussion is provided in the Recirculation section.
Reactor Head Cooline (4" NPS. Schedule 80 - 0.337" nominal thickness)
These 7 welds have all been inspected twice to GL 88-01 requirements, four of them three times.
Six of the seven have no recordable indications (NRI). The environment is steam. GPU Nuclear assumes a conductivity of 0.10 S/cm and an ECP of 0 mV, which correspot;ds to an oxygen content of about 0.2 ppm. This ECP would account for an expected higher ECP above the core than entering the core, even with HWC, due to the existence of short-lived oxidants such as peroxides.
If one were to assume a 10% flaw existed, the GL 88-01 crack growth equation predicts through-wall penetration in about one year. The BWRVIP-14 model predicts reaching 42% through-wall in about 1 year, and through-wall in 3 % years.
Assuming a 0.001-inch flaw exists, the GL 88-01 crack growth equation predicts through-wall penetration in about three years, whereas the BWRVIP-14 model predicts reaching 42% through-I wall in about 10 years.
I Core Snrav (8" NPS. Schedule 80 - 0.500" nominal thickness)
1940-98-20418 Page 10 of 19 These 8 welds have all been inspected twice to GL 88-01 requirements, four of them three times.
Six of the eight were IHSId in 12R (1988); all six of these have been inspected twice since the stress improvement and are now Category C welds. Six of the eight have no recordable indications (NRI), including the one weld that has not been stress improved. The environment is reactor water.
GPU Nuclear assumes a conductivity of 0.10 S/cm and an ECP of 0 mV, which corresponds to an oxygen content of about 0.2 ppm.
If one were to assume a 10% flaw existed, the GL 88-01 crack growth equation predicts through-wall penetration in about one year. The BWRVIP-14 model predicts reaching 42% through.-wall in about one year; through-wall in about 5 years.
Assuming a 0.001-inch flaw exists, the GL 88-01 crack growth equation predicts through-wall penetration in about 3 years, whereas the BWRVIP-14 model predicts reaching 42% through-wall in over 10 years.
Isolation Condenser (10" NPS. Schedule 80 - 0.594" nominal thickness)
These 26 welds have all been inspected twice to GL 88-01 requirements, eight of them three times.
All 26 were MSIPd in 15R (1994). All 26 have no recordable indications (NRI) following the stress improvement in 15R. The environment is reactor water on the condensate return side and steam on the steam supply side. GPU Nuclear assumes a conductivity of 0.10 pS/cm and an ECP of 0 mV, which corresponds to an oxygen content of about 0.2 ppm.
If one were to assume a 10% flaw existed, the GL 88-01 crack growth equation predicts through-wall penetration in about one year. The BWRVIP-14 model predicts reaching 42% through-wall in about 2 years; through-wall in about 4 % years.
Assuming a 0.001-inch flaw exists, the GL 88-01 crack growth equation predicts through-wall penetration in about 3 years, whereas the BWRVIP-14 model predicts reaching 42% through-wall in well over 10 years.
Shutdown Cooline (14" NPS. Schedule 80 - 0.750" nominal thickness and 10" NPS. Schedule 80 -
0.594" nominal thickness)
These 3 welds have all been inspected twice to GL 88-01 requirements, one of them three times.
Two were MSIPd in 15R (1994). One has no recordable indications (NRI) following the stress improvement in 15R. The 10" weld is the connection to the Isolation Condenser return line and was not stress improved because of the geometry. These three welds comprise the tee (14x14x10) that connects the Recirculation system, Isolation System return and the normally closed valve of the Shutdown Cooling system. As such, the environment is reactor water. And, because of the turbulence in this tee from the Recirculation system, GPU Nuclear expects that all three welds are f
l l
4
\\
1940-98-20418 Page ll of l9 l
exposed to the IIWC conditions in the Recirculation. Therefore, GPU Nuclear assumes a conductivity of 0.10 S/cm and an ECP of-200 mV, which corresponds to an oxygen content of l
about 2 ppb.
