ML13050A785
| ML13050A785 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 02/07/2013 |
| From: | NRC/OCM |
| To: | |
| SECY RAS | |
| References | |
| RAS 24138, 50-361-CAL, 50-362-CAL, ASLBP 13-924-01-CAL-BD01 | |
| Download: ML13050A785 (96) | |
Text
UNITED STATES OF AMERICA U.S. NUCLEAR REGULATORY COMMISSION BRIEFING ON STEAM GENERATOR TUBE DEGRADATION FEBRUARY 7, 2013 1:00 P.M.
TRANSCRIPT OF PROCEEDINGS Public Meeting Before the U.S. Nuclear Regulatory Commission:
Allison M. Macfarlane, Chairman Kristine L. Svinicki, Commissioner George Apostolakis, Commissioner William D. Magwood, IV, Commissioner William C. Ostendorff, Commissioner
2 APPEARANCES NRC Staff:
Bill Borchardt Director of Operations Eric Leeds Director, NRR Ken Karwoski Senior Level Advisor, NRR Steam Generators and Material Inspection Chris Jackson Branch Chief, NRR Reactor Systems Branch External Panel:
Jim Benson Program Manager, Steam Generator Management Program, Electric Power Research Institute Hitoshi Kaguchi Project Director, Nuclear Plant Production Division, Mitsubishi Heavy Industries, Ltd.
Jeff Fleck Manager, NSSS Damian Testa Product Manager, Steam Generator Management and Modification Programs, Westinghouse Electric Company Pete Dietrich Senior Vice President and Chief Nuclear Officer, Southern California Edison Michel Pettigrew Adjunct Professor, Ecole Polytechnique Daniel Hirsch, President, Committee to Bridge the Gap
3 P R O C E E D I N G S 1
CHAIRMAN MACFARLANE: Good afternoon. The Commission 2
meets today to discuss information about steam generators, including recent 3
operating experience and the NRC's regulatory oversight. While ongoing steam 4
generator tube degradation issues at the San Onofre Nuclear Generating Station 5
have made recent news, the NRC has been evaluating and addressing steam 6
generator tube issues for decades.
7 First we will be hearing from the NRC staff about some of that 8
history, the agency's oversight of licensee steam generator programs, recently 9
observed tube degradation mechanisms, operating experience, and details on 10 the design basis accident for steam generators. Following the staff presentation, 11 we'll hear from seven external panelists representing the Electric Power 12 Research Institute, Mitsubishi Heavy Industries, AREVA, Westinghouse, 13 Southern California Edison, Atomic Energy of Canada, and the Committee to 14 Bridge the Gap.
15 Let me say at the outset that it is important to stress that the 16 following issues should not be discussed with respect to the ongoing adjudicatory 17 proceeding associated with the March 27, 2012 Confirmatory Action Letter 18 issued by the NRC to Southern California Edison Company for the San Onofre 19 Nuclear Generating Station, two issues in particular: one, whether the 20 Confirmatory Action Letter constitutes a de facto license amendment that would 21
4 be subject to a hearing opportunity under the Atomic Energy Act, and if so; two, 1
whether Friends of the Earth's pending intervention petition satisfies the 2
procedural requirements for a petition to intervene.
3 Before we get started, let me ask if any of my fellow 4
Commissioners would like to make any comments. No? Okay. In that case, we 5
will turn it over to the NRC staff, in particular to the Executive Director of 6
Operations, Bill Borchardt.
7 BILL BORCHARDT: Thank you, Chairman. Actually, Eric's going 8
to begin the presentation today.
9 CHAIRMAN MACFARLANE: Okay, great.
10 ERIC LEEDS: Thank you, Bill. Thank you, Chairman.
11 Commissioners, good afternoon. Today the staff will be giving you a 12 presentation on steam generator tube degradation and its significance. Certainly 13 steam generator tubes are important to safety because they serve both as a 14 reactor coolant pressure boundary as well as a containment boundary. If not 15 managed effectively, steam generator tube degradation can be risk-significant.
16 As the Chairman mentioned, the NRC staff expends significant resources 17 monitoring steam generator performance across the industry, and we focus our 18 oversight on the safe operation of steam generators.
19 Personnel in the Office of Nuclear Reactor Regulation are 20 responsible for reviewing licensee proposals related to steam generator tube 21 integrity and for reviewing licensee inspection results. In the regions there are 22 inspectors that inspect licensee steam generator programs at every nuclear 23 power plant. And in addition, the Office of Nuclear Reactor Research has an 24 active steam generator research program which looks primarily at steam 25
5 generator inspection and integrity issues.
1 We have two staff presenters today. Chris Jackson, the branch 2
chief of the Reactor Systems Branch in NRR, will be discussing safety analysis 3
related to steam generators. Chris will be followed by Ken Karwoski, the senior-4 level advisor for steam generators. Ken will provide an overview of steam 5
generator degradation, including some recent issues, and the staff's oversight of 6
licensee steam generator programs. So at that, let me turn this over to Chris.
7 CHRIS JACKSON: Good afternoon. My name's Chris Jackson.
8 I'm chief of the Reactor Systems Branch. My branch is one of several branches 9
in NRR responsible for accident analysis. Next slide, please.
10 Here's a cartoon of a reactor, but just to make a point, we focus in 11 the accident analysis in the fission product barriers: the fuel, the core in the 12 middle, the reactor coolant system pressure boundary, and the containment. If 13 you look off to the left is the turbine and the money-making side of the house, 14 and those are inputs to the accident analysis. And you'll see that the steam 15 generator is a kind of in between. They fit in the middle there. Next slide, 16 please.
17 When we talk about safety analysis, we talk about the chapter 15 18 safety analysis. These are in the final safety analysis report. They cover 19 anticipated operational occurrences as well as design basis accidents. These 20 range from turbine trips and reactor coolant pump trips all the way to double-21 ended guillotine of a major pipe in the system or steam line break. The 22 objectives for the design basis accident are protecting the fuel design limits. The 23 RCS pressure boundary cools the containment. And then, of course, onsite and 24 offsite doses. Dose consequences remain within the limits. Next slide, please.
25
6 Steam generators transfer heat from the reactor to the turbine.
1 Steam generators also form a barrier between the reactor coolant system and 2
the steam system. This creates the possibility of a containment bypass situation, 3
so they're of significant importance to us. Next slide, please.
4 Steam generators provide an input to many of the accident 5
analysis. For example, RCS flow, heat removal, steam generator pressure, core 6
inlet and exit temperatures, power -- all are directly related to the steam 7
generators. And obviously, the failure of the steam generator tube is a design 8
basis accident as well. Next slide, please.
9 So steam generator tube rupture with regard to design basis 10 accidents. This is described in Standard Review Plan, Chapter 15.6.3, and it's 11 evaluated from a dose perspective. And this is because the steam generators 12 can form a barrier or a bypass scenario around the containment. We evaluated -
13
- if you follow the instructions on 15.6.3 -- from a conservative standpoint. The 14 flow characteristics are conservative. You assume the maximum RCS activity, 15 which is quite conservative, and you'll also assume an iodine peak. And it's 16 evaluated from a dose perspective. From a fuel design standpoint, it's not 17 limiting. It's much less significant than the major rupture of a reactor coolant 18 pipe, and from a containment, the containment's not challenged at all. So next 19 slide, please.
20 One thing that makes steam generator tube rupture a little bit 21 unique is the reliance on operators. Traditionally, we rely on automatic actions, 22 but tube rupture requires operators to play an important role in plant recovery.
23 First, they have to diagnose the event and they have to identify the steam 24 generator. And they have to cool the plant down to reduce primary coolant 25
7 pressure below secondary pressure. The operators are obviously trained and 1
tested on this on a routine basis. Next slide, please.
2 We have had steam generator tube rupture events in the United 3
States. We have learned from that operating experience. Principally in 1982, 4
there was a tube rupture event at Ginna. In this event, the steam generators 5
filled with water. Water entered the steam lines, and water was discharged to the 6
steam relief valves. Steam lines aren't designed for water solid conditions, and 7
the steam relief valves aren't designed to relieve water. So the operators were 8
challenged on this event. So we took action -- so 1982. Next slide, please.
9 We issued three generic letters: generic letter 8207, 8208, and 10 8211. And the licensees took action through an owners group initiative to 11 improve steam generator tube rupture recovery capabilities. This involved 12 instrumentation for the operators, help diagnose the event, as well as the 13 equipment needed to dump valves with and without power so that they could 14 mitigate the event. And this is documented in WCAP-10698A, which is on 15 ADAMS. Next slide, please.
16 So in conclusion, the steam generator tube rupture event is not a 17 limiting event from a fuel perspective or from a containment pressure 18 perspective. Additionally, the steam generator tube rupture event isn't analyzed 19 from a dose perspective, and the doses are well within regulatory limits. With 20 that, I'll turn it over to Ken Karwoski, who will talk about steam generator design, 21 steam generator degradation, as well as inspections and assessments.
22 KEN KARWOSKI: Good afternoon. My name's Ken Karwoski. I'm 23 the senior-level advisor for steam generators in the Office of Nuclear Reactor 24 Regulation. There are two major types of steam generators in use in the United 25
8 States. There's recirculating steam generators and once-through steam 1
generators. There are 62 units with recirculating steam generators. These 2
steam generators have the traditional U-shaped tubes. These types of steam 3
generators have been used in plants designed by Westinghouse and 4
Combustion Engineering. There are seven units with once-through steam 5
generators. These steam generators have straight tubes instead of the U-6 shaped tubes, and these steam generators are used in plants that were originally 7
designed by Babcock and Wilcox.
8 We usually group steam generators based on the tube material and 9
the heat treatment that that tube material has received, and that's because that 10 determines the corrosion resistance of the tubing to degradation. There are 11 three types of tube materials and heat treatments used in the United States. The 12 first generation steam generators have mill annealed alloy 600 tube material.
13 The second generation steam generators had thermally treated alloy 600 tube 14 material. And the current material of choice for the replacement steam generator 15 is thermally treated alloy 690. Next slide, please.
16 This slide has several illustrations of steam generators showing 17 both recirculating and once-through steam generators. The steam generator on 18 the left is a typical recirculating steam generator that would be used in a 19 Westinghouse-type plant. It has the traditional U-shaped tubes. The 20 recirculating steam generator in the middle is more representative of a steam 21 generator that would have been originally in service in a Combustion 22 Engineering-type plant. And if you look closely, in that steam generator, the 23 tubes are shaped more in a square rather than the traditional U-shape. They 24 have two 90-degree bends, especially in the larger radius tubes. The figure on 25
9 the right is an example of a once-through steam generator, just illustrating that it 1
has straight tubes. Next slide, please.
2 Steam generator tubes have degraded with time. We usually bin 3
steam generator tube degradation into two broad categories: corrosion category 4
and a mechanical degradation category. In terms of corrosion, that's usually 5
cracking or pitting or some type of -- form of corrosion. The mechanical type of 6
degradation is typically wear. In plants with the original mill annealed alloy 600 7
tube material -- the first generation steam generators -- they've observed both 8
cracking and mechanical type degradation along the entire length of the tube.
9 And the examples that are illustrated on this figure have basically been observed 10 in -- all these examples have occurred in mill annealed alloy 600 tube material.
11 In the second generation steam generators, thermally treated 600, they've 12 observed some cracking, but the dominant degradation mechanism in those 13 steam generators is still wear. In the thermally treated alloy 690 steam 14 generators, the dominant degradation mechanism is wear. There has been no 15 corrosion-related degradation in thermally treated alloy 690 steam generators, 16 either domestically or abroad. Next slide, please.
17 Steam generator tube degradation has led the industry to develop 18 various repair criteria and repair methods in order to extend the lifetime of those 19 steam generators. These review criteria and repair methods -- or repair criteria 20 and repair methods are submitted to the NRC staff for review and approval 21 because these requirements are contained in the plant technical specifications.
22 There are operating conditions and maintenance tasks that can affect the lifetime 23 of the steam generators. For example, plant can elect to operate at a higher hot 24 leg temperature, which may make the tubing material more susceptible to 25
10 corrosion-related degradation. From a maintenance standpoint, the utility can 1
elect to clean the deposits or impurities out of the secondary side of the steam 2
generator less frequently, which would make the tubing material more 3
susceptible to degradation. When the staff reviews implementation of a steam 4
generator program, it doesn't necessarily focus on the operating items or 5
maintenance items that may affect the lifetime of a steam generator. It focuses 6
on tube integrity and making sure that the tubes can perform their intended 7
safety function for the period of time between inspections. Next slide, please.
8 Steam generator tube degradation has led to replacement of steam 9
generators at a number of units. Fifty-seven of the 69 currently operating 10 pressurized water reactor units have replaced their steam generators. All of 11 these replacement steam generators have incorporated design enhancements.
12 They took the lessons learned from the original steam generators, and they 13 made modifications to the steam generators to limit the potential for degradation 14 in their replacement steam generators. Some of the design enhancements 15 include changing the tube material, changing the tube support plate material, 16 changing the design of the support plates in order to limit the potential for 17 corrosive impurities to accumulate by the tubing.
18 Since 1989, all of the steam -- licensees have evaluated their 19 replacements under the 5059 process to determine whether or not a license 20 amendment is required. In all cases, a license amendment has not been 21 required for the entire replacement project, although specific amendments may 22 have been required for specific aspects of the replacement. Although the design 23 of the steam generator may not have been reviewed by the staff, it is subject to 24 regional inspections. There's an inspection procedure in place that the regions 25
11 follow that addresses all aspects of the replacement project, including the 1
planning and design phase, the installation phase, and the post-installation 2
inspection and testing phase. Next slide, please.
3 This figure depicts the number of pressurized water reactor units in 4
operation as a function of year. It also shows the tube material in those steam 5
generators at those units. As you can see in the red, in the '70s, '80s, and '90s, 6
the majority of plants had mill annealed alloy 600 tube material, the tube material 7
that has been susceptible to degradation the most. Starting in 1980, plants 8
began to install steam generators with thermally treated alloy 600 tubing, and 9
then starting in 1989, plants began installing steam generators with thermally 10 treated alloy 690 material. Currently there are 48 units with thermally treated 11 alloy 690 steam generators in service, 17 units with thermally treated alloy 600, 12 and four units with mill annealed alloy 600. Next slide, please.
13 I now want to talk about two current steam generator issues. The 14 first issue is cracking in thermally treated alloy 600 tube material, and the second 15 issue is tube wear. As I just mentioned, there's 17 units with thermally treated 16 alloy 600 tube material in the United States. The average age of these steam 17 generators is approximately 25 years. In 2002, the first instance of cracking was 18 detected in this tube material. Since 2002, additional cracks have been detected 19 at various locations along the tube length. The number and severity of the 20 cracking, however, has been minor. However, we still have concerns with this 21 because cracking is a time-dependent phenomenon, and it tends to accelerate 22 with time. As a result, there's a possibility that significantly more tubes will be 23 affected by this degradation mechanism.
24 In addition, cracking is a more difficult degradation mechanism to 25
12 manage than a lot of the other degradation mechanisms such as wear. Cracking 1
is harder to detect, it's more difficult to size, and prediction of the growth rates of 2
the cracks is more difficult than other mechanisms. Next slide, please.
3 The second issue I wanted to discuss is steam generator tube 4
wear. Tube wear has been detected both in original and replacement steam 5
generators. It has been found along various lengths of the tube. It's been found 6
in the free span region of the tube as a result of the tube interacting with a loose 7
part or foreign object. It's been found in the free span region of the tube as the 8
result of tubes interacting with other tubes. It's also been found at the tube 9
supports, including the support plates and the U-bend supports. That's the 10 location where most of the tube wear has been found -- at the support locations.
11 The number of indications of tube wear detected varies from unit to 12 unit. Some units have virtually no wear; other units have thousands of 13 indications of wear. However, it's not the number of indications of wear that is 14 important. It's the severity of the wear: how much tube material has been 15 removed and whether or not that tube can perform its intended safety function.
16 Unlike cracking, wear is more easily managed. Wear is easily 17 detectable in most cases. It's readily sized and it's easy to predict the growth 18 rates of wear. Next slide, please.
19 Now I wanted to talk about two specific occurrences of tube wear.
20 The first is tube-to-tube wear and replacement once-through steam generators.
21 In the fall of 2011, a plant was doing its first in-service inspection, and they 22 identified some indications. They eventually attributed those indications to wear 23 as a result of tube-to-tube contact. Upon finding this information, they shared it 24 with the other utilities who have once-through steam generators, and it was 25
13 subsequently determined that four of those units also had -- or a total of four 1
units had wear as a result of tube-to-tube contact.
2 This wear is shallow and appears to be slow growing, and a root 3
cause evaluation is currently underway. It's important to note that this wear has 4
been detected in steam generators designed and fabricated by two different 5
vendors. Next slide, please.
6 This next slide is just a graphic showing that the wear is located in 7
the free span region of the tubing. The wear -- at least one plant -- ranged up to 8
at least 8 inches in length. All the tubes with this type of wear have had 9
adequate tube integrity. Next slide, please.
10 I now want to talk about the tube wear that has been observed at 11 the San Onofre Nuclear Generating Station. In 2010 and -- or San Onofre 12 Nuclear Generating Station Units 2 and 3 replaced their steam generators in 13 2010 and 2011, respectively. These steam generators were designed and 14 fabricated by Mitsubishi in Japan. In early 2012, Unit 2 shut down for a normal 15 refueling outage and performed a steam generator tube inspection. They found 16 wear at a number of different locations. They found wear that was attributed to a 17 foreign object in the steam generator. The number of tubes affected was very 18 small -- just a few indications -- and the severity of those indications was also 19 small.
20 Wear was also found at the tube supports. A large number of 21 indications were found at that location. However, the severity of those 22 indications was very limited. Wear was also detected at a structure referred to as 23 a "retainer bar." There were a few indications of wear at that location. That wear 24 was unexpected, and one of the indications was deep. Although the indication 25
14 was deep, that tube had adequate tube integrity. There were also two tubes that 1
were identified that had wear as a result of tube-to-tube contact. Those 2
indications were shallow, and those tubes had adequate tube integrity. Unit 2 3
had operated a full cycle at 100 percent power, and they had adequate tube 4
integrity at the time of that inspection. Next slide.