In the 14" welds, if on. were to assume a 10% flaw existed, the GL 88-01 crack growth equation predicts through-wall penetration in about 4 years. The BWRVIP-14 model predicts reaching 42%
through-wall in weh over 10 years.
In the 14" welds, assuming a 0.001-inch flaw exists, the GL 88-01 crack growth equation predicts through-wall penetration in about 5 % years, whereas the BWRVIP-14 model predicts reaching l
42% through-wall in well over 10 years.
In the 10" weld, if one were to assume a 10% flaw existed, the GL 88-01 crack growth equation predicts through-wall penetration in about 1 year. The BWRVIP-14 model predicts reaching 42%
through-wall in 7 % years.
In the 10" weld, assuming a 0.001-inch flaw exists, the GL 88-01 crack growth equation predicts through-wall penetration in about 4 years, whereas the BWRVIP-14 model predicts reaching 42%
through-wall in well over 10 years.
Recirculation (26" OD x.982 min. wall)
There are three welds in the GL 88-01 scope for 17R. One is the last Category C weld to be inspected in the 10 year period following the second inspection after stress improvement. It was last inspected in 12R (1988). This weld has had no recordable indications. The second is a full structural weld overlaid weld that has been inspected twice since the overlay was applied, not including the post-overlay baseline inspection. This was the weld from which a plug was removed for destructive evaluation in 12R. There has been no cracking detected in the weld overlay. The i
third is a weld that was stress improved containing IGSCC that is less than 30% through-wall; the cracking was detected afler IllSI was applied. The deepest part of the indication is about 22%
i through-wall. It has been inspected four times since the IGSCC call was made. GPU Nuclear has been tracking this weld as Category F for the purpose of cWsely monitoring it since it is close to the 30% limit allowed by the GL There has been essentially no change since the original sizing call.
The environment is reactor water and GPU Nuclear considers the system fully protected by IIWC, Therefore, GPU Nuclear considers that flaw initiation and growth is mitigated (i.e., there should be little or none).
Reactor Water Cleanun (6" NPS. Schedule 80 - 0.432" nominal thickness)
Inside Second CIV
1940-98-20418 Page 12 of 19 These 38 welds have all been inspected twice to GL 88-01 requirements, ten of them three times.
Thirty-two have no recordable indications (NRI). The environment is reactor water with IlWC.
GPU Nuclear assumes a conductivity of 0.10 pS/cm and an ECP of-200 mV, which corresponds to an oxygen content of about 2 ppb.
Ifone were to assume a 10% flaw exi: ed, the GL 88-01 crack growth equation predicts through-wall penetration in about one year. The BWRVIP-14 model predicts reaching 42% through-wall in l
about 1 year.
Assuming a.001-inch flaw exists, the GL 88-01 crack growth equation predicts through-wall j
penetration in about 3 years, whereas the BWRVIP-14 model predicts reaching 42% through-wall in well over 10 years.
Outside Second CIV There are nine welds not previously inspected in the 17R GL 88-01 scope. To date, GPU Nuclear has inspected 38 welds outside the second CIV. Of these,35 have had no recordable indications.
i The environment is the same as the system inside the second CIV; therefore, the crack growth estimates are the same.
i VII. Discussion Summary Key points from the earlier sections include:
1.
The crack growth equation from GL 88-01 is conservative for evaluating existing IGSCC in safety-related piping. Ilowever, it is excessively conservative for predicting crack growth immediately following assumed initiation.
2.
The crack growth equation from GL 88-01 accounts only for the stress dependency for l
crack growth. The BWRVIP-14 equation accounts not only for stress, but also water chemistry and ECP, both of which have a significant impact on crack growth rates.
l Therefore, the BWRVIP-14 is more realistic than the GL 88-01 equation.
l 3.
Using BWRVIP-14, only smaller diameter piping (<l2-inch NPS) welds show a likelihood l
of having a 10% through-wall flaw exceeding the Code allowable through-wall depth at the end of operation in September 2000. Ilowever, this would apply only to non-Sid welds.
Reinspection of uncracked, mitigated welds during the last three outages has detected no new indications ofIGSCC in stress-improved welds.
L__-_-___--___-
1940-98-20418 1
j Page 13 of 19 4.