5 Unit 3, in January of 2012, had operated for approximately half a 6
cycle when they observed primary to secondary tube leakage. The leak rate was 7
less than their technical specification limit, but the utility elected to shut down the 8
facility because the leak rate was increasing and the leakage was unexpected.
9 Like Unit 2, Unit 3 performed an inspection. They also identified wear at a 10 number of locations. They identified wear at tube supports, at the retainer bar, 11 and they also identified wear due to tube-to-tube contact. Unlike Unit 2, 12 however, some of the wear indications were significant; in particular, the wear 13 indications in the tubes affected by tube-to-tube contact. In all, there were eight 14 tubes that did not have adequate tube integrity. All eight of those tubes failed the 15 performance criterion in the technical specifications due to tube-to-tube contact 16 wear. The tube-to-tube contact wear has been attributed to the aggressive 17 thermal hydraulic conditions on the secondary side of the steam generator, 18 coupled with a lack of effective support in that region. Next slide, please.
19 I now wanted to spend a few minutes talking about our regulatory 20 framework, our oversight of licensee steam generator programs, and steam 21 generator performance. The inspection and repair of steam generator tubes are 22 covered in the plant -- are addressed in the plant's technical specifications. The 23 original technical specifications were developed in the 1970s, when wastage and 24 wall thinning were the dominant degradation mechanisms. When cracking 25
15 started to occur, it became evident that we needed to enhance our regulatory 1
framework. So the staff embarked on a multiyear effort to improve that 2
framework. Ultimately, the staff approved new generic technical specification 3
requirements, which were risk-informed and performance-based. The technical 4
specifications are performance-based in that we specify the criteria that the tubes 5
must meet at the time of the inspections. However, we do not specify the exact 6
details of how to go about achieving that objective. Since 2004-2005, all plants 7
have adopted these new risk-informed performance-based technical 8
specifications. Next slide, please.
9 The NRC oversees and monitors the implementation of licensees 10 steam generator programs. We have a multi-tiered approach, which involves 11 both regional and headquarter activities. From a regional perspective, they have 12 onsite resident inspectors. And from a steam generator tube integrity standpoint, 13 those inspectors would monitor reactor coolant system leakage and primary to 14 secondary leakage, which are indicators of possible loss of tube integrity. The 15 regions also have region-based inspectors who, from a steam generator tube 16 integrity standpoint, would go and monitor the licensees and service inspection 17 program for the steam generator tubes. In addition, as I indicated previously, 18 they would also inspect the steam generator replacement project.
19 There are also a number of activities here at headquarters with 20 respect to steam generators. The headquarters staff will have calls -- or 21 discussions with select licensees during their outage to monitor the scope and 22 results of the tube inspections being performed at those facilities. In addition, the 23 plant technical specifications require licensees to submit reports summarizing the 24 results of their inspections. All those reports are reviewed by the staff. In 25
16 addition to these activities, the staff will, as on an as-needed basis, meet with 1
licensees to discuss their inspection results. In addition, there are semiannual 2
meetings with the industry to discuss issues that may be generic to the steam 3
generator community. These interactions are available to the public. The 4
meetings are available to the public, our review of licensees steam generator 5
reports are documented, and those reviews are publicly available; as is when the 6
staff has discussions with the utilities, we document those discussions and those 7
summaries are made publicly available. Next slide, please.
8 The NRC has had a research program since the 1970s, when it 9
became evident that the steam generator tubes were susceptible to extensive 10 degradation. We look at three broad areas in that research program. We look at 11 the capabilities of the in-service inspection program to detect degradation. We 12 look at the assessments of tube integrity and the models that are used to 13 evaluate the flaws that are identified in the tubes. And in the past, we've looked 14 at the various corrosion mechanisms in place in the steam generator. The Office 15 of Nuclear Regulatory Research also has an international steam generator tube 16 integrity program, where a number of foreign entities are a part of that. That 17 program has been invaluable in ensuring that we keep abreast of research that's 18 going on overseas and also the operating experience in other countries. Next 19 slide.
20 In addition to the staff responding to the emerging degradation in 21 the 1970s and 1980s, the industry also took a number of steps. They formed a 22 group referred to as the Steam Generator Owner Group, which addressed steam 23 generator issues. Today the industry has a standardized program for addressing 24 steam generator issues. They developed guidelines which basically cover all 25
17 aspects of a steam generator program. They have primary and secondary water 1
chemistry guidelines. They have inspection guidelines that govern the inspection 2
of the steam generator tubes. They also have guidelines on how to perform 3
integrity assessments for the flaws that are observed in the tubing. In addition, 4
they have primary to secondary leakage guidelines, which provide guidance on 5
how to monitor for leakage and how to respond to that leakage. Next slide, 6
please.
7 Steam generator performance has improved since the 1970s. In 8
the next few slides, I'll show you a couple graphs that depict that improvement in 9
performance. Prior to 2012, the last time a plant did not have adequate tube 10 integrity was 2003. So in general, the performance has been very good, 11 especially over the last decade. Over the last 25 years, most of the losses of 12 tube integrity have been a result of cracking type degradation rather than 13 mechanical types of degradation. Next slide, please.
14 This slide depicts the forced outage frequency as a result of 15 primary to secondary tube leakage as a function of year. And as you can see, 16 there's a general declining trend with time. In the last few years, there have only 17 been a few primary to secondary leaks that have resulted in a forced shutdown.
18 Next slide, please.
19 This next slide depicts the steam generator tube rupture frequency 20 as a function of year. Once again, you see a declining trend as a result of 21 improved performance. One question that we always get is, is the improved 22 performance a result of the steam generator replacement, or does it actually 23 reflect improvement in performance of the inspections that are being performed?
24 And so on this graph, we have two lines. The red line, or solid line, shows the 25
18 tube rupture frequency for the first generation steam generators, with the mill 1
annealed alloy 600. And as you can see, that line has a declining trend. And 2
that's attributed to improvements in the steam generator programs that licensees 3
implement. Next slide, please.
4 We have a defense-in-depth approach to address steam generator 5
issues. And it starts with the design phase. We try to achieve a high-quality 6
design where the steam generator tubes are resistant to degradation. However, 7
we recognize that steam generator tubes may degrade. As a result, as Chris 8
Jackson indicated, a steam generator tube rupture during normal operation is a 9
design basis accident. In addition, in the unlikely event of a tube failure, 10 licensees train their operators to respond to primary to secondary leakage and 11 also to steam generator tube ruptures during normal operation. Even though we 12 try to achieve a high-quality design and have degradation-resistant materials, we 13 still require tube inspections to be performed on a periodic basis, and we require 14 assessments of those findings.
15 Tube inspections are performed when the unit is shut down.
16 However, the tubes normally degrade when the plant is operating. As a result, 17 we have various operational parameters that monitor to give us an indication of a 18 possible loss of tube integrity. These include reactor coolant system leakage and 19 primary to secondary leakage. In addition, we've assessed the risk significance 20 associated with steam generator tube degradation. We've assessed the risk 21 significance associated with single and multiple steam generator tube ruptures 22 during normal operation, design basis accidents. And we've assessed the risk 23 associated with tube degradation and severe accidents. As a result of these 24 reviews, we have not identified the need for regulatory action. Next slide, please.
25
19 In summary, steam generator tubes may degrade. However, this 1
degradation can be managed. The NRC staff monitors steam generator 2
operating experience, and our focus is on tube integrity. In general, the steam 3
generator performance has improved since the 1970s. That concludes my 4
presentation.
5 ERIC LEEDS: All right. That concludes the staff's presentation, 6
and turn it over to the Chairman.
7 CHAIRMAN MACFARLANE: Okay, great. Thank you. Thank you 8
very much. It's very clear presentations. Appreciated that. We will start with 9
questions with Commissioner Apostolakis.
10 COMMISSIONER APOSTOLAKIS: Thank you, Madam Chairman.
11 And thank you for your presentations. My interest in today's topic is very high.
12 And I very much value the information that we are getting from you and will get 13 from the second panel. However, as the Chairman mentioned, there's an 14 ongoing adjudicatory proceeding, so that limits the number and nature of 15 questions I can ask. And I can assure you, however, that perhaps the 16 shallowness of my questions does not reflect my interest in this topic.
17 Ken, you mentioned that there's degradation or the possibility of 18 degradation of tubes when there are aggressive thermal hydraulic conditions on 19 the secondary side. And I guess the result of that may be another technical term 20 that I see in all the documents: fluid elastic instability. Now, not all of us have 21 spent a career studying steam generators, so can you explain in simple terms 22 what that means?
23 KEN KARWOSKI: The term "fluid elastic instability"?
24 COMMISSIONER APOSTOLAKIS: Yes.
25
20 KEN KARWOSKI: That's a condition when the velocity of the water 1
on the secondary side of the steam generator causes the tubes to become 2
unstable and vibrate excessively in different modes of operation. And so in the 3
case of that specific condition, the tubes will start to vibrate excessively. There's 4
two forms -- that vibration can be both in the plane of the U bend or it can be out 5
of the plane of the U bend. In the case of San Onofre, for example, it was in 6
plane fluid elastic instability that resulted in a lot of the tube-to-tube contact.
7 COMMISSIONER APOSTOLAKIS: And that instability, as I 8
understand it, is related to the natural frequencies of the tubes, right?
9 KEN KARWOSKI: Yes. The support conditions of the tubing would 10 affect the susceptibility of the tubing to fluid elastic instability. The more well 11 supported a tube is, the less susceptible it will be to fluid elastic instability.
12 COMMISSIONER APOSTOLAKIS: Thank you. On Slides 26 and 13 33, you talked about the risk significance of this tube rupture event. And you said 14 that there was an assessment regarding risk significance, but you didnt tell us 15 what the results were. What is the risk significance?
16 KEN KARWOSKI: Steam generator tube degradation, if not 17 effectively managed, could be risk significant. However, we performed various 18 risk assessments that have looked at both tube ruptures during normal operation 19 and design basis accidents and have concluded that the risk associated with that 20 is within acceptable limits.
21 COMMISSIONER APOSTOLAKIS: Okay. So its the totality of 22 everything we do and so on that results in a lower risk.
23 KEN KARWOSKI: Right.
24 COMMISSIONER APOSTOLAKIS: Thank you Madam Chairman.
25
21 No, you want to add?
1 CHRIS JACKSON: Yeah. I mean, in Kens presentation he was 2
discussing globally, we had generic issue 163 and generic issue 188, which 3
looked at risk, you know, from a broad perspective -- accident induced tube 4
ruptures, multiple tube ruptures, severe accident -- and based on all the work 5
above and beyond the significant work we were doing with the performance 6
degradation, performance-based approach, we chose not to do any additional 7
work. With regard to San Onofre and the risk significance of the event itself, the 8
operators -- leakage was detected, primary to secondary leakage was detected, 9
and the operators chose to shut the plant down. They chose to shut the plant 10 down well below the tech spec limit. You know, they were at a procedural limit 11 and they chose to shut the plant down. Thats exactly what we would look for in 12 the performance degradation. You identify the leakage before its a rupture. So 13 they were well below tech spec leakages of 150 gallons per day, so this is quite a 14 small amount of leakage. The plant was shut down safely. The operators took 15 appropriate action. So from that perspective, the public wasnt put in any risk 16 from the event when the leakage was discovered.
17 COMMISSIONER APOSTOLAKIS: Thank you. Thank you Madam 18 Chairman.
19 CHAIRMAN MACFARLANE: Commissioner Magwood.
20 COMMISSIONER MAGWOOD: Thank you Chairman, and thank 21 you for your presentations today. As all this matter has evolved over the last 22 year or so, and as Ive spoken -- havent had a chance to sit down with you. I 23 look forward to doing that, but as Ive talked to staff and Ive talked to industry 24 people and other experts about all this, a lot of the discussion seems to be, I 25
22 guess, I would use the -- Commissioner Apostolakis, we talked about this earlier 1
today, empirically based. Its all experiential. You look at how things have 2
behaved in the past and thats how you design programs going forward. And to 3
some degree, a lot of our discussion has been empirical. Youve talked in terms 4
of experience rather than physics, say. And that raises a question with me and 5
Id like to hear your thoughts about this. It seems to me that as we design 6
sophisticated equipment these days, we rely a lot more on analysis than we do 7
simply on, you know, whats worked in the past. So I expect that vendors are 8
using very highly sophisticated codes and models to design their systems. Are 9
we using highly sophisticated models and codes to analyze those systems? And 10 Id like to understand where are we on the regulatory side of this in terms of 11 understanding -- we see a steam generator design. What is our ability to analyze 12 that design from a physics standpoint, from a modeling standpoint as opposed to 13 simply referencing whats worked in the past?
14 CHRIS JACKSON: Well, you make a good point, and as we get 15 more, you know, weve redesigned many of the generators, mainly with like-to-16 like replacements, but as we do, we use more modern tools. And by and large 17 as industry does that, that improves the plant design. So when we do our 18 reviews -- we call it the standard review plan -- so the standard review plan 19 outlines the entire scope of our review and we use those for new reactors, we 20 use them for operating reactors. As we learn, we update those. So if you look at 21 the design, typically we focus on the safety significance, on the safety analysis, 22 and on the consequences. So we rely on industry codes and industry standards 23 for the design. The diesels and the valves are designed to as ASME standard or 24 a concretes designed to an ACI standard. So the design, rarely do we get into 25
23 the heavy details of the design itself. We look at the consequences and the 1
integrated safety. So, you know, obviously for fuel and some other components, 2
we look heavily at the design. So, for the steam generators, the thermal 3
hydraulic conditions on the secondary side are not something thats within our 4
standard review plan. We dont look at that in excruciating detail when we do the 5
review. So, thats something thats more financial based. If they make that 6
mistake it costs them money because the generators dont perform from a 7
financial standpoint. But from a safety standpoint with our defense-in-depth 8
approach, thats less important. Obviously if we learn the operating experience 9
when we go forward we may revisit that. But at the moment, the standard review 10 plan doesnt have us looking in detail at thermal hydraulic conditions in the 11 secondary side for the design basis.
12 COMMISSIONER MAGWOOD: I wanted to let Jennifer jump in, 13 but just let me make sure I understand what you just said because it sounds a bit 14 like -- and just give me a chance to clarify this a little bit for me -- because it 15 sounds a bit like -- because as I think it is, as both you and Ken indicated, the 16 tubes are -- they are a containment barrier, basically. So one would expect that 17 there would be some analysis that would go with this to assure that that barrier 18 will remain intact during the operating life of the equipment. So is that not how 19 we look at that? Can you clarify that a bit for me? Because it sounds like you 20 were saying we dont analyze that and --
21 JENNIFER UHLE: I think I can maybe help answer that question.
22 Im Jennifer Uhle. Im currently the deputy director for Nuclear Reactor 23 Regulation, but until about a month ago I was the deputy director of the Office of 24 Nuclear Regulatory Research where, as Ken has indicated, weve had an 25
24 extensive steam generator tube integrity research program for many years since 1
about the 70s. And I would say to help answer specifically your question, 2
Commissioner Magwood, areas that we have advanced our modeling has been 3
in looking at the -- the modeling that takes place once a crack or a degradation 4
has occurred, then there are analytical approaches that are used to determine 5
what the integrity of the tube would be if it would be challenged under various 6
operating conditions. And our approach so far has been to take a look at what 7
the industry has been providing. We model cracks. Cracks are hard to size and 8
to characterize specifically, so we have, you know, an approximation method to 9
model the geometry of the crack and then, you know, with an analytical approach 10 determine what pressures and temperatures and how much more cracking could 11 occur before it would become a problem during that next operating cycle. So 12 weve made some advances there, but as Chris has indicated, we do not use the 13 same level of design tools to take a look at the integrity of the tubes themselves, 14 largely because of the design process involving ASME type codes.
15 COMMISSIONER MAGWOOD: And another observation I would 16 make is that we focus, and I think most of the discussions focus on the tubes 17 themselves, as opposed to the mechanics of the structure.
18 JENNIFER UHLE: Yes. Right.
19 COMMISSIONER MAGWOOD: But even, to some degree, thermal 20 dynamic behavior, because I think a little bit of what Chris was saying, the 21 thermal dynamic behavior is more of an economic issue as opposed to safety 22 issue. But it does have safety implications obviously because that leads to fluid 23 elastic instability and other --
24 JENNIFER UHLE: We have, in the research program, looked at 25
25 more physics-based corrosion approaches. I can point to, well, what we call the 1
model boiler program, that we found because we -- because each of the plants 2
have a different chemistry on their secondary side, we could not possibly come 3
up with correlations specific for each plant. So at this point in time, we use more 4
of the -- or as Ken has indicated, the integrity approach. And so if the crack 5
growth rate is a certain amount, as long as it can continue operating until the next 6
inspection cycle, then we find that adequate protection.
7 BILL BORCHARDT: Just as a non-expert in many of the technical 8
aspects, our concern has always been public exposure. I mean, we wanted to 9
make sure that we had a program in place so that if a tube did develop a 10 problem, it wouldnt result in release off-site that exposed the public 11 unacceptably. It wasnt to make sure that there would never be a tube problem, 12 right? So the fact that a tube develops a leak, theres a tech spec which governs 13 that, requires the plant to shut down, do in-service inspections and plug tubes 14 that look like they wont make it until through the next cycle, thats our interest.
15 COMMISSIONER MAGWOOD: Thats interesting, and I think that 16 certainly leads you to -- and I know the tech specs allow for a certain amount of 17 leakage, but it does lead to a lack of an analytical basis for any changes in 18 design. You know, we rely on the experience that weve had in the past to 19 design our inspection program, and without an analytical basis, if there are 20 changes in, you know, steam generator design, how do we know that those --
21 had experience that applies to the new designs? How do we sort through that?
22 BILL BORCHARDT: Just to clarify one aspect, I think, of what you 23 were saying, there is a technical basis, I believe, for understanding crack growths 24 and different failures, that once detected through ISI, we can have a reasonable 25
26 expectation of how long that tube will last and whether or not it needs to be 1
plugged. So there is a very good, I think, basis for that. I think youre right 2
though about the internal thermal hydraulics of the steam generator.