Stress improvement has been an effective mitigator of continued growth of shallow IGSCC at Oyster Creek. There has been essentially no change in the flaw dimensions for the two Recirculation system welds with unrepaired IGSCC in a stress improved weld.
5.
There is no weld with no recordable indications from earlier GL 88-01 inspections that has developed IGSCC foll<. wing those inspections. Every weld with IGSCC at Oyster Creek also contains root ge,unetry, counterbore, root condition, and etc. indications.
6.
Water chemistry control at Oyster Creek since the 15R Outage has been excellent, averaging 0.09 S/cm..
7.
Ilydrogen water chemistry performance since 15R has been excellent with availability around 90%
VIII. Proposal for Alternative GL 88-01 Scope for the 17R Outage Based on the above evaluation, GPU Nuclear proposes the following alternative initial u: ope for the upcoming 17R Outage.
- 1. ReactorIIead Cooling:
Inspect all seven welds, since they represent the highest risk of exceeding Code allowab!c depth if an indication ofIGSCC was missed in the last inspection. This is primarily due to the small diameter, wall thickness and environment.
- 2. Core Spray:
Inspect the two Category D welds, since they represent a higher risk of exceeding Code allow *able depth if an indication ofIGSCC was missed in the last inspection and the fact that they are not stress improved. This is primarily due to the small diameter, wall thickness and environment.
}
- 3. Isolation Condenser:
Inspect no welds, because they were stress improved and have no recordable indications. A missed flaw in the last inspection is l
unlikely to exceed Code allowable depth at the end of Cycle 17.
i
- 4. Reactor Recirculation:
Inspect the two welds that contain IGSCC in the as-stress improved condition. The purpose is to verify that no change, as would be expected in the stress improved condition and with Ilydrogen Water Chemistry (IIWC), in the flaw size has occurred.
I i
j
F i
I 4
I 1940-98-204i8 Page 14 of 19 I
- 5. Shutdown Cooling:
Inspect the one weld that is not stress improved and contains indications not interpreted to be IGSCC. Even thouj;h GPU Nuclear 1
l considers that the weld is protected by HWC, it does represent a slight risk that IGSCC may exist due to the ID indications.
l
- 6. Reactor Water Cleanup (RWCU):
Inside Second Containment Isolation Valve (CIV) i inspect three welds that represent one-half of the welds that contain l
indications not interpreted to be IGSCC. Even though GPU Nuclear j
considers the welds protected by HWC, they do represent a slight risk that IGSCC may exist due to the ID indications.
Outside Second CIV
(
Inspect no welds. GPU Nuclear considers them fully protected by HWC, and the fact that we have detected no IGSCC in any RWCU I
welds over the last 5 outages. Additionally, the four CIVs are being i
modified during 17R to resolve the concem about the ability to close the valves under blowdown conditions (i.e., a pipe break outside the second CIV).
A listing of the current GL 88-01 scope and the proposed scope in provided in Table 2.
Samnie Expansion Should GPU Nuclear detect IGSCC in previously IGSCC-free welds, GPU Nuclear proposes to expand the inspection scope. The general criteria GPU Nuclear would use includes increasing the sample among welds of the same category, weld condition, and environment. For example, should l
GPU Nuclear find IGSCC in the Reactor Head Cooling, Core Spray, Shutdown Cooling, and i
Recirculation system initial samples, GPU Nuclear would not perform any additional inspections I
since GPU Nuclear is doing 100% of the welds in the reswtive Category in those systems. Should GPU Nuclear detect IGSCC in the initial sample in the RP.?U system, GPU Nuclear would inspect the remaining Category D welds containing indications other than IGSCC. If GPU Nuclear detects IGSCC in the second sample, GPU Nuclear proposes to inspect the remaining Category D welds inside the second CIV.
w
i j
1940-98-20418 i
Page 15 of19 l
Additionally, should GPU Nuclear detect significant IGSCC, in terms of either number of new welds containing or growth of existing flaws, GPU Nuclear will evaluate what additional actions l
are required and discuss them with NRC.
i l
l l
I t
I
1940-98-20418 Page 16 of19 TABLE 1 Inspection History of 17R Outage GL 88-01 Scope System: Closure Head
- llR 12R
' 13R 14R 15R 16R 17R.