3 COMMISSIONER MAGWOOD: I mean, if someone were to have 4
a, I dont know, a helical steam generator say, just to pick something at random, 5
how would we analyze that? Whats -- how do with do that? Do we do that?
6 Here comes Jennifer again.
7 JENNIFER UHLE: Sorry. This is Jennifer Uhle again from Nuclear 8
Reactor Regulation. I would, in case of large deviations from the state of 9
practice, then we would obviously invest more time in analyzing the structural 10 integrity -- I mean, the overall behavior of the structure of the steam generator.
11 And I would expect that all of our methods that we currently use to look at tube 12 integrity, we would revisit under the research program. For example, these types 13 of simplified analytical approaches to determine the effect of a crack and its 14 impact on structural integrity, when we saw the changes to 690 -- excuse me, the 15 Alloy 690 material, weve gone back and just verified that, in fact, these 16 approaches are still valid.
17 COMMISSIONER MAGWOOD: Okay. Well thank all of you for 18 your -- Chris, did you want to --
19 CHRIS JACKSON: Well, yeah, just to follow up, I mean by and 20 large when we dont have a lot of information, you know, for new reactors, you 21 just go conservative -- or where we lack information. I think in this instance, the 22 thermal hydraulic conditions in the secondary side surprised the utility and 23 surprised us. And the consequences were what are surprising and that this was 24 a new degradation and were going to study it, so, you know.
25
27 COMMISSIONER MAGWOOD: Appreciate it. Thank you for your 1
answers. Chairman, for the meeting -- SRM, it might be worthwhile to ask 2
Research to give us some thoughts about what direction research in this area 3
should take going forward with all the experience that weve recently 4
accumulated. Thank you.
5 CHAIRMAN MACFARLANE: Thank you, Commissioner Magwood.
6 Commissioner Ostendorff.
7 COMMISSIONER OSTENDORFF: Thank you, Chairman. Thank 8
you all for the presentations. Im going to start out, Ken, with you. I remember 9
back in the Navy in the 1970s we had a significant problem with steam generator 10 tube cracking associated with stress corrosion cracking phenomena, chemistry 11 control sludge, chemical hideout, pH phosphate control issues that were of 12 significant concern, resulted in a lot of special inspections. I remember doing 13 one as engineer of a submarine back in 1980 in Charleston, and we plugged the 14 tubes routinely at that time. It was like Bill said, very straight stick approach.
15 You go in there, you do an inspection, see cracks, you plug the tubes.
16 Something that was said earlier today, and your presentation led me to ask this 17 question, these generators were basically used for 30 years. Thats was the 18 service life of a submarine. The ones that are in -- you know, the generators that 19 are in service in PWRs today, what kind of a time period are they designed for?
20 Is there a standard? Is it 40 years? Is it something more than that? How do we 21 look at that as far as standardization or expectation of life? Because I was 22 looking at all the ones that have been replaced as opposed to just ones that had 23 tubes plugged, and theres a big difference between those two evolutions.
24 KEN KARWOSKI: In terms of the steam generator design, most 25
28 steam generators were originally designed, I believe, for 40 years based on 1
general corrosion of the tubing. They did not anticipate cracking, and as a result, 2
a lot of those steam generators did not make 40 years; some replaced, I think, 3
the earliest was like after seven years of operation, and some have operated 4
over 30 years. So in the design phase usually what youre looking at is general 5
corrosion of the tubing. If cracked, then you try to select materials that are not 6
susceptible to mechanisms such as stress corrosion cracking.
7 COMMISSIONER OSTENDORFF: Okay. So as these -- lets say 8
for generators replaced in the last 10 years. Are they still being nominally 9
designed for a 40-year period based on general corrosion or for cracking or tube-10 to-tube wear? Is that thinking design methodology evolved with experience and 11 if so, how?
12 KEN KARWOSKI: Okay, our replacement steam generators were 13 thermally alloy 690. In general, I dont want to say that that material is immune, 14 but it is more resistant to corrosion and in particular stress corrosion cracking.
15 So its difficult to model the lifetime of a steam generator on whether or not it may 16 or may not be susceptible to cracking. So, a new steam generator would still 17 have, you know, a licensee can specify in their specification for a replacement 18 steam generator 30 years, 40 years, and I think some of the new reactors, the 19 design might be 60 years and that would be based on fatigue, analysis, and 20 general corrosion of the tubing material. So in some respects, its up to the utility 21 to determine what the lifetime is. But as I indicated in my presentation, our focus 22 is making sure that the tubes have adequate tube integrity for the period of time 23 between inspections. The lifetime is more of an economic issue.
24 BILL BORCHARDT: Yeah. You might want to ask the second 25
29 panel this. They certainly have better information. But I dont think you can look 1
at the number of steam generator replacements that have been made to date 2
and assume that those generators were in a condition of failure or had reached 3
the maximum number of tube plugs allowed. But rather, the decisions, were in a 4
number of cases made, along with the license renewal decision, and once the 5
authorization was granted or the decision was made to operate from 40 to 60 6
years over the life of the plant, then they had a 30-year-old generator and said, 7
Well, well just replace the generator now and thatll get us to the end of life.
8 And so that decision was made more -- I cant say more -- but influenced by the 9
business economic decision of having a new generator with enhanced 10 performance that would relieve some operator burden, that would reduce dose 11 exposure during outages because of the number of tubes that were required to 12 be plugged and the ISI programs. And so there were a lot of factors, I think, that 13 went into the steam generator replacements that have been conducted so far.
14 COMMISSIONER OSTENDORFF: Okay. Thank you. Thats a 15 very important distinction and I appreciate your identifying that. One comment 16 that Jennifer made that kind of caught my attention if I understood her correctly --
17 you want to come back to the podium. You dealt with chemistry, I guess -- or did 18 I understand you to say that the chemistry varies significantly even for the same 19 type of generator being installed at different plants? You might see different 20 approaches to chemistry control?
21 JENNIFER UHLE: I would say over time they have evolved, but Ill 22 ask Ken to say, but yes, they vary between plants.
23 KEN KARWOSKI: Yeah, and thats the key. The industry operates 24
-- I believe they all follow the EPRI primary and secondary water chemistry 25
30 guidelines and stay within the limits and action levels within those guidelines.
1 However, when youre talking about steam generator tube degradation, you have 2
an integrated effect. So a steam generator that was put into operation 35 years 3
ago has the integrated effect of every chemical transient that it had experienced 4
over the lifetime of the plant, and that is very difficult to model on, you know, how 5
much deposits are on the secondary side, you know, what type of impurities may 6
exist in there because of the materials that are used on the secondary side of the 7
plant, which can vary from plant to plant. So that is very difficult to model on a 8
generic basis and make a generic type of conclusion.
9 COMMISSIONER OSTENDORFF: Okay. Thank you. Let me shift 10 gears a little bit. Ken, you mentioned -- and Im going to direct this question to 11 you and to Chris -- you talked about the international exchange of information 12 and I believe that Chris you were over at ASN for a while --
13 CHRIS JACKSON: Yes, I was.
14 COMMISSIONER OSTENDORFF: -- and I was curious since the 15 French have a lot of PWRs, Im curious about what experience the French have 16 had with similar issues on tube degradation or cracking. Whoever wants to take 17 that.
18 CHRIS JACKSON: Well, Ill start and then Ill let Ken follow up. But 19 I think the French steam generators are very similar to ours. The original designs 20 were very similar to our designs and theyve evolved. So we use the same 21 constructors and vendors; so they are very similar and we share information.
22 KEN KARWOSKI: Yes. So in terms of, you know, tube cracking, 23 theyve observed tube cracking, theyve had similar issues. They do have some 24 modifications in the designs which may make their steam generators more or 25
31 less susceptible to certain forms of degradation, but for the most part theyre very 1
similar, as Chris indicated. And we share operating experience. There are some 2
differences, but we share operating experience and we factor that in.
3 COMMISSIONER OSTENDORFF: Okay. Let me shift gears and 4
stay with Ken and Chris, and Eric, feel free to add in here if you want to, or Bill, I 5
want to talk technology. I was out at SONGS in July of last year and I believe it 6
was AREVA that had -- was coming in to look at tube-to-tube gap, looking at 7
different dimensional checks of their tubes. I wanted -- I think, as I understood at 8
the time, that theres some very specific technology being used for non-9 destructive evaluation of those tube-to-tube gaps. Are there standard practices 10 from where we sit as regulators that would suggest theres a proper way of 11 looking at this as far as the gaps to measure for proper clearances?
12 KEN KARWOSKI: When a specific issue comes up at a plant, 13 frequently theres no generic techniques. There have been two proximity signals 14 in other steam generators, so utilities have developed various techniques to try to 15 ascertain, you know, are tubes in closer proximity and could there be a potential 16 issue there. But in terms of have we reviewed and approved a specific technique 17 for measuring those gaps, no, that would normally be up to the licensee. And to 18 the extent that thats relied upon in our decision-making process, then we would 19 ask for the appropriate technical basis and then review that if its subject to, you 20 know, our review and approval.
21 COMMISSIONER OSTENDORFF: I think you have some backup 22 here at the podium.
23 JENNIFER UHLE: Its Jennifer Uhle from NRR, and I would add to 24 that that our -- the research program is heavily involved in looking at the 25
32 effectiveness of the various NDE techniques that are being deployed for those 1
parameters that we find important in affecting the tube integrity; so its a very 2
active part of what were doing in research.
3 COMMISSIONER OSTENDORFF: But is it the staffs general 4
impression that theres reliability in these gap measurements that are being 5
reviewed across industry, not just one plant --
6 KEN KARWOSKI: Well, in general, we know that the current 7
technique is capable of seeing outside the tube wall, detecting loose parts that 8
are metallic that might be next to a tube wall. We do know that utilities do use 9
various techniques to monitor different things like the ovality of the tube or the 10 diametrical expansion or the profilometry. In this particular case, part of the AIT, 11 or our technical evaluation report, we may look at that specific aspect, depending 12 on how its used in the specific review.
13 COMMISSIONER OSTENDORFF: Okay.
14 JENNIFER UHLE: But I would just add to that that if there was a 15 parameter that a licensee was going to rely on to then feed into their analytical 16 approach to determine whether or not the tube was likely to have integrity over 17 that continued operating cycle, then we would be very interested in 18 understanding the reliability of that NDE technique. The parameter that you are 19 citing, I do not personally know if that parameter is currently used. So I think 20 thats -- youre getting some of the --
21 COMMISSIONER OSTENDORFF: I understand. Thank you.
22 Thank you, Chairman.
23 CHAIRMAN MACFARLANE: Okay. Thank you. All right. Eric or 24 Chris, can you back up a sec -- and I know Chris, in your slides you talked about 25
33 a limiting event, and I think thats a bit deep into the lingo there, so Id like to 1
clarify that. And Id like you guys to explain what -- just put on the record here, 2
whats the worst that can happen during a steam generator tube rupture 3
scenario?
4 CHRIS JACKSON: A limiting event. I apologize. That is lingo. It 5
comes so naturally. But the fuel is designed for various challenges, and a 6
reactor coolant system is designed for various challenges, and so is the 7
containment. And this event doesnt approach any of the limits. Theres other 8
events that are much more significant. So from a fuel standpoint, this is not a 9
challenging event for fuel. Many of the other design basis accidents do 10 challenge the fuel when you approach limits, and they dont exceed them, but 11 this is not a challenging event for the fuel. So other events will uncover the fuel, 12 will cause cladding to raise temperature, oxidize the fuel, which can be 13 acceptable, depending on the type of event. This one does not. Theres no 14 temperature excursion, so the fuel remains intact for the design basis event. The 15 same goes for the containment. The limits, other events, the steam line break, 16 the large break LOCA will challenge the containment, you can approach limits; 17 but this one doesnt, so its not a challenging event from that perspective. But 18 what makes it challenging is the operators have to take action and you have the 19 containment bypass scenario; so thats what makes this event more interesting.
20 CHAIRMAN MACFARLANE: So then explain whats the worst that 21 can happen during a tube rupture scenario.
22 CHRIS JACKSON: Well, I mean, the worst that can happen is you 23 would -- I guess theres several scenarios. If your tube ruptures your limiting 24 event, the worst that could happen is you wouldnt have safety systems, you 25
34 would drain the reactor coolant system and fail the fuel and have a direct release 1
path to the environment. That would be a very bad event. You could have 2
another event, a steam line break that would result in fuel failure as well. So the 3
worst event would be a fuel failure with a containment bypass.
4 ERIC LEEDS: Im not sure if thats what youre asking for, 5
Chairman. I think that what youre talking about is for steam generator --
6 CHAIRMAN MACFARLANE: Yes.
7 ERIC LEEDS: -- tube degradation --
8 CHAIRMAN MACFARLANE: Yes. Yes.
9 ERIC LEEDS: -- is it possible to have multiple tube ruptures? How 10 would the plant react to multiple tube ruptures?
11 CHAIRMAN MACFARLANE: Right. Right. Yeah.
12 ERIC LEEDS: And for that type of a scenario, very similar to a 13 single tube rupture, what youre doing is you rely on the operators to spot the 14 leakage -- and they have technical specification limits on leakage -- and shut 15 down the plant, and allow the pressure to equalize between secondary side, the 16 side that goes to the steam -- to the turbine, the money-making end, and the 17 primary side. So what youd have is youd have a bigger probability for a primary 18 to secondary leak, you get more of the primary fluid into the secondary fluid.
19 That could release more radionuclides into the secondary side, so the possibility 20 is there for more of a possibility for a release off site or within the turbine building.
21 CHRIS JACKSON: Yeah. I apologize. I dont think I answered the 22 question. So if you had multiple tube ruptures, that would be a more challenging 23 event from some perspectives but from other perspectives it wouldnt 24 necessarily. So one of the challenging things on a tube rupture, particularly a 25
35 small one, is diagnosing it. With multiple tube ruptures, youd have an earlier 1
plant transient and youd be able to identify the tube or the generator quicker, 2
safety systems would react; so multiple tube ruptures would challenge the 3
operators in a different way. But we have studied that from a risk perspective 4
and we chose not to take regulatory action or regulatory action wasnt necessary.
5 So in some aspects the operators were benefitted by automatic systems and 6
easier diagnosis, but the timing would create another challenge for them, so...
7 CHAIRMAN MACFARLANE: Let me ask Ken a question. So Im 8
interested in alloy 690 and 600. You said there are still some plants that use 9
alloy 600, the TT one. Whats the compositional difference between those?
10 KEN KARWOSKI: In general, its the chromium content of the 11 alloy. Alloy 600 has lower chromium. Alloy 690 has higher chromium. The mill 12 annealed versus thermal treatment refers to a heat treatment that those materials 13 receive that make them -- that may change the micro-structure in terms of -- the 14 carbide distribution around the grain boundaries, which may make it more or less 15 susceptible to stress and cracking.
16 CHAIRMAN MACFARLANE: I understand that. Limiting events, 17 no. but that, yes.
18
[laughter]
19 Good, okay. So what do you think then is the biggest challenge 20 right now facing the replacement steam generators? Is it tube wear, or not? Or 21 is there something else?
22 KEN KARWOSKI: I would say that the biggest challenge --
23 recognize the thermally-treated alloy 600 includes both original and replacement 24 steam generators. But Id say the challenge I see in the future is managing the 25
36 cracking and that thermally-treated alloy 600 tube material. As far as I know, 1
none of those plants have plans to replace their steam generators. Right now, 2
theres only been a limited amount of cracking, but we know from past 3
experience cracking tends to follow an exponential curve, so I -- from my 4
perspective, I believe that the cracking in the thermally-treated alloy 600 is 5
probably the biggest challenge.
6 CHAIRMAN MACFARLANE: Okay, so once -- so have any steam 7
generators been replaced that -- so there were replacement steam generators 8
but they have the alloy 600 in it?
9 KEN KARWOSKI: Yes. Recognize we usually distinguish two 10 different forms of alloy 600.
11 CHAIRMAN MACFARLANE: Yeah, I know. The thermally-treated 12 13 KEN KARWOSKI: The mill annealed alloy 600 and the thermally-14 treated alloy 600. There has been one plant, because of the timing and the 15 replacement, actually replaced their original mill annealed alloy 600 steam 16 generators with steam generators with mill annealed alloy 600 tubing. Theres 17 been one unit that has done that. There have been approximately eight units 18 that have replaced their original mill annealed alloy 600 tube material with 19 thermally-treated alloy 600 tube material.
20 CHAIRMAN MACFARLANE: Okay. So theres -- there are some 21 plants that have done that. All right.
22 KEN KARWOSKI: Yes.
23 CHAIRMAN MACFARLANE: Okay. So then let me try to 24 understand. You guys talked about tube ruptures at other plants, maybe one in 25
37 particular. I want to understand a little bit more about other experience of tube 1
ruptures.
2 KEN KARWOSKI: Theres been nine domestic tube ruptures, if my 3
memory serves me correctly. The last tube rupture occurred in the year 2000 at 4
Indian Point 2. Most of the tube ruptures occurred in the 70s and 80s, and in 5
1993 there was a tube rupture at Palo Verde Unit 2. So the last two tube 6
ruptures, one was in 2000, one was in 1993, and there were several in the 80s 7
and the 1970s.
8 CHAIRMAN MACFARLANE: Okay. And then what about 9
international experience with tube wear and tube cracking? I mean, is there --
10 some countries plagued more than others?
11 KEN KARWOSKI: I would say in general these steam generators 12 are very similar, and as Chris indicated, a lot of the designers and fabricators are 13 used throughout the world. So the experience is comparable from country to 14 country.
15 CHAIRMAN MACFARLANE: Yeah. Okay. Good. And Bill, you 16 mentioned something about the new generators have enhanced performance.
17 What does that mean?
18 BILL BORCHARDT: Well, I was just referring to the improved 19 materials that are inside.
20 CHAIRMAN MACFARLANE: Okay. So thats pretty much the only 21 thing thats been improved or changed in the new generators is the materials?