17R Weld Number sis Insp.
Insp.
Insp.
Insp.
Insp.
Insp.
Insp.
Cat MV-5-001 No No No Yes No
.Yes No Yes D
MV-5-001 A No No Yes Yes No Yes No Yes D
MV-5-002 No No Yes Yes No Yes No Yes D
NR022-576 No No Yes Yes No Yes No Yes D
NR024-576 No No No Yes No Yes No Yes D
NR026-576 No No Yes Yes No Yes No Yes -
D RHC-HS-002 No No No Ye.J No Yes No Yes D
System: Core Spray
' l l R-12R 13R -
14R 15R 16R 17R 17R
Insp.
Insp.
Insp.
Insp.
Insp.
Insp.
Cat NZ-3-039 12R No Yes No Yes No No Yes C.
NZ-3-040 12R Yes Yes No Yes No No Yes C
NZ-3-042 12R Yes Yes No Yes No No Yes C
NZ-3-082 No Yes No Yes No Yes No Yes D
NZ-3-083 No Yes No Yes No Yes No
. Yes D
NZ-3-084 12R No Yes No Yes No No Yes C
NZ-3-086 12R No Yes No Yes No No Yes C _
NZ-3-088 12R No Yes No Yes No No Yes C
System: Isolation Condenser l
llR
.12R 13R 14R 15R 16R 17R 17R-Weld Number s
' SI Insp.
Insp.
Insp.~
.Insp.
Insp.
Insp.
Insp.
Cat F.
NE-2-052 15R No No Yes No Yes No Yes Cl NE-2-053 15R Yes No Yes No Yes No Yes Cl NE-2-054 15R No No Yes No Yes No Yes Cl NE-2-054A 15R No No Yes No Yes No Yes Cl NE-2-055 15R No No Yes No Yes No Yes Cl 1
1940-98-20418 Page 17 of 19 TABLEI
_ Inspection History of 17R Outage GL 88-01 Scope NE-2-056 15R No No Yes No Yes No Yes Cl NE-2-057 15R No Yes Yes No Yes No Yes Cl NE-2-058 15R No Yes Yes No Yes No Yes Cl NE-2-058A 15R No Yes Yes No Yes No Yes Cl NE-2-059 15R No No Yes No Yes No Yes Cl NE-2-060 15R No No Yes No Yes No Yes Cl NE-2-061 15R No Yes Yes No Yes No Yes Cl NE-2-062 15R No No Yes No Yes No Yes Cl NE-2-063 15R No Yes Yes No Yes No Yes Cl NE-2-063A 15R Yes No Yes No Yes No Yes Cl NE-2-116 15R Yes No Yes No Yes No Yes Cl NE-2-117 15R No No Yes No Yes No Yes Cl NE-2-I l 8 15R No Yes Yes No Yes No Yes Cl NE-2-119 15R No Yes Yes No Yes No Yes Cl NE-2-120 15R No Yes Yes No Yes No Yes Cl NE-2-121 15R No No Yes No Yes No Yes Cl NE-2-122 15R Yes No Yes No Yes No Yes Cl NE-2-123 15R No No Yes No Yes No Yes Cl NE-2-124 15R No Yes Yes No Yes No Yes Cl NE-5-201 15R Yes No Yes No Yes No Yes Cl NE-5-209 15R No No Yes No Yes No Yes Cl System: Reactor Water Cleanup (RWCU) llR 12R 13R 14R 15R 16R 17R 17R Weld Number SI Insp.
Insp.
Insp.
Insp.
Insp.
Insp.
Insp.