22 KEN KARWOSKI: The tube materials have been enhanced, but 23 other things can affect the degradation of tubes. So the tubes support materials 24 have changed. They used to be carbon steel, they used to corrode and create 25
38 issues. Theyve been replaced with stainless steel. Theyve changed -- when 1
you bend a tube you put a lot of stresses in the tube; that can make it more 2
susceptible to cracking. Theyve -- for the tubes that are the most tightly bent, 3
the smallest radius tubes, theyll do a thermal stress relief to help reduce those 4
stresses. Theyve changed the method by which they expand the tubes within 5
the tube sheet to reduce the stresses in the tube sheet at the expansion 6
transition region. So there have been a number of design changes that have 7
taken place in order to limit the potential for degradation.
8 CHRIS JACKSON: Water chemistry has changed for the better, 9
too, in many areas. So, you know, I know plants are working quite a bit to 10 remove things in their feed water system that could --
11 CHAIRMAN MACFARLANE: What are the problems in the water 12 chemistry that different --
13 KEN KARWOSKI: Different chemical species can cause 14 accelerated corrosion. Theres certain chemicals that you never want to get into 15 your steam generators, like lead. Lead can lead to stress corrosion cracking very 16 quickly. So there are certain chemical species that utilities definitely try to avoid 17 their use anywhere on the secondary side, and then they monitor those species 18 that can be detrimental to the tubing and keep the concentrations of those 19 species below limits that are normally specified in the EPRI water chemistry 20 guidelines.
21 CHAIRMAN MACFARLANE: Okay.
22 CHRIS JACKSON: And from an operating experience standpoint, 23 the industry and us learned quite a bit from the 1982 Ginna event, where they did 24 have challenges bringing the plant down into a safe condition; so thats all been 25
39 implemented into the regulatory process for the owners group initiative.
1 CHAIRMAN MACFARLANE: Okay. Okay. Great. Thank you.
2 Commissioner Svinicki.
3 COMMISSIONER SVINICKI: Im batting cleanup today, and I was 4
thinking about Commissioner Apostolakis began this round of questioning by 5
referring to the fact that although there are a lot of important dimensions of the 6
topic were talking about today, not all of them are appropriate for pursuit and 7
resolution in this venue, so that makes it difficult for us to -- even the technical 8
information is being used as the bases to resolve certain investigatory and 9
adjudicatory matters. So I think I will just compliment you all on your 10 presentations, and I will tease Eric and Bill a little bit by saying its easy to look 11 good when you have people like Chris and Ken and Jennifer, who took a very 12 wise position near the podium, just so I compliment -- its easy to look good when 13 you have people that have such an impressive command of all of the history and 14 the technical details. So thank you for the work you do and I will yield back.
15 Thank you.
16 CHAIRMAN MACFARLANE: Great. Thank you. Thank you all.
17 Thanks very much for excellent presentations. We will now take a five-minute 18 break before we convene the external panel.
19
[break]
20 CHAIRMAN MACFARLANE: All right. I think were had a bit of a 21 break, and now were back, and to get us started with the external panel, Im 22 going to turn to Jim Benson, the program manager for the Electric Power 23 Research Institute Steam Generator Management Program. Mr. Benson.
24 JIM BENSON: Well, good afternoon. Im going to start by 25
40 providing a review of the EPRI Steam Generator Management Program, and if 1
we could go to Slide Number 3. So, EPRI was established in 1973 as an 2
independent non-profit organization to conduct research on key issues facing the 3
Electric Power Research Institute. And in 1975 EPRI established a steam 4
generator program first beginning with the Steam Generators Owners Group, and 5
now were turning into the Steam Generator Management Program, which exists 6
today, one of the longest standing issue programs within EPRI. Next slide, 7
please.
8 The specific objectives of the EPRI SGMP include identifying and 9
prioritizing and conducting steam generator research to address materials 10 degradation. Additional objectives include performing a long-term R&D in 11 various technical areas such as water chemistry, non-destructive examination, 12 materials and thermal hydraulics; all of those that have a significant impact on 13 the steam generator operation. And the last objective Ill mention is the 14 development of necessary technologies, processes, procedures, and tools to 15 support the assessment of steam generator tube integrity. Next slide, please.
16 The SGMP develops many documents, technical documents, and 17 guideline documents. The guideline documents come about by the research we 18 do as well as industry experience, and we have currently six guideline 19 documents that are put out to the industry. These documents, although called 20 guidelines, they do have requirements in them that the utilities do put into their 21 steam generator programs. Those guideline documents address secondary 22 water chemistry, primary water chemistry, examination, the techniques that are 23 used, as well as the period between inspections or the inspection interval. They 24 address primary to secondary leakage, monitoring for leakage, and actions that 25
41 need to be taken when leakage is identified. There are also tube integrity 1
assessment guideline documents to assess the current condition of the steam 2
generators, as well as making predictions for future degradation that may occur.
3 And the last document Ill mention is the in situ pressure test guideline to actually 4
assist in determining tube integrity. Next slide, please.
5 At this point, Im going to change over to look at some of the 6
historical degradation mechanisms that have been observed in the industry.
7 Next slide, please.
8 This slide here, I just wanted to mention that there was a previously 9
identified in the NRC presentations, just identifying the various tube materials 10 and the trends toward the alloy 690 material for the replacement steam 11 generators. Next slide, please.
12 This slide provides a snapshot of causes of steam generator tube 13 repair, and by tube repair I dont just mean plugging up tubes, but also consider 14 any degradation that exceeds a repair limit. So, sleeving of tubes, rerolling, and 15 alternate repair criteria; basically trying to capture the more significant 16 degradation thats been observed in this particular plot. And Im plotting the 17 percentage of tubes versus the year. And its on percentage, so from this chart 18 you cant tell numbers of tubes, just the percentage on a given calendar year.
19 And it identifies the reasons for the repair, going from the early repair, which 20 would be the thinning, and then in the early 1970s, followed by preventive repair 21 or denting of the tubes. Then there was pitting type of repairs performed. So, 22 the degradation mechanism that you see on here mostly in the blue shaded area 23 is stress erosion and cracking, and that has been one item that has affected the 24 industry to monitor, and inspect for, and maintain the operation of steam 25
42 generators has been a focus on stress corrosion and cracking. Next slide, 1
please.
2 So, the next three slides are going to talk about the three different 3
tubing materials that are used in the steam generators here in the U.S. The first 4
chart here is showing the number of tubes repaired versus year, and in this chart 5
were actually showing the actual number. And you look at the scale on the left 6
side. It goes up to 9,500 tubes repaired in a given calendar year, and thats 7
important to point out now because when I talk about the 600 thermally treated 8
and the 690 thermally treated on the following slides, the scale is going to go 9
down from a 9,500 down to 200 or 300 tubes repaired. And the only thing to 10 point out on the mill annealed tubing is the stress corrosion cracking, is the 11 numbers, the larger numbers, that is shown in the chart in the gray shaded area.
12 Next slide, please.
13 This slide shows the alloy 600 thermally treated tubing experience 14 relative to repair. And youll see a lot of middle color blue, the mid blue in the 15 center, and that is reflecting structure wear, and the structure wear has been a 16 large number of repairs that have been performed on the 600. And I say large 17 number; thats relative. This chart, looking at this, its showing maybe 200 18 repairs is the maximum for the fleet of the 600 TT tubing which has in the order 19 of 17 units that are shown on this particular plot. One thing to note in the brown 20 or tan colored boxes to the right is the occurrence that has begun on stress 21 corrosion cracking in the alloy 600 thermally-treated tubing. Not all plants have 22 experienced the cracking that have thermally-treated tubing; however, we are 23 noticing that that is an occurrence that is starting to be observed. The next slide, 24 please.
25
43 This slide is showing the repairs on the alloy 600 TT. This is the 1
material of choice to date for the replacement steam generators. Represents 2
different numbers of plants that are represented by this chart. As the plants are 3
replaced in the early 1990s, mid 1990s, we were averaging maybe one to three 4
plants that are replaced per year. So, when we get to 2011, that represents a far 5
greater number of plants on the order of approximately of 50 or so represented 6
on the far right, in 2011. And were seeing from this chart that the predominant 7
mechanism, the yellow really stands out here, is the tube structure where thats 8
been observed at AVBs and support plate locations. Next slide, please.
9 This chart, Im switching from previous charts, were talking about 10 tubes. This chart is talking about indications. The number of indications -- there 11 can be many indications -- any current indications of tube wear that occur in one 12 tube, because theres many support locations along a tube length. So, what Ive 13 done here is Ive taken those steam generators and those plants that have 14 observed the greatest number of wear indications, and I plotted that here in this 15 chart. So, the chart goes up from zero to about 12,000 indications reported for 16 any given plant, and Ill focus on plant A to start with. For plant A in the lower 17 left, Im comparing the first ISI to the second ISI, and Im showing the change 18 between the first and the second to be very small. Realizing that -- and if we look 19 at the depths, there are depths of -- flawed depths of these, and only the largest 20 flawed depths tend to be repaired, usually over 35 or 40 percent through wall.
21 So, there are very few tubes that are repaired that are shown on this chart. Most 22 of these remain in service, are monitored, and have a very controlled growth rate 23 for the wear. And so the only difference from one inspection to another is the 24 change in the height of the bar showing in plant A, showing very little change; 25
44 meaning the same indications that were present in the first inspection are there in 1
the second and with very little new indications that were reported.
2 Now on plant B, theres a little bit of a change there. We see an 3
increase in the bar by about 50 percent of the height, going from about 6,000 4
indications to maybe 9,000 indications. So, thats showing an increase. And 5
then well jump over to plant E at the far right showing basically a threefold 6
increase in the number of indications. I wanted to point out one item on this 7
particular bar is that the blue on the far right, on plant E, which has the most 8
number of indications, less than 10 percent through wall. So, less than 10 9
percent through wall is very, very small, just able to be detected by the Eddy 10 current. Its very shallow. So, these will be monitored over time to determine if 11 they grow and at what rate they are growing, but you can see from this chart on 12 the percentage through wall.
13 The last plant Ill talk about is actually plant C and D. This unit has 14 actually only had one inspection, and Im putting this down here just to show 15 because of the numbers, and it fits the category of the top units for seeing wear, 16 that after one inspection of each of the two plants there just to show the 17 differences in the wear and the distribution of flaws. Next slide, please.
18 This is a chart that was shown earlier in a different format from the 19 NRC presentation showing the forced leak or outage was reduced. And I just 20 wanted to point out that except for the leak or outage in 2012, there was a period 21 of six years that had proceeded without any forced shutdowns due to tube leak.
22 And the trends of this going down, we talked about the improved inspection 23 programs to control leakage, and then also obviously the replacement of the 24 steam generators. Next slide, please.
25
45 And for conclusions Id like to say that the SGMP provides the tools 1
to develop technically strong steam generator programs which focus on 2
maintaining steam generator tube integrity. The tools include results of EPRI 3
research, guidance documents, and access to the worldwide steam generator 4
operating experience. The steam generator programs are continuously updated 5
based on new industry experience, industry guidance, and research results.
6 And, for example, Id like to mention that the industry guidance has been flexible, 7
flexible enough to address emerging issues that might not have been observed 8
before. We were able to handle the stress corrosion cracking observances, 9
foreign objects that get in the generators and cause wear of the tubes, as well as 10 tube-to-tube wear thats recently been observed, and that concludes my 11 presentation.
12 CHAIRMAN MACFARLANE: Thank you. Okay, next were going 13 to hear from Dr. Hitoshi Kaguchi, project director in the Nuclear Plant Production 14 Division for Mitsubishi Heavy Industries. Dr. Kaguchi.
15 HITOSHI KAGUCHI: Thank you very much. My name is Hitoshi 16 Kaguchi. I am a project director with Mitsubishi Heavy Industries. My 17 presentation today describes MHIs experiences related to tube vibration and 18 wear in steam generators during a faulty case of plant operation in Japan, in the 19 U.S., and also in the other countries. MHI has manufactured 116 steam 20 generators in total, and six of them were for the U.S. plants. Our steam 21 generators have experienced two major issues. One was flow-induced vibration 22 wear between the joist and the anti-vibration bar. We call them AVBs. And they 23 were observed in several plants in Japan, in the 1980s to the early 1990s. The 24 other was the tube rupture event due to fluid elastic instability, or we call it FEI, in 25
46 one plant in 1991. Until 2012, MHI steam generators have not experienced tube 1
wear issues since the early 1990s.
2 At first, I would like to explain about the early tube to AVB wear.
3 The wear was caused by fluid-induced vibration. The vibration occurred because 4
of the too large gap between the tubes and the AVBs. The nominal gap at the 5
time was over 10 mils. To prevent this tube to AVB wear MHI set the design gap 6
of the later steam generator to approximately 3 mils. For already installed steam 7
generators MHI replaced the existing AVBs with expandable type AVBs in 36 8
steam generators on-site in Japan. Similar repairs were also performed by the 9
other vendors in the United States.
10 The second major operating experience was a tube rupture event 11 that occurred in 1991, at Mihama Unit 2 in Japan, after 19 years of operation.
12 The direct cause of the rupture was a fatigue failure of a tube because of a out-13 of-plane fluid elastic instability due to improper installation of the AVBs during 14 fabrication. Over the time sludge accumulated between the tube support plate 15 hose and the tubes. This resulted in the tube became tightly fixed at the tube 16 support plate. Damping of the tubes decreased and the out-of-plane FEI studied.
17 After this event, based on the extensive research under the development 18 program, new design guidelines were developed by the JSME, Japanese Society 19 of Mechanical Engineers. And the standard was implemented by Japanese MHI.
20 Of course, the control of the AVBs becomes more strict after that.
21 Next I would like to explain about in-plane FEI in SONGS. The SONGS 22 steam generators were the largest steam generator fabricated by MHI. The 23 design, the SONGS steam generator was based on the established design 24 practices, ASME, and other industrial codes, and the detailed customer design 25
47 specifications. A major focus of the SONGS design -- SONGS steam generator 1
design was to minimize the tube wear. We changed the design nominal gap 2
between the tube and the AVB from our standard 3 mils to the 2 mils. This is the 3
-- 2 mils, is the cold condition and fixes the zero gap at the hot operating 4
condition. This was intended to reduce the -- prevent the tube wear caused by 5
out-of-plane FEI. MHI also confirmed a severity issue against the out-of-plane 6
FEI with conservative assumptions. The tube leakage at SONGS Unit 3 7
occurred due to tube-to-tube wear caused by in-plane FEI, and the in-plane FEI 8
discovered at SONGS is the first occurrence in the nuclear industry. It is also the 9
first evidence that the in-plane FEI conditions could be achieved in our operating 10 steam generator. For in-plane FEI to occur, three conditions must exist 11 simultaneously. The first one is the high steam quality. It means a very dry 12 steam. This is related to the tube damping. And the second is the high steam 13 velocity, flow velocity. This is related to the energy input to the tube. The last 14 one is the low contact force between the tube to AVB. This is related to the 15 effectiveness of the supports.
16 The steam quality and the velocity there depends on the size, and 17 the design, and the operating condition of steam generator. The low contact 18 force was the desire to the effort to minimize the wear due to out-of-plane FEI.
19 MHI applied tighter damage wear control and the process improvement to the 20 achieved uniformity distributed gaps. These actions result in unexpected in-21 plane FEI. MHI has confirmed that the condition found in the SONGS unit do not 22 exist in other MHI design steam generators, including operating steam 23 generators and the proposed USABWR design. SONGS Unit 3 had significant 24 tube-to-tube wear by in-plane FEI, but Unit 2 did not. MHI examined the reason 25
48 for this difference in detail. The same hydraulic conditions in both units were 1
almost the same. The only difference was the support condition between the 2
tubes and the AVBs. After the completion of the Unit 2 assembly, MHI improved 3
fabrication processes. The main difference was the damage control of AVBs.
4 MHI recently developed a very detailed three-dimensional model shown on this 5
slide. Using this model we confirmed that the small change in damage control 6
resulted in the large change in contact force. Contact force in Unit 3 are less 7
than half of the dose in the Unit 2. We now understand that the AVB supports in 8
Unit 2 are more effective to prevent in-plane FEI compared to Unit 3.
9 The following two slides show the improvement of the hydraulic 10 condition by changing operating condition. This graph shows the quality. Power 11 reduction show -- power reduction to 70 percent, for example, it would improve 12 the steam quality in SONGS steam generators by more than half. This would 13 bring the steam quality where it's in a range in other steam generators fabricated 14 by MHI.
15 The similar result obtained for the flow velocity. Flow velocity would 16 also be less than half, as reached in other steam generators.
17 My conclusion today, first, the in-plane FEI observed at SONG is 18 the first occurrence in operating steam generators. The second, MHI has 19 identified technical causes of tube wear from the in-plane FEI. Third, based on 20 the technical causes, we can say in-plane FEI can be prevented by reduced 21 steam quality, reduced flow velocity, and/or greater contact force between the 22 AVBs and tubes. Thank you very much.
23 CHAIRMAN MACFARLANE: Thank you. Okay, now we will move 24 on to hear from Jeff Fleck, who is manager of Nuclear Steam Supply System 25
49 Mechanical Engineering at AREVA. Mr. Fleck.
1 JEFF FLECK: Good afternoon. Today I would like to share with 2
you the background and experience that AREVA has in the area of tube vibration 3
and wear, starting on Slide 3. Since 1989, AREVA has designed and 4
manufactured replacement steam generators. And as a global organization this 5
experience includes components for both international and domestic utilities.
6 Next.
7 The components are different due to the overall size and design of 8
the plants for which they were designed, and include 67 recirculating steam 9
generators, and four once-through steam generators. However, the materials of 10 construction and the fabrication techniques are essentially the same, utilizing 11 advanced and higher performing materials than the original components as well 12 as enhanced fabrication practices and techniques. In the U.S. there are four 13 plants with a recirculating design currently installed, one more to be installed this 14 fall; and two plants with the enhanced once-through steam generator design 15 installed.
16 For the U.S. replacements, in-service tube inspections have been 17 performed, with most of the components receiving at least two. The U.S.
18 inspection methodology and techniques provide for identification of tube wear at 19 low depth thresholds as was shown previously, which allows for trending and 20 larger data sets for use in engineering assessments. Next.
21 In the AREVA U.S. components, tube wear has been identified at 22 various locations, including anti-vibration bars, tube support plates, outer bundle 23 supports, and also wear from foreign object intrusion. The operating experience 24 with the AREVA recirculating steam generators has not shown any tube wear 25
50 due to in-plane fluid elastic instability. Tube to tube wear in the once-through 1
design, however, has been identified and was summarized in NRC generic letter 2
2012-7, as well as in Mr. Karwoskis presentation earlier today.
3 The root cause for this mechanism is still in progress, but instability 4
is not a factor, based on the results of the three inspections performed at the 5
oldest plant and little to no increase in the flaw depths over time. Next.
6 As the designer of these components, AREVA has performed 7
studies and evaluations to determine the causes of the wear, in order to ensure 8
corrective actions have been identified and implemented as appropriate. Next 9
slide.
10 Moving into some considerations that AREVA takes when 11 designing a component, we work to an owner certified design specification that 12 defines the design requirements that must be met. As a class one vessel, the 13 component must also meet ASME requirements as well as industry and best in-14 house practices. Next.
15 Among the design requirements are those related to tube vibration 16 and wear, therefore the following conditions must be considered during the 17 design phase: thermal hydraulic, including flow rates, steam pressure, and void 18 fraction; support configurations, the numbers and the design of each; the tube 19 bundle configuration, for instance the tube spacing and the U-bend, otherwise 20 known as incrimination; material selection related to wear resistance, and the 21 coefficients of wear associated between material interaction; and flow-induced 22 vibration of responses of the tubes and other internal components of the steam 23 generator.
24 Included in the flow-induced vibration analysis, are stability 25
51 determination, turbulence, and nonlinear wear calculations, all of which must 1
meet the established acceptance criteria. In all cases conservatism is utilized in 2
the analysis to ensure adequate design margins are maintained in the 3
component. AREVAs design codes have been benchmarked through either 4
laboratory testing or comparative analysis with other established codes. If during 5
the design process changes are proposed or suggested, they are compared to 6
previously proven technology and thoroughly evaluated for any detrimental 7
effect. As a learning organization, AREVA utilizes lessons learned, corrective 8
actions, and operating experience for applicability, and we incorporate them as 9
appropriate. The final component design is certified as meeting the spec by a 10 professional engineer, and owner acceptance of the component is established.
11 Now Id like to move into our experience in the steam generator and 12 inspection and repair business that we have been engaged with for the past 30 13 years. This work also includes engineering functions associated with steam 14 generator assessment work. Our experience includes many of the challenging 15 mechanisms that existed in the original alloy 600 steam generator components 16 presented by Mr. Benson. Most recently, however, tube wear is one of the most 17 common mechanisms affecting the replacements. However, its been observed, 18 based on our experience, that tube wear is detectible at low levels using EPRI 19 qualified technique, has size uncertainties that are well quantified, has structural 20 correlations that are conservative and based upon laboratory testing of real 21 specimens, and typically has large margin to the tube integrity performance 22 criteria.
23 The industry framework defined by NEI 97-06 and the EPRI Steam 24 Generator Management Program establishes rigorous programmatic 25
52 requirements, especially in the qualification of people and techniques. It also 1
standardizes methodology for assessments and provides for operational 2
experience sharing and exchange. Finally, it also established the current steam 3
generator tech specs that were also referred to earlier. These tech specs are up 4
to date and appropriate for effective in maintaining safety margins, especially 5
those with tube wear.
6 For over a year AREVA has been involved in support of steam 7
generator activities at the San Onofre site. Tube-to-tube wear as a result of fluid 8
elastic instability and in-plane vibration is a new phenomenon that has required 9
significant inspections, engineering, and conservative tube repair strategies. In 10 Unit 2, the majority of the tube plugging was preventative and used conservative 11 decision-making based on the Unit 3 characteristics to establish margin for return 12 to service. Our condition monitoring and operational assessment was complete 13 and was submitted as part of the SCE CAL response letter. Based on our 14 calculations, at 70 percent power, there is margin to in-plane instability, which will 15 then mitigate the potential for additional tube to tube wear, as well as place Unit 2 16 into an operating regime that has demonstrated successful performance relative 17 to this phenomenon as presented by Dr. Kaguchi.
18 In summary, historically tube wear has been a manageable 19 mechanism in operating steam generators. Cycle lengths and tube integrity 20 margins have not been challenged, even in cases where large populations of 21 wear scars are affecting the steam generator. A common strategy taken for the 22 management of tube wear is to preventively plug tubes that are less than the 40 23 percent threshold criteria established in the plant tech specs. This increases 24 margin in the tube integrity -- the calculations that are performed to support the 25
53 next inspection interval. Finally, in our experience, the current industry protocol 1
programs and requirements provide sufficient safety margins for all steam 2
generator tube degradation mechanisms, including tube wear. Thank you. That 3
concludes my presentation.
4 CHAIRMAN MACFARLANE: Thank you very much. Okay, next 5
were going to hear from Damian Testa, who is project manager for Steam 6
Generator Management and Modification Programs at Westinghouse. Mr. Testa.
7 DAMIAN TESTA: Good afternoon.
8 CHAIRMAN MACFARLANE: Got to have the red dot there.
9 DAMIEN TESTA: Okay. Thank you. Good afternoon. Today Id 10 like to share with you some of Westinghouses experiences with respect to flow-11 induced vibration. With respect to design of Westinghouse steam generators, 12 weve manufactured generators for over 40 years. The potential for flow-induced 13 vibration is routinely analyzed in every steam generator that we design and 14 manufacture. This includes vibration in tubes, and also moisture separators in 15 dryers.
16 The input parameters for the methodology used in the FIV analysis 17 are documented in technical literature, and theres been also extensive testing 18 performed to support this analytical methodology and literature. Improvements in 19 the analytical methods have been made as operating experience and test data 20 have evolved over the years. Replacement steam generators have incorporated 21 enhancements in the U-bend assemblies and in all these generators 22 manufactured. Field modifications have also been performed and they have 23 been effective in resolving original steam generators flow-induced vibration 24 issues; things like complete anti-vibration bar replacements such as Dr. Kaguchi-25
54 san has mentioned earlier, and also modifications made in pre-heater designs.
1 Westinghouse manufacturing processes have improved for the 2
RSGs and also our new design steam generator such as the AP1000. The 3
advanced AVB design since 1990s have been incorporated in these models, and 4
that consists of tighter dimensional controls on compounds and improved 5
assembly oversight in the documentation at our manufacturing facilities.
6 With respect to our original steam generators, weve had relatively 7
good performance with a few observed issues. Weve had a limited amount of 8
AVB wear in different models. Weve had some short-term rapid wear early in 9
life due to manufacturing issues, but over the long term, with the few exceptions, 10 AVB wear has not challenged the pressure boundary or the tube integrity in 11 these generators.
12 I mentioned AVB replacements. Weve replaced AVBs, anti-13 vibration bars, which are the supports of the tubes, in 19 original steam 14 generators between the years of 1985 and 1993. This process was effective in 15 minimizing the AVB wear in the Model 51-and Model F-style steam generators.
16 This modification consisted of incorporating expandable AVB design into the 17 generators, which reduce the gaps. Other Westinghouse models during this 18 timeframe did not experience such issues as those two models.
19 With respect to our Model D3, we have an FIV issue that resulted in 20 a tube leak in a Westinghouse pre-heater-style steam generator. There were 21 field modifications made to that generator, to divert flow in the pre-heater. Also 22 our models D4 and D5 that are also pre-heater designs, the wear was not as 23 severe as the D3. There were no tube leaks; however, we did do tube 24 expansions to better support the tubes, and we also made modifications to split 25
55 the feed water flow to reduce the velocities as seen by the tubes.
1 With respect to replacement steam generators, theres been no 2
significant operational issues observed. A fraction of one percent of the tubes 3
have experienced AVB wear, which is a very low number, and many 4
Westinghouse RSGs have no AVB wear indications after one or more cycles of 5
operation.
6 Some significant events that have occurred; in 1983, we had AVB 7
wear that resulted in a tube leak in a Model 33 steam generator. That was 8
determined to be related to a manufacturing issue. We had a Model D3 pre-9 heater wear issue. It was caused by turbulence in out-of-plane fluid elastic 10 instability, and that was resolved by flow control in modifications to improve 11 tubing support. Also there was a tube rupture due to high cycle fatigue in 1987 in 12 a Model 51 generator. This was caused by denting at the top tube support plate 13 and a variation in the AVB insertion depth. It was addressed by analysis and the 14 installation of sentinel plugs, and stabilizers in a few tubes in this particular plant.
15 And there was an NRC bulletin that resulted in that -- 88-02. A rapid wear event 16 occurred in 1992 in a Model F. It was determined to be related to a 17 manufacturing issue. It was one steam generator in a three-loop plant. That was 18 resolve by replacement of all the AVBs in that particular generator, and the utility 19 decided to replace the AVBs in the other remaining generators at the same time.
20 With respect to our experience at SONGS, we performed an 21 evaluation for SONGS Unit 2 addressing tube wear at AVBs, tube-to-tube wear, 22 and the potential for in-plane instability. We concluded that tube-to-tube wear 23 observed in tubes in Unit 2 resulted from proximity of the tubes and out-of-plane 24 vibration and/or in-plane turbulence, and not in-plane instability. How we came 25
56 to that conclusion is based on the following; Eddy current data showed that there 1
was no extension of the wear scars beyond the width of the AVB itself, and not 2
only in these tubes, but other tubes in Unit 2 as well. Also, vibration due to in-3 plane instability will cause extension of the wear scars beyond the width of the 4
AVBs as observed in Unit 3. Additionally, the two tubes with tube-to tube wear 5
have no indications of the top tube support plate wear, as found with tubes that 6
had tube-to-tube wear in Unit 3. All this was included in our operational 7
assessment that we performed, and has been delivered to the NRC for review.
8 So in summary, we have observed issues related to FIV in the past 9
with our original steam generators. As a result of our experience with this wear 10 and fatigue issue over the past two decades, weve incorporated enhanced 11 designs, manufacturing and oversight into our RSG in recent new steam 12 generator manufacturing. We strive for zero wear for our design in 13 manufacturing, and as a result, we have minimal wear. Tube wear in the original 14 steam generators in service currently is managed in accordance with NEI 97-06, 15 and as a result of that, we feel that they are very safe to operate. That concludes 16 my presentation.
17 CHAIRMAN MACFARLANE: Good, thank you very much. Next up 18 is Pete Dietrich, who is the senior vice president and chief nuclear officer of 19 Southern California Edison.
20 PETE DIETRICH: Thank you, Chairman. Good afternoon. Im 21 pleased to have this opportunity to share our experience and explain what we 22 have learned at San Onofre regarding tube wear in our replacement steam 23 generators. Next slide, please. The tube wear we have observed at San Onofre 24 falls into two categories: what I will characterize and what we have characterized 25
57 as tube-to-tube wear; and what I will refer to as other tube wear mechanisms, the 1
normal wear that has been discussed in some of the other presentations. And I 2
will discuss each briefly in this slide to shape them. Tube-to tube wear occurs 3
when adjacent tubes vibrate and contact each other during operation. Other 4
plants have experienced tube-to-tube wear in the free span straight length of 5
tubes between tube support plates. At San Onofre, tube-to-tube wear has 6
occurred in the U-bend region for the first time in the industry. Other tube wear 7
includes more prevalent tube wear mechanisms occurring in the industry as well 8
as at San Onofre. This includes wear of tubes as the vibrate and contact anti-9 vibration bars, and as they contact tube support plates. The SONGS 10 replacement steam generators also experience another tube wear mechanism 11 resulting from vibration of thin retainer bars on the periphery in the U-bend 12 region, as well as one instance of wear due to a foreign object. The wear rates 13 resulting from each of these other mechanisms are within the experience of other 14 operating plants and the industry steam generator management program 15 described by the previous speakers. More precisely, the wear rates from these 16 mechanisms at San Onofre have been determined to allow safe operation of the 17 Unit 2 steam generators at 100 percent power for a full operating cycle, based on 18 proven regulatory framework and established industry precedent. Now, because 19 the type of tube-to-tube wear observed at San Onofre is unprecedented, I will 20 focus the majority of the remainder of my presentation on this mechanism. Next 21 slide, please.
22 This slide compares the tube-to tube wear observed in the two San 23 Onofre units. The depiction is a plan view of half the tube bundle showing 24 damage or wear indications in the U-bend region for one of the steam generators 25
58 in each of the units. Unit 3 steam generator, on the left side, experienced deep 1
tube-to-tube wear in multiple tubes, including through-wall wear on one tube. By 2
contrast, Unit 2, which is on the right side of the slide, operated twice as long as 3
Unit 3 and had shallow tube-to-tube wear in only one pair of tubes. The deep 4
tube-to-tube wear in multiple Unit 3 U-bend tubes resulted from in plane vibration 5
caused by fluid elastic instability. As you have heard from the earlier 6
presentations, this phenomenon has not been observed previously in an 7
operating steam generator. For Unit 2, to be conservative, Southern California 8
Edison is assuming that the single case of tube-to-tube wear in Unit 2 may also 9
have been caused by in-plane fluid elastic instability, although as you have just 10 heard, there is another plausible explanation for this instance of shallow tube to 11 tube wear. Next slide.
12 This slide depicts the critical elements in an operating steam 13 generator required for fluid elastic instability to occur. The three components 14 needed to act concurrently to cause the in-plane fluid elastic instability are loose 15 or ineffective support conditions, high steam velocity, and high steam dryness 16 which results in low damping. The size of each circle is indicative of the relative 17 presence of each. In Unit 3, all three conditions existed concurrently for a 18 significant number of tubes. The anti-vibration bar support conditions at Unit 2 19 were more effective than in Unit 3, therefore this circle is smaller in the three-ball 20 diagram on the right. I will discuss the reasons for this in the next slides.
21 Consequently, the overlap of the three conditions required for fluid elastic 22 instability was minimal for Unit 2, at 100 percent power, and deep tube-to-tube 23 wear did not occur in this unit. Next slide, please.
24 As we saw in an earlier slide, deep tube-to-tube wear from fluid 25
59 elastic instability occurred in multiple tubes in Unit 3 but not in Unit 2. This 1
difference in performance was caused by differences in fabrication between Unit 2
2 and Unit 3. The anti-vibration bars that support the U-bends are long, flat, and 3
relatively thin. They are formed into a V-shape, and thus can be subject to 4
twisting and bending. The Unit 3 anti-vibration bars were flattened after the 5
forming, using a press, and the force applied to the Unit 3 anti-vibration bars was 6
three times that of the force used in the Unit 2 forming of the anti-vibration bars.
7 The resulting contact between tubes and anti-vibration bars was thus minimized.
8 The effective of this difference was confirmed in pre-service measurements of 9
contacts between tubes and anti-vibration bars. Unit 2 had many more such 10 indications of contact. Next slide, please.
11 Mitsubishi Heavy Industries has developed a very detailed full tube 12 bundle model, which calculates the effect of the fabrication differences on contact 13 forces between the individual anti-vibration bars and tubes at steam generator 14 operating conditions. In the Mitsubishi Heavy Industrys presentation, you saw 15 that the effect on contact forces, and thus the relative looseness of the tubes is 16 substantial. Unit 2 has increased contact force and reduced looseness, and is 17 therefore less susceptible to fluid elastic instability. The more effective anti-18 vibration bar support in Unit 2 resulted in a full run cycle at 100 percent power for 19 approximately 12,500 hours0.00579 days <br />0.139 hours <br />8.267196e-4 weeks <br />1.9025e-4 months <br /> with minimal tube-to-tube wear. Anti-vibration bar 20 wear was also less extensive in Unit 2. For example, the total number of anti-21 vibration bar wear indications in both Unit 3 steam generators was about 50 22 percent greater than in Unit 2, even though Unit 2 had run twice as long. Next 23 slide, please.
24 We have learned quite a bit from the SONGS experience with our 25
60 replacement steam generators, and particularly that in-plane fluid elastic 1
instability can occur in operating steam generator U-bends. As importantly, we 2
have learned how to mitigate and avoid this new phenomenon in our operating 3
plants, namely prevent the concurrent presence of inadequate supports, high-4 steam velocity, and high-stream dryness. Ensure acceptable stability ratios for 5
in-plane fluid elastic instability using conservative inputs and assumptions. And 6
lastly, we provide adequate justification through testing when extrapolating 7
beyond proven operating experience. Next slide, please.
8 As a plant operator, I have operational control over two of the three 9
components needed for in-plane fluid elastic instability. Specifically, reducing 10 power can reduce steam velocity and reduce steam dryness sufficiently to 11 preclude in-plane fluid elastic instability since the conditions for fluid elastic 12 instability no longer occur concurrently. Next slide.
13 The previous slides summarize some of the specific technical 14 understanding gained. Finally, I would like to reaffirm the context for this 15 understanding. At San Onofre existing industry and plant measures, specifically 16 operator response, for detecting a leak and shutting down the plant promptly 17 worked effectively at San Onofre. The existing steam generator management 18 program, as discussed by previous presenters, is effective and broad enough to 19 address emergent issues such as in-plane fluid elastic instability, as well as the 20 more prevalent anti-vibration bar and tube support plate wear. And also 21 independent and open expert participation is very important. Thank you.
22 CHAIRMAN MACFARLANE: Thank you very much. All right, next 23 well hear from -- is it Michel or Michael?
24 MICHEL PETTIGREW: Michel.
25
61 CHAIRMAN MACFARLANE: Michel Pettigrew, who is adjunct 1
professor at Ecole Polytechnique, Montreal, and a principal research engineer 2
Emeritus for the Atomic Energy of Canada Limited, Chalk River. Mr. Pettigrew.
3 PETE DIETRICH: Its on. Yes, sir. Youre ready to go.
4 MICHEL PETTIGREW: Good, all right. Could I have the first --
5 yes, thank you.
6 ANNETTE VIETTI-COOK: Speak into the microphone.
7 MICHEL PETTIGREW: Yeah, Ill get closer, yeah. My talk is going 8
to be about vibration and in particular fluid elastic instability. On the left of the 9
figure, you see a steam generator. I think you all know enough about steam 10 generator. What Id like to do is to focus on a peculiarity of a steam generator.
11 So, on the right you see a steam generator being assembled. And what you see 12 there is U-bends that are being inserted in the steam generators. So, you notice 13 the U-bend -- the plane of the U-bend is being installed, and on top of the U-14 bends are bars. They are anti-vibration bars. And so you can see here that from 15 the point of view of out-of-plane motion, the tubes are really very well supported 16 because you have a large number of bars all around; but from the point of view of 17 in-plane motion, theres really no positive restraint here to prevent the tube to 18 move in the in-plane direction. Essentially, it relies on friction forces to limit the 19 vibration. And that is one part of the problems that we are discussing here.
20 Could I have the next one?
21 Okay, the other concern, if you like, is the vibration excitation, 22 which is in this case, is in the form of flow velocity, high-flow velocities. It may be 23 a bit difficult to see on this slide, but on the right, at the top near the U-bend, 24 there is an area thats whiter than the rest. And this corresponds to a flow 25
62 velocity of about 6.5 meters per second, or 20 feet per second. So, there are 1
quite high-flow velocities in that region of the steam generator coupled with the 2
possibility of less-than-effective support. Its the kind of thing that can lead to 3
problems. Next slide, please.
4 Now, weve taken a very pragmatic approach to these problems.
5 Essentially, we have subjected various section or markup of various parts of 6
steam generators to operating conditions. So, in this case in here we have, on 7
the right-hand side, there is a test section here and its essentially a cantilever 8
tube bundle. And we are subjecting it to flow and we can, with this loop in here, 9
we can have air, water, or a mixture of air-water, so we can simulate liquid, two-10 phase and gas flow. And we subject the tube bundle to these flow conditions 11 and we reach instability that way. And if I could have the next slide.
12 Okay, so this is now looking at the end of the tube bundle and there 13 is a Plexiglas porthole there, if youd like that allows us to look at the end of the 14 tube. Thats the free end of the tube. The other end is clamped. And youll see, 15 if you can get the simulation going -- okay. You see whats happening in here.
16 This is condition of instability. Its a square tube bundle. The flow is going up; its 17 liquid, so in that case its water. Okay, so now what were going to do is increase 18 the flow velocity by 10 percent. Were not changing anything at all; were just 19 increasing the flow by 10 percent. So, could you -- would you do it, okay. Can 20 you -- yeah. You see, now, we have instability. Its the same condition except 10 21 percent more flow velocity, and you see we have instability now in a completely 22 different mode of instability, all right?
23 Okay, from an academy point of view, this is all very interesting.
24 From a practical point of view, youll really want to avoid this in your component.
25
63 Theres no question about it. You have amplitudes here where you have tubes 1
contacting each other, meaning vibration amplitude in the order of a quarter of an 2
inch or more.
3 Okay, can I have the next one? Here, we analyze this information 4
of fluid elastic instability. We present it in the form of an instability diagram, 5
which you can see now. So, on the Y-axis, you have the dimensionless flow 6
velocity; its flow velocity divided by frequency, diameter. And on the X-axis is a 7
mass-damping parameter, its a dimensionless mass-damping parameter. So, 8
what Ive done here is over the years Ive reviewed the literature and come up 9
with something, like 500 data points, which I have analyzed and put on this graph 10 in here. And the message is clear here. If youre below the lines -- theres a line 11 there, if youre below the line, thats stable; if youre above the line, its unstable.
12 So, you want your component to be below the line, not above, you know, its that 13 simple.
14 Okay, the next slide is going to -- next one -- is going to be whats 15 happened now in two-phase flow and in in-plane direction. Now, Id like you to 16 focus on the figures on the right; Figures C and B, okay? In this case the tubes --
17 thats a cross section across a test section, and the tubes that are numbered are 18 instrumented tubes and they are flexible tubes. All the tubes that are not 19 numbers are essentially rigid. In C you have a column of tube, which we could 20 constrain to vibrate only in the in-plane direction, or allow it to vibrate in the 21 axisymmetrical fashion in every which way. On D we have two half columns 22 there, and they are constrained to vibrate only in the in-flow direction or in-plane 23 direction. So, the next slide shows the results.
24 Okay, so if you look at the upper part result, we have a column of 25
64 tube that is constrained to vibrate in the in-plane. And on the right, top right, you 1
see the response. There is really no evidence of instability; whereas, if for the 2
same column, I allow it to vibrate every which way, then it really goes unstable as 3
it shows on the left-lower diagram. In the left-lower diagram you have the 4
vibration response times the velocity, starting by the left, its in liquid flow. And 5
on the very right, its at 95 percent void fraction, and then in between -- in 6
between. But clearly, you know, you do not have instability for a single column 7
thats constrained to vibrate in the flow direction, the in-plane direction, and you 8
do if they are un-constrained, although the frequency in that case was a lot 9
higher, meaning that the rigidity of the tubes were much higher. So, in spite of 10 that, it really goes unstable.
11 Now, the next slide shows a comparison between one column and 12 two columns. One column is on the left and the top, two columns is on the right 13 at the bottom. And in both cases here, were only allowing the tube to vibrate in 14 the in-plane direction. And of course, for the one column, then it doesnt go 15 instable, as weve shown before; but if you have two columns, then it does go 16 unstable, as you can see on the drawing on the left bottom. On the left bottom, 17 you have vibration amplitude versus flow velocity. And theres three lines, and 18 here one is void fraction of 80 percent, the next one is 90 percent, and the last 19 one is 95 percent. Now, were going to see an animation of this, if I can get the 20 next one.
21 So, lets go on the left in here and get it going. You see on the left 22 this is the single row, but thats allowed to vibrate every which way in an 23 axisymmetrical fashion. And you see it goes unstable. Whereas, if the tube will 24 constraint to vibrate only in the in-plane direction, it would not go unstable as 25
65 weve seen before. Now, if we look at the two-column one now. You see the 1
two-column one does go unstable. Thats really the interesting part in here is 2
that one column one does not go unstable; two-column one does go unstable.
3 What weve shown in here is that you can have fluid elastic instability in the in-4 plane or the in-flow direction in two-phase flow, but sometimes its not so easy to 5
do. Its a lot more difficult to get fluid elastic instability, to achieve fluid elastic 6
instability when the tubes are constrained to vibrate in the flow direction. If I 7
could have my final slide?
8 Okay, so this is a similar diagram of instability on the Y-axis; theres 9
a dimensionless flow velocity on the X-axis in the mass-damping terms. And the 10 bottom line is for results from two bundles that were allowed to vibrate 11 axisymmetrically; whereas the upper line is really to look at what happens when 12 you are constraining the vibration into the in-plane direction. So, you have to go 13 to a higher velocity to achieve instability when the tubes are on constraints. So, 14 its more difficult to get in-plane instability, but its possible to happen. Thank 15 you. Thats all Ive got to say.
16 CHAIRMAN MACFARLANE: Okay, thank you very much. And 17 then our final speaker this afternoon is Daniel Hirsch, president of the Committee 18 to Bridge the Gap, and lecturer at the University of California-Santa Cruz. Mr.
19 Hirsch.
20 DANIEL HIRSCH: Chairman Macfarlane, members of the 21 Commission, thanks very much for the invitation to be here today. I am a lecturer 22 on nuclear policy at UC Santa Cruz and president of the Committee to Bridge the 23 Gap, but the views here are my own today. If I could have Slide 3. This is 24 obviously the San Onofre reactor. When it is operating, there are billions of 25
66 curies of radioactivity inside each of those domes, and outside of them reside 1
8.5 million people within 50 miles, about four times as many as reside near 2
Fukushima. Your job, that goes without saying, is to assure that that radioactivity 3
remains inside those domes and it never gets out to expose those people. Next 4
slide, please.
5 Steam generators are an absolutely critical safety feature for 6
assuring that that happens. They perform two vital functions. One is they are 7
necessary for cooling the core, preventing the fuel from melting or releasing its 8
radioactivity. Secondly, they provide a direct pathway out of containment to the 9
environment. So theyre unique. They can both cause the melt and they can 10 provide a pathway for that radioactivity to expose people. That must never 11 happen. Next slide, please.
12 The steam generators at San Onofre are of the recirculating type.
13 And so, as youve heard before, there are four basic places where you can have 14 wear. Theres a free-span area in the U tube region where the tubes can bang 15 against each other, if the thermal hydraulic conditions are bad, but there are also 16 tube support plates, anti-vibration bars, and retainer bars where they can rub 17 against those supports. Unfortunately, were having all four kinds of damage at 18 San Onofre. Next slide, please.
19 Edison makes two claims, and youve heard their entire team 20 pushing for this restart of Unit 2 at 70 percent power. The two fundamental 21 claims that they make to support that -- the first -- next slide -- is that the wear in 22 Unit 2 is far less extensive than the wear in Unit 3, that there were 300 tubes with 23 unexpected tube-to-tube wear in Unit 3 and only two tubes with minor wear in 24 Unit 2. Next slide, please.
25
67 For months, however, Edison and NRC staff refused to release the 1
actual data about the degree of damage within Units 2 and 3. It took Senator 2
Boxers intervention for these data to be revealed. And youll see very quickly 3
why when you look at these tables. Yes, there were only two tubes showing 4
tube-to-tube wear in Unit 2, but there were thousands of indications of wear at 5
the anti-vibration bars in the tube support plates. There are 1,600 tubes that 6
have been damaged just in this first cycle of operation in Unit 2. And as youll 7
see from the next slide, there are about 1,800 tubes that have been damaged in 8
Unit 3. And the next slide you can just see this graphically that theres very little 9
difference. The number of tubes damaged between Unit 2 and Unit 3 are very 10 similar. Unit 3 has a slightly higher fever, but both of these are very sick 11 reactors. They both need to be in intensive care. Next slide.
12 So, when those data were released, Edison made a second claim 13 that the nature of the support structure wear is not unusual in new steam 14 generators and is part of the equipment settling in. So, I asked NRC staff if they 15 have any data about this. They said theyve heard the same claim, said it was 16 based on anecdotal information that they had no data. So, my students and I 17 went and assembled the data that, frankly, the NRC staff should have, and the 18 results are quite extraordinary. Next slide.
19 You will see and in the slides that follow that rather than being a 20 normal amount of wear it is very much beyond the norm -- orders of magnitude 21 beyond the norm. Next slide.
22 You will see, for example, here that the number of indications of 23 wear for the steam generator tubes, the median is four nationally. And for Unit 2, 24 the one thats supposed to be better and good to go, they have over 4,700 25
68 indications of wear, a thousand times more. Next slide.
1 The number of damaged steam generator tubes in Unit 2, again the 2
good one to go supposedly, is about 1,600. The median nationally is four. And 3
lastly, the next slide, youll see -- one more slide, yeah, thats it, Im sorry -- that 4
there are 510 tubes that have been plugged in Unit 2 when the median nationally 5
is zero. There were more tubes plugged in one cycle of operation for the new 6
steam generator in Unit 2 than in all of the new steam generators in the country 7
combined.
8 So, an additional claim has now been made by Edison thats 9
saying, Yeah, we have a lot of wear in Unit 2, but this levels off over time. I 10 asked NRC staff if they have information as to whether thats true, they said, 11 again, theyve heard it anecdotally, but they had no data. So, I had to have my 12 students accumulate the data once again. And to just give you one example, the 13 Palo Verde plant, which Edison has identified as one that they claim is similar, for 14 Units 1, 2, and 3, the number of tubes damaged dont level off. They continue to 15 increase, and indeed the rate of increase continues to increase, generally, from 16 one in-service inspection to the next. So, there go those claims.
17 Why does it matter? Well, for two reasons. San Onofre in just one 18 or two years has experienced more damage than steam generators normally 19 experience in decades. And in one to two years, theyve chewed through nearly 20 half of their 8 percent plugging limit and they have thousands of indications of 21 wear on tubes that have not been plugged that if this wear continues, theyll have 22 to plug as well. They concede that the wear is due to random vibration, not the 23 fluid elastic instability causing the tube-to-tube damage, and it will continue to 24 concede if they restart. So, you have steam generators that cannot run for very 25
69 long even if theres no breakage, no disaster, no release. But gross failure is 1
clearly possible. Edison, as youve heard, claims that the reason Unit 3 is in 2
somewhat more trouble than Unit 2 is that the supports were more effective in 3
Unit 3 than 2, but supports are exactly whats getting worn down in Unit 2. And 4
so, that contact force is diminishing, that support is diminishing; and the problem 5
with the tube-to-tube wear is its sudden and unpredictable. You have a nice 6
ramp function perhaps for other kinds of wear, but for the tube-to-tube its a step 7
function. So if you allow them to restart Unit 2 and you get more of that support 8
wear loosening the fit, you are running the risk of this running out of control with 9
the damage thats not predictable or controllable. Next slide, please.
10 Perhaps the most extraordinary aspect of Edisons restart request 11 is found in the following three sentences from their transmittal letter. First -- next, 12 please. They -- keep going, sorry. Again, again, again, again, here we go. They 13 say that they plan to run at 70 percent power for five months is not a fix but an 14 interim compensatory action. Next slide.
15 They say after those five months, they want to shut down and see if 16 their theory is correct. This is clearly experimenting with safety. But the critical 17 one is the next sentence. They say, In addition, Edison has established a 18 project team to develop a long-term plan for repairing the steam generators. So, 19 Edison knows that those steam generators need to be repaired or replaced and 20 is asking you to let them run without repairing or replacing them. And I would say 21 to you, and its in this next slide, that you should simply say no. It would be 22 unwise to permit San Onofre Unit 2, with its critical steam generators that need 23 repair or replacement, to operate without repairing or replacing them.
24 Now, last couple comments in the last minute or two I have. This 25
70 episode has demonstrated not simply that the steam generators at San Onofre 1
are damaged and need repair, but theyve exposed some breakage in the NRCs 2
regulatory structure itself that needs to be repaired. I am told that it took the 3
NRC staff a total of only one day, when they finally looked at this matter, to 4
determine that the computer code projections were wrong. One day of review, a 5
billion dollars of expense, and a steam generator that had eight tubes that would 6
have burst that wouldnt have been safe. One day of review that was bypassed.
7 And Edison is now asking you to do the same thing. They want you to rely on 8
the computer models. Theyre now relying on the ATHOS model, which also 9
failed to predict the problem. The same people that designed the steam 10 generators that failed, that operated and approved them, the two other 11 companies that have okayed this, the consultant all telling you, Yeah, maybe it 12 wasnt done right, but trust us, we can do it again. There are 8.5 million people 13 on the other side of those containment domes. And the only way it would be 14 appropriate would be to make sure that that is really safe. If you had a car that 15 had brakes that were failing, you would not get away with saying, Ill just keep it 16 off the freeway and drive at 50 miles an hour. Youd have to get the brakes 17 repaired. With a car you can only kill a few people. There are hundreds of 18 millions of curies of cesium and strontium and iodine inside those domes that 19 youve got to keep in. And letting them run with damaged steam generators 20 without repairing them would be very unwise. Thank you.
21 CHAIRMAN MACFARLANE: Thank you very much. Okay, well 22 move on to questions and start with Commissioner Apostolakis.
23 COMMISSIONER APOSTOLAKIS: Thank you very much for the 24 presentations. There were a couple of statements that confused me a little bit.
25
71 Mr. Fleck, on your Slide 15, you say that at San Onofre tube-to-tube wear due to 1
in-plane fluid elastic instability is a new phenomenon, but youre implying that this 2
is what happened, correct?
3 JEFF FLECK: Yes, sir.
4 COMMISSIONER APOSTOLAKIS: Okay. Then we go to Mr.
5 Testa, who says on Slide 14 that Westinghouse concludes that the problem was 6
the result of out-of-plane vibration and/or in-plane turbulence and not in-plane 7
instability. And then Mr. Dietrich says that -- you heard of their alternative 8
interpretations -- SCE goes with instability. Well, first of all is there a 9
disagreement as to what caused the wear? I mean, youre saying it was in-plane 10 turbulence, and you say it was in-plane instability. And Im not an expert on 11 these things, but the words are different.
12 JEFF FLECK: I dont think theres a disagreement. I think what we 13 have done in our operational assessment and our strategy is to conservatively 14 assume that it is tube-to-tube wear, based on the results from a similarly 15 designed and operated unit next door. So, we were not -- based on the Eddy 16 current results that we have available to us, it was determined that we would 17 conservatively assume that those two tubes did experience some level of 18 instability and as a result wore against each other during that operating time 19 period.
20 COMMISSIONER APOSTOLAKIS: Whats the difference between 21 in-plane turbulence and in-plane instability, Mr. Testa?
22 DAMIAN TESTA: I would call those two the same, in-plane. The 23 question is in-plane or out of plane.
24 COMMISSIONER APOSTOLAKIS: But you make it very clear 25
72 here. You say, In-plane turbulence and not from in-plane instability, so that tells 1
me that theyre different, and youre saying not different.
2 DAMIAN TESTA: Yes, well, yes, thats correct. We do not believe, 3
based on the evidence, that you had in-plane instability. And that was just based 4
on --
5 COMMISSIONER APOSTOLAKIS: AREVA, says that --
6 DAMIAN TESTA: -- any current evidence that I quoted in my 7
presentation. Its an approach we took thats less conservative than the 8
approach that AREVA took, which is a more conservative approach.
9 COMMISSIONER APOSTOLAKIS: Why is it more conservative to 10 assume instability? I mean, Mr. Dietrich said the same thing. Its more 11 conservative in what way?
12 JEFF FLECK: Well, from the standpoint that Unit 2 falls into the 13 realm of being susceptible despite having differences between Unit 2 and Unit 3, 14 fabrication and other things that were mentioned earlier. Thats --
15 COMMISSIONER APOSTOLAKIS: Yeah, but if I decide that the 16 cause was in-plane instability, youre telling me that -- no, if actually, -- if there is 17 a question whether is due to in-plane turbulence versus in-plane instability, 18 youre telling me its more conservative to assume in-plane instability. Mr.
19 Dietrich, you have a comment on that?
20 PETE DIETRICH: Yes. First, Commissioner, I want to be clear 21 that the two presentations that youre referring to refer to conclusions about Unit 22 2 --
23 COMMISSIONER APOSTOLAKIS: Yeah.
24 PETE DIETRICH: -- and actually what occurred in Unit 2. Both 25
73 companies have concluded that what occurred in Unit 3 was clearly due to in-1 plane fluid elastic instability.
2 COMMISSIONER APOSTOLAKIS: Okay.
3 PETE DIETRICH: So these conclusions are about Unit 2.
4 COMMISSIONER APOSTOLAKIS: Yeah.
5 PETE DIETRICH: Westinghouse concluded that perhaps during 6
the installation process of the steam generators, these two tubes, where we saw 7
the very minimal indications, just at the edge of detection, might be because of 8
what we call tube-to-tube contact, where the tubes came into contact with each 9
other after fabrication, during shipping, and during installation. And that is as the 10 steam generator went into service, they wore as they moved apart. We feel that 11 that is not as serious a degradation mechanism as in-plane fluid elastic 12 instability. So, in my comments where I said we have assumed the conservative 13 route that Unit 2 is seeing very small amounts of fluid elastic instability in these 14 two tubes, thats what helped us shape our proposed corrective actions, the 15 corrective actions for our proposed period of operation. So, that is how we 16 managed the differing opinions about Unit 2, but again, both companies agree on 17 what occurred and caused the deeper damage on Unit 3. Does that help?
18 COMMISSIONER APOSTOLAKIS: Yeah, you agree?
19 JEFF FLECK: Yes.
20 COMMISSIONER APOSTOLAKIS: Mr. Pettigrew, your 21 presentation was what I used to see in my previous life. You spoke to the slides, 22 which I congratulate you for.
23 MICHEL PETTIGREW: [laughs]
24 COMMISSIONER APOSTOLAKIS: But on Slide 6, if we can have 25
74 it back, you saw an animation there and it wasnt clear to me what was in-plane 1
and what was out-of-plane? Help me understand it. Slide 6 of Mr. -- or Dr.
2 Pettigrews slides, please? Okay, so lets go to the animation. Are they now --
3 whats happening there? Which one is in-plane and which is out-of-plane?
4 MICHEL PETTIGREW: Theyre mostly out-of-plane.
5 COMMISSIONER APOSTOLAKIS: So, horizontal is out of plane?
6 MICHEL PETTIGREW: Yeah, yeah, thats right. The flow was 7
upward in here and thats mostly out-of-plane. And more of the time in single 8
phase flow -- thats in single-phase flow --
9 COMMISSIONER APOSTOLAKIS: Okay.
10 MICHEL PETTIGREW: Okay, more of the time when you get 11 instability, its in the out-of-plane or what is called the lift direction, thats the 12 terminology thats used.
13 COMMISSIONER APOSTOLAKIS: And this is something that 14 people knew -- people in the business knew that you could have out-of-plane 15 vibrations?
16 MICHEL PETTIGREW: Yeah.
17 COMMISSIONER APOSTOLAKIS: My understanding is that in-18 plane is a new phenomenon as people claim, correct?
19 MICHEL PETTIGREW: Well, certainly in two-phase flow, yeah, it is 20 a new phenomenon, yeah.
21 COMMISSIONER APOSTOLAKIS: And then is the next slide 22 animated, too?
23 MICHEL PETTIGREW: No, the previous one is.
24 COMMISSIONER APOSTOLAKIS: Five, lets look at 5.
25
75 MICHEL PETTIGREW: Yeah. Okay, 5, essentially, the first sign of 1
instability. And what we did from the -- the vibration in this case is mostly in the 2
in-flow direction, which is a little unusual. Most of the time its in the --
3 COMMISSIONER APOSTOLAKIS: So this is in-plane?
4 MICHEL PETTIGREW: This is vibrating in-plane, but if you look 5
carefully enough, there is a component out-of-plane.
6 COMMISSIONER APOSTOLAKIS: Im sorry, what? That?
7 MICHEL PETTIGREW: There is a component out of plane.
8 COMMISSIONER APOSTOLAKIS: Youre right, youre right.
9 MICHEL PETTIGREW: Okay. So, in the next slide then all we did 10 is increase the flow by 10 percent. Everything is all the same, and we get a 11 different mode of instability. And the point that Im making is that theres, you 12 know, possibly and quite a number of different modes that you could have. And 13 from an academic point of view, thats all very interesting. But from a practical 14 point of view, you just want to avoid all of this. And so essentially, its the first 15 one thats reached, okay, we cant go any further. So we use that as a guideline.
16 We want to be below the lowest critical velocity for instability all the time.
17 COMMISSIONER APOSTOLAKIS: Thank you very much.
18 Chairman?
19 CHAIRMAN MACFARLANE: Okay. Thank you. On to 20 Commissioner Magwood.
21 COMMISSIONER MAGWOOD: Thank you, Chairman. Thank -- I 22 thank all of you for your presence today and your presentations. Youve come 23 from all over to talk about this issue, so we appreciate that. I wanted to follow up 24 on my colleagues line of question with Mr. Pettigrew. I had some of the same 25
76 questions. I appreciate you going over that. But one -- you heard my dialog with 1
the staff earlier, I think, about our ability to model some of these phenomena.
2 They seem rather complicated, this instability. Is this something that you have 3
looked at modeling? Is this something that we can model? And, you know, 4
where are we in doing that?
5 MICHEL PETTIGREW: Yeah. Yeah, well, thats a good question.
6 Weve been trying to model these phenomena for quite a number of years, and I 7
guess weve made some progress, but so far we havent been able to predict 8
what we have seen in the lab, you know, within plus or minus 10, 20 percent, and 9
were out by a factor of 50 percent or more. And its particularly difficult in two-10 phase flow, because in two-phase flow, you have an extra parameter, which is 11 void fraction or steam quality, if you like.
12 And then youve got also to talk about flow regime. What kind of 13 flow regime exists inside the tube bundle? Okay, so very little work has been 14 done on that, because its very difficult to do, for one thing, and in cross-flow in 15 particular. So you -- weve made an attempt at looking at the tail of this with --
16 experimentally to start with, and weve built test sections with tubes about twice 17 the size of steam generators so that we can look at the detail of whats 18 happening as the flow goes through the tube bundle. And this was revealing 19 actually. What you find is that the flow tends to create a path between the tube, 20 which you could represent by a series of 60-degree elbows. And you find that 21 the void fraction distribution across that gap -- we took 19 measurement points 22 across the gap at every millimeter. And what you find is that the void fraction 23 distribution goes from being high on one tube and low on the other tube, and the 24 next tube is the other way around. So theres a tremendous mixing taking place 25
77 in the tube bundle, and if youre looking at the mixture of the flow, its really quite 1
a fine flow.
2 Now, between -- now Im talking about the flow going between the 3
tube, but between -- so the -- an upstream and downstream tube, if you like.
4 Then theres a region here of more stagnant flow and more -- lower void fraction 5
tends to be mostly liquid. So whats happening is that the flow on one end is 6
streaming between the tube and the tube bundle, and in between, between 7
upstream and downstream tube, theres an accumulation of liquid there and 8
some cyclic activities as well, which, again, is something that we were not 9
expecting. We were expecting random turbulence excitation to be quite broad 10 banded, and what we found is that there is some periodicity in this two-phase, 11 flow end tube bundle.
12 COMMISSIONER MAGWOOD: Right, when you were running 13 your experiments, was -- were you able to identify transition points where the 14 flow took you either to in-plane or out-of-plane instability?
15 MICHEL PETTIGREW: No, I dont think so.
16 COMMISSIONER MAGWOOD: Really?
17 MICHEL PETTIGREW: Yeah. Yeah.
18 COMMISSIONER MAGWOOD: So that sounds like an area of --
19 for additional --
20 MICHEL PETTIGREW: Well, exactly. Well, whats being done 21 now, as you may know, we have a so-called share program at the Ecole 22 Polytechnique in Montreal. So weve set up a lab specifically devoted to look at 23 two-phase flow-induced vibration, and so we have developed instrumentation to 24 do that, and doing experiments, and some modeling. And to model two-phase 25
78 flow, then what you have to do is to look at the basics. So theyre doing 1
experiments now on one bubble in a liquid environment and looking at what that 2
bubble is doing, and eventually, now theyve added -- theyve got five bubbles 3
now -- so looking at whats happening, you know, when you have a train of five 4
bubbles in the liquid. And I think this is all, you know, very exciting work to do, 5
but its going to take some time before we can take the results of that kind of 6
work and fit it into design guidelines.
7 COMMISSIONER MAGWOOD: I appreciate that as a --
8 MICHEL PETTIGREW: Yeah.
9 COMMISSIONER MAGWOOD: May have to put you on my visit 10 list, I think. Yeah, let me turn to Mr. Testa and Mr. Fleck. I mean, is work like 11 this part of your everyday lives? I mean, do you look at work such as Professor 12 Pettigrews as youre thinking about future designs? Does -- do you get that level 13 of detail in your work? Yeah, you please go first, Mr. Testa.
14 DAMIAN TESTA: Weve performed similar tests in the past, and 15 thats what our analytical codes are based on, the data from the tests weve 16 performed in the past. They may not be as extravagant as Dr. Pettigrews, but 17 yeah, we perform tests, and thats what our models are based on.
18 COMMISSIONER MAGWOOD: I think you mentioned during your 19 presentation that you model -- routinely model fluid instability, you know, fluid-20 induced vibration, and you didnt really specify whether you modeled both in-21 plane or out-of-plane. Im assuming you just model out-of-plane or is --
22 DAMIAN TESTA: Only out of plane.
23 COMMISSIONER MAGWOOD: Okay.
24 JEFF FLECK: Same here. And our analysis codes also are based 25
79 upon testing that was performed in mockups and boilers in France, which is 1
where most of the design work occurs for the replacements. But they -- those 2
tests were performed in the late 80s and early 90s to validate some of the 3
design changes that they were making to the components.
4 COMMISSIONER MAGWOOD: Let me stay with the vendors for a 5
moment. What -- recognizing that, you know, all of you are focused very much 6
on whats happened with the fluid -- with the in-plane fluid-induced instability, but 7
looking -- sort of looking down the road, whats the next step? Where are we 8
going with steam generator design? I mean, have we -- we still have some to 9
replace there are new reactors on the drawing board. Whats the next step?
10 Where do we go from here?
11 DAMIAN TESTA: Want me to go first?
12 JEFF FLECK: Either way.
13 DAMIAN TESTA: Basically, were using the same tools that weve 14 used. Were just staying within our comfort levels. Were not pushing our design 15 limits. Were staying with what we know, whats been proven to work in the past.
16 COMMISSIONER MAGWOOD: So weve seen the apex of PWR 17 steam generator design? This is it? This is as good as it gets for that?
18 DAMIAN TESTA: Well, you can always improve on fluid elastic or 19 thermal hydraulic codes; and theres some new codes that are in the works, and 20 were looking at those. Theyre currently being developed, but whether theyll be 21 as good as what were using or not, well have to evaluate them when they come 22 out.
23 COMMISSIONER MAGWOOD: And Mr. Fleck, whats next in the 24 design?
25
80 JEFF FLECK: In the same context, we are in France developing a 1
new thermal hydraulic code. Thats been underway, but as you might imagine, 2
the development of a code that has to handle so many variables and these 3
conditions that are very uncertain, its time -- you know, you have to vet the 4
process and make sure that, again, youre staying in the bounds of what youve 5
known and your technology that youve used, and continually use that to 6
benchmark anything new that youre working on or that youre developing. But 7
as far as the pinnacle of the replacement market or the replacement design, I 8
would say that most of it is fairly standard, you know, at this point. I dont think 9
theres anything outside of the norm that anyone is looking at.
10 COMMISSIONER MAGWOOD: So pretty much the same as with 11 MHI?
12 HITOSHI KAGUCHI: Yeah, we have all sorts of experience here at 13 SONGS, and the, as I said, the three factor flow and the quality, dryness, and 14 also the contact forces. So we have to be careful about these three things. And 15 two things -- and the flow and the quality is based on the TH, thermal hydraulic 16 condition, so it can be more precisely analyzed. And also we have to be very 17 careful because we are very concerned about the gap between the tube and the 18 AVB -- to vibrate in this direction. Now, if the contact force is very weak, tube 19 cant start this way, so we have to also be careful in the manufacturing in this 20 point. Yeah.
21 COMMISSIONER MAGWOOD: All right, thank you very much.
22 And just -- Mr. Hirsch, you mentioned your students several times. What kind of 23 course was it that was --
24 DANIEL HIRSCH: I teach Introduction to Nuclear Policy, so this 25
81 was a useful project for them to try to accumulate some of the data that NRC has 1
but hasnt assembled itself, and in the process identified, I think, some useful 2
suggestions for you folks that you ought to be assembling these routinely on your 3
own rather relying on my students.
4 COMMISSIONER MAGWOOD: Always curious, but people know I 5
sort of like education -- even though Im not a professor, I like -- but what kind of 6
students are they? Are they technical students?
7 DANIEL HIRSCH: Theyre many disciplines. Theyre interested 8
both on the policy side and the technical.
9 COMMISSIONER MAGWOOD: Okay, so its multidisciplinary.
10 Sounds interesting. Thank you. Thank you, Chairman.
11 CHAIRMAN MACFARLANE: Thank you. Commissioner 12 Ostendorff?
13 COMMISSIONER OSTENDORFF: Thank you, Chairman. Thank 14 you all for your presentations. I want to add my comments to those of my 15 colleagues, Commissioner Apostolakis and Svinicki, who indicated that though 16 we may not ask particular questions, it does not mean were not interested, 17 consistent with the Chairmans comments at the opening of the hearing today.
18 So I thank everybody here for your presentations.
19 Let me go to Mr. Benson here for a minute. You know, talking at a 20 high level, I asked the first staff panel comments or question concerning design 21 considerations for the life of a steam generator. And I wanted to ask you from 22 your EPRI standpoint, what are some high-level -- not focusing on any particular 23 design, but just across the industry, can you talk a little bit about design life 24 considerations and what goes into looking at a particular set of optimization 25
82 criteria for design life?
1 JIM BENSON: Yes. The work that weve done thus far in 2
specifications for steam generators was mostly focused on the tubing, and weve 3
done numerous studies on tube material and corrosion studies, and weve put 4
out documents on specifications for the manufacture of tubing and requirements 5
on the cleanliness, ovality, and Eddy current noise as well. So thats pretty much 6
the extent as far as a design standpoint on specifications. However, what we do 7
provide is information that we gather from the utilities on operating experience 8
and similar to some of the information Ive provided today. And that is shared 9
amongst the utilities and the vendors, and then from those experiences, they can 10 determine where the weaknesses are in their designs and strive to improve those 11 designs.
12 COMMISSIONER OSTENDORFF: Okay. I want to give our three 13 vendor representatives any opportunity if you want to add to Mr. Bensons 14 comments, if you had anything else you want to say on that.
15 DAMIAN TESTA: Yeah. Nothing in particular.
16 JEFF FLECK: If your question would evolve or revolves around the 17 design life of a replacement steam generator, 18 COMMISSIONER OSTENDORFF: Typically 20, 40, 60 years.
19 JEFF FLECK: Minimum 20. Ive seen them 40, Ive seen 60. Its 20 very plant-specific.
21 COMMISSIONER OSTENDORFF: Yeah.
22 JEFF FLECK: It depends on perhaps the life of the plant, for when 23 they reach that point in the economic basis when its more appropriate to replace 24 the unit than it is to limp along with the, you know, a unit that has stress corrosion 25
83 cracking or some of these other mechanisms that weve talked about. It --
1 theres a wide range, but usually, or typically 40 years, similar to the original 2
component, is the design life thats asked for in the specification.
3 COMMISSIONER OSTENDORFF: Dr. Kaguchi, anything you want 4
to add that?
5 HITOSHI KAGUCHI: To design, we need, anyway, design life.
6 The design life, influence of the design life on fatigue or something is linear to the 7
design. So 20 years, 40 years is just a linear double of the number.
8 COMMISSIONER OSTENDORFF: Okay.
9 HITOSHI KAGUCHI: So its not so tremendous. Design itself is the 10 same.
11 COMMISSIONER OSTENDORFF: Okay, thank you. Going back 12 to Mr. Benson here, in one of your slides, in your Slide 8, you showed -- I 13 understood from the causes of steam generator tube repair, if Im looking at the 14 color-coding correctly and I understood your presentation, the most common 15 cause of tube repair has been stress corrosion cracking.
16 JIM BENSON: Yes, thats the blue color.
17 COMMISSIONER OSTENDORFF: So Im looking -- you know, that 18 continues to be the case over recent years, and I think this -- our NRC staff panel 19 earlier has made a comment along the lines of, Perhaps theres variability in the 20 industry as to chemistry practices, and I was curious from an EPRI standpoint, 21 do you see -- is there a consensus on chemistry to lessen stress corrosion 22 cracking for a particular design of steam generator, or are there any different 23 viewpoints on that?
24 JIM BENSON: So from the standpoint of the chemistry, because 25
84 obviously theres also stress considerations of the tubing, but from the standpoint 1
of the chemistry, that is being considered in the guidance that was put out to the 2
industry based on the corrosion studies that have performed, numerous 3
corrosion studies on the various tubing alloys. So the cracking that were seeing 4
in this particular chart is from the 600 mill annealed -- just the few plants that are 5
left have a significant number of cracks that theyre observing, and the newly 6
identified, thermally treated 600 tubing in the cracks thats identified in that 7
material as well. So this cracking that were seeing now is from the -- not from 8
the 690 material, but the chemistry that weve developed to assist in minimizing 9
the cracking, slowing the rate of cracking, and not just the chemistry itself but 10 also processes through a chemical cleaning to keep the tubes clean and the 11 crevices cleaned out, and that also is technical documents that are provided to 12 the utilities on optimization of chemistry practices and cleaning practices.
13 COMMISSIONER OSTENDORFF: Do you have any feel as to 14 what extent the utilities are actually following that guidance?
15 JIM BENSON: Well, the guidance that we put out is for the U.S.
16 utilities. They are all following the guidance we provide. If there are any 17 exceptions to that guidance, because there are many, many requirements in our 18 guideline documents, then they would identify through a technical write-up that 19 they would provide to both the NRC and to EPRI, and there are a couple of 20 instances where the technical information has been provided, but its very rare to 21 have that deviate. So for all of the many requirements, theyre followed pretty 22 much to the letter.
23 COMMISSIONER OSTENDORFF: Okay. Thank you. I want to go 24 back to a question that Commissioner Magwood asked, and as I understood that 25
85 question, he was commenting on Professor Pettigrews work and to what extent 1
is that body of that work reflected in the design considerations for the vendors 2
designing steam generators. And I believe, Mr. Testa, you commented that you 3
use your own model. So can you comment briefly in a little more detail about 4
what model perhaps is used, the AP1000 design certification?
5 DAMIAN TESTA: The same models that are used in the 6
replacement steam generators. We refer back to the testing that was done 7
numerous years ago. Mr. Fleck even referred back to some testing that was 8
performed in France, and that work is still used today.
9 COMMISSIONER OSTENDORFF: And on the vendor side, is 10 there a consensus viewpoint on that modeling as to its applicability?
11 DAMIAN TESTA: We dont have any challenges with it. Its 12 worked well for us. Our track record is very good with it. I --
13 COMMISSIONER OSTENDORFF: Im not trying to challenge your 14
-- you know, Im just trying to see to what extent you guys are - weve got a 15 number of different vendors at the table here. Im trying to see, is there generally 16 a consensus, or are there very different approaches by one vendor approach to 17 the next?
18 DAMIAN TESTA: I dont think theres a whole lot of approaches in 19 this particular area. There are differences with respect to design and 20 manufacturing. We have a lot of differences that we dont share with each other 21 22 COMMISSIONER OSTENDORFF: Sure.
23 DAMIAN TESTA: -- obviously.
24 COMMISSIONER OSTENDORFF: Okay.
25
86 DAMIAN TESTA: But I think when it comes to FIV, I think our roots 1
are very similar.
2 COMMISSIONER OSTENDORFF: Okay. Do you want to add 3
anything, Mr. Fleck, or...
4 JEFF FLECK: No, I think that that adequately covers it.
5 COMMISSIONER OSTENDORFF: Okay. Ill give a last chance for 6
Professor Pettigrew. Is there anything you want to comment on and the -- from 7
your perspective as to how you see the academic-type work that youve been 8
doing being reflected in vendor applications?
9 MICHEL PETTIGREW: Well, usually what we do, whenever we 10 come up with something that is significant and can improve things, then we put 11 them in our specification for flow induced vibration; so there is a direct path here 12 between whats done in the lab and what goes into the specification.
13 COMMISSIONER OSTENDORFF: Okay. Thank you. Thank you, 14 Chairman.
15 CHAIRMAN MACFARLANE: Thank you very much. Okay. Got a 16 couple of questions here. Can you just go back to -- for Mr. Testas Slide 14, and 17 pick up on the question that Commissioner Apostolakis was asking when you say 18 in-plane turbulence and not from in-plane instability, and can you just explain to 19 me -- I must be a little thick -- what the difference is between turbulence and 20 instability?
21 DAMIAN TESTA: Ill try to do that. Turbulence would be similar to 22 a tube rattling around --
23 CHAIRMAN MACFARLANE: [affirmative]
24 DAMIAN TESTA: -- in a support, relatively closely supported. It 25
87 will move, but it will not severely deform, as in -- as it was described by Dr.
1 Pettigrew.
2 CHAIRMAN MACFARLANE: [affirmative]
3 DAMIAN TESTA: So structurally, it will not bend or deform 4
significantly. Its more or less just a rattle around in a tube, in a support case. I 5
guess thats one -- best way I could --
6 CHAIRMAN MACFARLANE: And whats instability? That was 7
[inaudible], right?
8 DAMIAN TESTA: Instability is a change in the mode shape.
9 CHAIRMAN MACFARLANE: Change in the...
10 MALE SPEAKER: In the mode shape of the structure of the tube.
11 CHAIRMAN MACFARLANE: So a change in the structure of the 12 tube. The tube itself is deforming as opposed to moving.
13 DAMIAN TESTA: And Ive deferred it - Dr. Pettigrew to see if --
14 CHAIRMAN MACFARLANE: And do you all agree to those 15 definitions?
16 MICHEL PETTIGREW: Well, they -- if I may say something here, 17 there are basically four vibration excitation mechanisms. One is called periodic 18 wake shedding, you know, where you may have a cylinder and flow, and it would 19 generate vortices, and these vortices can excite the structure to vibrate or excite 20 other structure to vibrate; so that is periodic wake shedding.
21 Second one is acoustic resonance. Its usually not a problem in 22 steam generators, but if you have a situation whereby the frequency of the 23 vortices that are being shed by the tube bundle coincide with the acoustic natural 24 frequency of the container, then youd have a resonance problem, and it can do a 25
88 lot of damage. But that doesnt happen in steam generators.
1 Then you have whats been called buffeting or rendering 2
turbulence. Theres been different terms for the same thing. This is coming from 3
the flow itself. Theres always turbulence in flow, and two-phase flow in 4
particular, its a very active flow, theres turbulence. This generates pressure 5
pulsation or -- pressure pulsation along the surface of the structure, and that 6
excites it to vibrate. But usually, this is a fairly mild mechanism, and were 7
worrying about this one in the longer term. If you want a steam generator to last 8
60 years, then you have to do that kind of analysis and do it well.
9 And the last one is fluid elastic instability. To have fluid elastic 10 instability, youve got to have a coupling between the motion of the structure and 11 the fluid. So if you imagine that you have a tube inside a tube bundle and theres 12 flow; if you displace the tube a little bit, then the hydraulic forces around that tube 13 are going to be affected. Theyre going to change. If you get a situation whereby 14 the displacement of the tube causes hydraulic forces to change -- so you may 15 have, you know, a change in position of the structure, which changes the flow 16 and increases the fluid forces on the structure, so you get a situation where you 17 have -- the more you have motion, the more you have hydraulic excitation. The 18 more you have hydraulic excitation, the more youve got motion. And it goes 19 unstable more and more until everything breaks down or is limited by age in 20 tubes or supports.
21 CHAIRMAN MACFARLANE: So when you see in-plane tube-to-22 tube wear, which of those four mechanisms is it?
23 MICHEL PETTIGREW: Well, okay, I divide this thing in two. First 24 of all, you go from the flow and then you calculate the response of the structure.
25
89 And once youve got the response of the structure, then you calculate the wear 1
that will take place the -- if the structure responds to that level. Okay? So there 2
are two different things here. There is vibration, and there is vibration damage.
3 CHAIRMAN MACFARLANE: [affirmative]
4 MICHEL PETTIGREW: Okay, so you do the vibration response 5
calculation first, and then you go through your wear analysis and see how much 6
damage will result.
7 DAMIAN TESTA: And the point I was trying to make in my Slide 14 8
was that these tubes were not assumed to be very much in close proximity to 9
one another, possibly even touching, which is not the design condition.
10 CHAIRMAN MACFARLANE: Okay. So then, just -- lets take this a 11 little bit further, and so Im trying to understand these processes, which actually 12 sound pretty complex. And Im trying to understand how well we know them and 13 whether we can understand whether they progress linearly or not.
14 MICHEL PETTIGREW: Okay. In time.
15 CHAIRMAN MACFARLANE: In time.
16 MICHEL PETTIGREW: In time, yes, yeah. Well, thats a very good 17 question. [laughs] Yeah, maybe I could comment on this a little bit?
18 CHAIRMAN MACFARLANE: Maybe somebody else has a 19 comment, too? Go ahead.
20 MICHEL PETTIGREW: Yeah. All right. The vibration part of it 21 probably doesnt change too much in time. Were talking about the same fluid 22 velocities and the same kind of structures, so I dont think we get a big change in 23 time. From the wear point of view, our wear studies have shown that wear for 24 the same excitation is the same, remained the same in time. So the wear 25
90 calculations are based on the parameter, which is called the work rate.
1 CHAIRMAN MACFARLANE: [affirmative]
2 MICHEL PETTIGREW: And this work rate, to put it very simply, is 3
the integral of the contact force times the sliding distance.
4 CHAIRMAN MACFARLANE: [affirmative]
5 MICHEL PETTIGREW: Okay? So the greater is your impact force 6
7 CHAIRMAN MACFARLANE: You dont take into account the 8
properties of the material?
9 MICHEL PETTIGREW: Yeah, you do that, yeah. But thats 10 another step. Okay. So this work rate, you can calculate it and estimate it. And 11 the wear damage, the wear volume damage according to our chart, the theory --
12 so the wear volume is really the product of a wear coefficient times a work rate.
13 Okay? The wear coefficient, this you have to obtain experimentally.
14 CHAIRMAN MACFARLANE: [affirmative]
15 MICHEL PETTIGREW: Its the wear properties of the material 16 combination that you have.
17 CHAIRMAN MACFARLANE: Are there -- is there experimental 18 data --
19 MICHEL PETTIGREW: Yeah.
20 CHAIRMAN MACFARLANE: Is there an experimental database?
21 MICHEL PETTIGREW: Yeah. There is a limited amount, not 22 greatly, because you have to do those tests at high temperature and high 23 pressure -
24 CHAIRMAN MACFARLANE: [affirmative]
25
91 MICHEL PETTIGREW: -- so it means in a pressure vessel. If you 1
do things in a pressure vessel, it gets very costly and very complicated, and to go 2
and try to measure inside those --
3 CHAIRMAN MACFARLANE: [affirmative]
4 MICHEL PETTIGREW: -- what you need, you need to measure the 5
contact forces and the sliding distance. Were talking about measuring microns 6
at 300 degrees C, measuring fraction of newtons at 300 degrees C. You need 7
something like $80,000 worth of transducer just for one such test facility. So --
8 but to answer your question, there is some data available. Some come from the 9
EPRI program, and some have come from Westinghouse, because they have 10 asked us to do that work for them.
11 CHAIRMAN MACFARLANE: And its been done -- its been 12 collected on the alloy 690?
13 MICHEL PETTIGREW: I cannot speak for what Westinghouse has 14 done with test data. Ours has been generally published, so its available to 15 everybody. Yeah.
16 PETE DIETRICH: Chairman, what we have concluded, getting to, 17 you know, how does fluid elastic instability propagate over time - the wear --
18 thats why its critical to prevent it, is because we do think that it starts with a set 19 of tubes and propagates to tubes around it and can grow if the phenomenon is 20 allowed to occur. And thats why, as we depicted in our presentation, not 21 allowing the three components to occur concurrently is very important. Its 22 critical. And by preventing fluid elastic instability, you prevent -- or by preventing 23 those three conditions from occurring concurrently, you prevent the onset of fluid 24 elastic instability.
25
92 CHAIRMAN MACFARLANE: Do you guys have any comments?
1 DAMIAN TESTA: No, I agree.
2 JEFF FLECK: Yeah, I agree, and Id also add that, you know, that 3
when we changed materials from 600 thermally treated to 690, and stainless 4
steel supports that, you know, the common practice would be to do wear testing 5
to understand what the coefficients are in order to, you know, perform some of 6
the wear analysis that Dr. Pettigrew referenced.
7 JIM BENSON: One point I wanted to make is that EPRI has done 8
some studies on wear coefficients of different materials, the -- a 600 material, the 9
690 support materials, stainless steels, carbon steels, and then made that 10 available to the industry in our published reports. So we do have those types of 11 studies that were done.
12 HITOSHI KAGUCHI: We also referred EPRI data and also some 13 material slightly changed from the -- for the -- especially AVB material is for 14 viable use, the use of 410. So we confirmed that the data is adequate or not.
15 We did our own test.
16 CHAIRMAN MACFARLANE: Okay. Thank you. Commissioner 17 Svinicki.
18 COMMISSIONER SVINICKI: I just want to add my thanks for all of 19 you traveling here today and for sharing your information and perspectives. I 20 appreciate also your responses to my colleagues questions, and I dont think I 21 have anything further. Thank you.
22 CHAIRMAN MACFARLANE: Okay. Any further questions from my 23 colleagues? No? Yes?
24 COMMISSIONER APOSTOLAKIS: I dont think we got a straight 25
93 answer whether research results like the ones that Dr. Pettigrew produces diffuse 1
into the design or -- you gentlemen said, Oh, we do our own. I mean, are you 2
feudal lords and isolated and dont care what the professor in Canada does?
3 DAMIAN TESTA: No, we care. We listen to what he says. He 4
publishes papers along with other academics. And our engineers read the 5
published papers out there, and its considered in the design.
6 JEFF FLECK: Yeah.
7 DAMIAN TESTA: Were just referring back to the data that we 8
have based on our own tests and staying within the comfort levels of certain 9
parameters.
10 COMMISSIONER APOSTOLAKIS: So youre aware of what the 11 literature is?
12 DAMIAN TESTA: Yes.
13 COMMISSIONER APOSTOLAKIS: To some extent.
14 DAMIAN TESTA: Yeah, yeah.
15 COMMISSIONER APOSTOLAKIS: Dont read it every day.
16 JEFF FLECK: I agree.
17 MICHEL PETTIGREW: Maybe I would -- in the light of your 18 question, maybe additional comments. We -- in Canada, we work through 19 a research chair that is called an Industrial Research Chair, meaning it 20 has industrial partners; so in our case, its Babcock and Wilcox Canada, 21 and Atomic Energy of Canada are the chair sort of supporter, if you like.
22 And this allows to create a research group and a lab with two-phase 23 facilities and so on, and theyre very active in the process of this chair.
24 We have meetings twice a year, and sometimes they direct the work that 25
94 needs to be done, and sometimes we tell them what they should be 1
interested in.
2 COMMISSIONER APOSTOLAKIS: Well, but you also publish in 3
international journals like the Journal of Pressure Vessel Technology and so on.
4 MICHEL PETTIGREW: Thats worked very well.
5 COMMISSIONER APOSTOLAKIS: Okay, this is better now.
6 Thank you very much.
7 CHAIRMAN MACFARLANE: Okay. Okay, great. Anybody else?
8 No? Okay. All right, well, I want to thank you all very much, both the external 9
panel and our NRC panel, for excellent presentations. This is a very technical 10 topic, and I commend everybody for their patience, and I think weve learned 11 quite a bit this afternoon. I really appreciate, again, all of you for coming out 12 here, and I will now say that we are adjourned.
13
[whereupon, the proceedings were concluded]
14
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION In the Matter of
)
)
SOUTHERN CALIFORNIA EDISON CO.
)
)
)
Docket Nos. 50-361-CAL (San Onofre Nuclear Generating Station - ) 50-362-CAL Units 2 and 3)
)
CERTIFICATE OF SERVICE I hereby certify that copies of the foregoing TRANSCRIPT of February 7, 2013 Briefing on Steam Generator Tube Degradation have been served upon the following persons by Electronic Information Exchange and by electronic mail as indicated by an asterisk*.
Office of Commission Appellate Adjudication U.S. Nuclear Regulatory Commission Washington, DC 20555-0001 E-mail: ocaamail@nrc.gov Atomic Safety and Licensing Board Panel U.S. Nuclear Regulatory Commission Mail Stop - T-3 F23 Washington, DC 20555-0001 E. Roy Hawkens Chief Administrative Judge E-mail: roy.hawkens@nrc.gov Anthony J. Baratta Administrative Judge Email: anthony.baratta@nrc.gov Gary S. Arnold Administrative Judge Email: gary.arnold@nrc.gov U.S. Nuclear Regulatory Commission Office of the Secretary of the Commission Mail Stop O-16C1 Washington, DC 20555-0001 Hearing Docket E-mail: hearingdocket@nrc.gov Southern California Edison Company Douglas Porter, Esq.*
Director and Managing Attorney Generation Policy and Resources Law Department 2244 Walnut Grove Ave., GO1, Q3B, 335C Rosemead, CA 91770 Email: douglas.porter@sce.com Counsel for Licensee Morgan, Lewis & Bockius, LLP 1111 Pennsylvania, Ave. N.W.
Washington, D.C. 20004 Paul M. Bessette, Esq.
Kathryn M. Sutton, Esq.
Stephen J. Burdick, Esq.
Steven P. Frantz, Esq.
William E. Baer, Jr.*
Mary Freeze, Legal Secretary Lena M. Long, Legal Secretary E-mail: pbessette@morganlewis.com sburdick@morganlewis.com ksutton@morganlewis.com wbaer@morganlewis.com sfrantz@morganlewis.com mfreeze@morganlewis.com llong@morganlewis.com
San Onofre Nuclear Generating Station, Units 2 and 3, Docket Nos. 50-361 and 50-362-CAL TRANSCRIPT of February 7, 2013 Briefing on Steam Generator Tube Degradation 2
U.S. Nuclear Regulatory Commission Office of the General Counsel Mail Stop - O-15 D21 Washington, DC 20555-0001 Edward Williamson, Esq.
David Roth, Esq.
Catherine Kanatas, Esq.
David Cylkowski, Esq.
Email: edward.williamson@nrc.gov david.roth@nrc.gov catherine.kanatas@nrc.gov david.cylkowski@nrc.gov OGG Mail Center: ogcmailcenter@nrc.gov Friends of the Earth Ayres Law Group 1707 L St., NW Suite 850 Washington, D.C. 20036 Richard E. Ayres, Esq.
Jessica L. Olson, Esq.
Kristin L. Hines, Esq.
Email: ayresr@ayreslawgroup.com olsonj@ayreslawgroup.com hinesk@ayreslawgroup.com Natural Resources Defense Council Geoffrey H. Fettus, Esq.
1152 15th Street, NW Suite 300 Washington, DC 20005 Email: gfettus@nrdc.org
[Original signed by Herald M. Speiser ]
Office of the Secretary of the Commission Dated at Rockville, Maryland this 19th day of February, 2013