Cat ND-1-001 No Yes No Yes No Yes No Yes D
ND-1-003 No Yes No Yes No Yes No Yes D
ND-1-004 No No Yes Yes No Yes No Yes D
ND-1-004A No No No Yes No Yes No Yes D
ND-1-004B No No Yes Yes No Yes No Yes D
ND-1-004C No No No Yes No Yes No Yes D
ND-1-005 No Yes No Yes No Yes No Yes D
ND-1-006 No Yes No Yes No Yes No Yes D
i ND-1-007 No No Yes Yes No Yes No Yes D
[
ND-1-008 No No No Yes No Yes No Yes D
ND-1-009 No No No Yes No Yes No Yes D
f
(
l940-98-20418 Page 18 of 19 TABLE 1 Inspection History of 17R Outace GL 88-01 Scope ND-1-010 No No Yes Yes No Yes No Yes D
ND-1-011 No No No Yes No Yes No Yes D
ND-1-012 No No No Yes No Yes No Yes D
ND-1-013 No No Yes Yes No Yes No Yes D
ND-1-020 No No No Yes No Yes No Yes D
ND-1-021 No No No Yes No Yes No Yes D
ND-1-023 No No No Yes No Yes No Yes D
ND-1-024 No No No Yes No Yes No Yes D
ND-1-025 No No No Yes No Yes No Yes D
ND-1-027 No No No Yes No Yes No Yes D
ND-1-027A No No No Yes No Yes No Yes D
ND-1-028 No No No Yes No Yes No Yes D
ND-1-029 No No No Yes No Yes No Yes D
ND-10-001 No No Yes Yes No Yes No Yes D
ND-10-002 No No Yes Yes No Yes No Yes D
ND-10-003 No No No Yes No Yes No Yes D
ND-10-004 No No No Yes No Yes No Yes D
ND-10-005 No No Yes Yes No Yes No Yes D
ND-10-006 No No Yes Yes No Yes No Yes D
ND-10-007 No Yes No Yes No Yes No Yes D
ND-10-008 No Yes No Yes No Yes No Yes D
ND-10-009 No Yes No Yes No Yes No Yes D
ND-10-010 No Yes No Yes No Yes No Yes D
ND-10-011 No No Yes Yes No Yes
'No Yes D
ND-10-017 No No No Yes No Yes No Yes D
ND-10-018 No No No Yes No Yes No Yes D
ND-10-020 No No No Yes No Yes No Yes D
ND-2-009 No No No No No No No Yes G
ND-2-010 No No No No No No No Yes G
ND-2-011 No No No No No No No Yes G
ND-2-012 No No No No No No No Yes G
ND-2-013 No No No
'No No No No Yes G
ND-2-014 No No No No No No No Yes G
ND-2-015 No No No No No No No Yes G
ND-2-016 No No No No No No No Yes G
, ND-2-017 No No No No No No No Yes G
c-______
1940-98-20418 Page 19 of 19 TABLEI Inspection History of17R Outage GL 88-01 Scope System: Recirculation llR -
12R 13R 14R 15R 16R' 17R.
Insp.
Insp.
Insp.-
Insp.
Insp.
Insp.
Cat NG-D-011 RI 11R Yes Yes Yes No Yes No Yes E
NG-D-018 llR Yes Yes Yes Yes Yes Yes Yes F
NG-E-013 llR Yes Yes No No No No Yes C
System: Shutdown Cooling i
> llR 12 R--
13R 14R 15R 16R 17R.
17R-Weld Number
. SI Insp.
Insp.
Insp.
Insp.
-Insp.
Insp.
Insp.
Cat NU-4-001 15R No Yes Yes No Yes No Yes E
NU-4-002 No No-No Yes No Yes No Yes D
NU-4-003 15R No No Yes No Yes No Yes C
l l
i
r '---
~
1940-98-20418 s
Page 20 of 19 TABLE 2 Number of Welds by System and Category for Both Scopes l
\\
System GL 88-01 Proposed Reactor Head Cooling Category D 7
7 RWCU Category D 38 3
l Category G 9
0 Core Spray Category C 6
0 j
Category D 2
2 Isolation Condenser i'
Category C1 26 0
Shutdown Cooling Category C1 2
0 Category D 1
1 l
Recirculation j
Category C 1
0 Category E 2
2 Total 94 15 Note: Category Cl indicates stress improved with one post-stress improvement inspection.
I l
l
- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _