Failures in massive Japanese nuclear components

  1. You have chosen to ignore posts from AppDev. Show AppDev's posts

    Failures in massive Japanese nuclear components

    The $680 million project to retrofit the two active reactors at the San Onofre nuclear power-plant near San Diego with new steam generators has run into a major problem: new tubing expected to last 40 years or more has failed after less than two years. The four replacement steam generators, each weighing 640 tons, were designed and built in Japan by Mitsubishi Heavy Industries, calling into question the ability of the Japanese nuclear industry to service a type of reactor favored in  the U.S. but disfavored in Japan. [ Michael R. Blood, Associated Press, California nuclear power-plant to remain shut for testing, WTOP (Washington, DC), March 28, 2012, at ]

    San Onofre units 2 and 3 pressurized-water reactors, commissioned in 1983 and 1984, are among the largest and most advanced second-generation nuclear power-plants. However, they were designed just prior to the discovery of accelerated corrosion in steam generator tubing made with high-performance Inconel 600 alloy. As with most such reactors designed later, Mitsubishi used Inconel 690 tubing in its replacement components. Ordinarily, that should improve resistance to both corrosion and metal fatigue at operating temperatures.

    Vague reports so far on recent problems indicate wear-related failures induced by vibration. Mitsubishi did not produce equivalents to the original steam generators. In addition to changing materials, it omitted a stiffening cylinder, added more heat-exchange tubes, redesigned tubing supports and altered flow rates of pressurized water in the reactors' primary loops. The U.S. NRC allowed those changes without full technical reviews. [ Fairewinds Associates, Steam generator failures at San Onofre, March, 2012, available at ]

    In April, 2010, NRC reported cracks detected in the replacement steam generators before they were installed, indicating defective workmanship. However, the cracks reported then were not in tubing but in other parts of the steam generators. [ Welding defects in replacement steam generators, U.S. NRC, April 5, 2010, at ]

  2. You have chosen to ignore posts from AppDev. Show AppDev's posts

    Project loaded with unnecessary risks

    Southern California Edison, operator of the San Onofre nuclear power-plant, greatly increased risks in its project to replace steam generators, by allowing Mitsubishi Heavy Industries to develop many new design elements. Although they were all intended to improve performance, every such new element lacks operating experience to assure its long-term reliability. [ Boguslaw Olech and Tomoyuki Inoue, Improving like-for-like replacement steam generators, Nuclear Engineering International 57(687):36-39, 2012, at ]

    The result was not really like-for-like replacement--needed to escape full technical reviews, as described in NRC regulation 10 CFR 50.59--but instead a new and significantly unproven design, meeting only overall parameters. Flow-induced vibration, now a key issue at San Onofre, is a well known hazard for steam generators. [ Assessment and management of ageing of major nuclear power-plant components: steam generators, International Atomic Energy Agency, 1997, at ]

    The many changes made by Mitsubishi Heavy Industries in tube layout and support structures left the new San Onofre design at risk of unknown vibration modes. Descriptions of the design efforts by project engineers do not mention analysis of vibration modes. However, vibration-induced tube wear has been severe, with more damage to some heat-exchange tubes in two years or less than was expected in 40 years. [ Unattributed, Worn tubes in San Onofre 2 reactor, Reuters Market News, March 29, 2012, at ]

    Now Southern California Edison is already incurring costs from unplanned outages that could run as high as $2 million a day, and there could be large, unplanned expenses ahead from having to rework parts of the replacement steam generators or having to accept operating lifetimes far shorter than expected.

  3. You have chosen to ignore posts from topaz978. Show topaz978's posts

    Re: Failures in massive Japanese nuclear components

    Well these are the responsibility of the contracting company. While many things are the responsibility of the NRC this is not clearly so. Many designs may fit the needs but the NRC is not the lead agency. The operator is. If they got snookered that is too bad. The issue is that stainless steel design has not been properly looked at for all components. Many but not all. Vibration has many resolutions other than material changes. Circulation rate, pump design, and tube diameter are just a few. There are a whole host of vibration solutions includeing a cheap resonant chamber added to allow dampening. It does have to be designed for the frequency range but it will damp the issue at a low cost.
  4. You have chosen to ignore posts from AppDev. Show AppDev's posts

    Clever Hans makes a mess

    The previous reader could have made better sense of the situation by reading references provided. They show that Southern California Edison (SCE) represented the project to NRC so as to evade technical reviews. One of the references clearly shows SCE materially involved in efforts changing the project from straightforward replacement of a component to an almost entirely new design.

    Ironically, engineers from SCE and Mitsubishi were patting themselves on the backs for that coup in print, just as the $680 million project was coming unglued. Some way will probably be found to control vibration, but now it looks to be too late. Some of the heat exchange tubes have been wrecked, and many others have been worn so badly that their useful lifetimes are much reduced.

    The public stands to suffer from higher electricity rates, if local generation capacity is impaired for long. Residents of southern California will easily remember their electricity supply crisis of 2001.

  5. You have chosen to ignore posts from AppDev. Show AppDev's posts

    Southern California at risk of blackouts

    Gross failures in all four of the giant new, Japanese nuclear components at the San Onofre power-plant in southern California have left the region at high risk of electricity blackouts. [ Michael R. Blood, Associated Press, Blackouts likely with idled California power-plant, WTOP (Washington, DC), April 5, 2012, at ]

    Backup power in the region is minimal because of retirements of AES plants in Redondo Beach and Huntington Beach, purchased from Southern California Edison in 1998. AES plans to repower these inefficient plants, coal-fired and single-cycle natural gas fired respectively, with combined-cycle natural gas but is two or more years away from having repowered plants online.

  6. You have chosen to ignore posts from AppDev. Show AppDev's posts

    A gamble goes bad

    As the number of damaged heat-exchange tubes reported in replacement steam generators at Southern California Edison's San Onofre nuclear power-plant continues to rise, so does speculation about causes of the early failures. Tubing in the original steam generators lasted almost 30 years, but in the replacements it began to fail in less than two years.

    Engineers from Edison and Mitsubishi, the Japanese company that designed and manufactured the replacement units, have described ways the replacement design differed from the original. Some had potential to affect vibrations that have damaged tubing: about four percent more tubes than the original design, installed in place of a so-called "stay cylinder" brace, and a thicker "tubesheet" into which tubing is welded, making it stronger and stiffer to compensate for the "stay cylinder." [ Boguslaw Olech and Tomoyuki Inoue, Improving like-for-like replacement steam generators, Nuclear Engineering International 57(687):36-39, 2012, at ]

    While adding more tubes and thickening the "tubesheet" may have been adequate for the weights and operating pressures, such changes can also affect vibration modes and intensities--much like adding or removing material on the sounding boards of a violin. With such a complex assembly, vibrations can be unusually challenging to understand. So far, the engineers have not described what analysis they did, if any.

    Arnie Gunderson, a former nuclear engineer and operator who now acts as a consultant to advocacy groups, recently released a second report calling attention to these design changes. He comes across a bit strong about weights and pressures, when it is highly likely that engineers would have followed well established safety codes for those issues, but he has little to say about vibration modes or intensities. [ San Onofre cascading steam generator failures, Fairewinds Associates (Burlington, VT), April 12, 2012, available at ]

    Why would engineers take such a risk, making changes whose effects can be hard to understand? The added tubing might be used to justify a small increase in power, but it is hard to see how the economics of that factor alone are great enough to justify the risks. Another reason for wanting the "stay cylinder" out is that it tends to get in the way of the tube inspections that must be performed at every fuel change.

  7. You have chosen to ignore posts from topaz978. Show topaz978's posts

    Re: Failures in massive Japanese nuclear components

    The issue is that the changes in fact did not compromise the "normal" level of operations. They made sense. Only that the changes might have effected higher levels of operation. As each of these assemblies is a unique custom build (this is not a ford fiesta) critical initial assembly inspection plays a huge role. Larger than the engineering issues normally.
     The design changes often do not require a large suite of harmonic testing as very similar structures passed similar tests with the same result. Normally such a small difference would not lead to a large new design test suite. Further the design issues are particular to the plant. Maybe the original install was an 1/8 note short of a harmonic but the contractor needed a new valve location and pushed the unit into the sweet spot while trying to fit a required earthquake valve?
     The violin metafor is "cute" but stupid. Get real. These are not children building mud pies. They model harmonics before build on super computers. You are useing serious backyard engineers as an argument as opposed to poor quality inspection? No engineer who worked on the project can say anything about it. That is the contract and is no big surprize particularly if there was a build fault. It is not their problem.
  8. You have chosen to ignore posts from AppDev. Show AppDev's posts

    No strong musical analogy

    The musical analogy of a violin won't take us far, but design changes made by Mitsubishi to the original Combustion Engineering steam generators at Southern California Edison's San Onofre plant were vibrationally similar to removing a violin's soundpost. That would turn a Stradivarius into a boom box.

    While removing a "stay cylinder," adding more heat-exchange tubes over its cross-section and thickening the "tubesheet" to compensate might well achieve the same performance with respect to weights and pressures, it can produce substantial changes in vibration modes that are very challenging to calculate accurately. As someone with a physics background would know, supercomputers are of little help, since the problem scales factorially. So far, no analysis of these effects has been disclosed to the public, if there was one.
  9. You have chosen to ignore posts from topaz978. Show topaz978's posts

    Re: Failures in massive Japanese nuclear components

    You are incorrect as to the engineering. The modeling does not even take a supercomputer. We currently have desktop software able to do this. Takes a couple of days, but that is not the issue. The vibrational nodes can be internally caused or external. That is the problem for any refit. The design spec of an upgrade refit that increases plant capacity is an untested combination.
     Each plant is built to spec by the owner. The engineers assume several issues on an upgrade. These also are not tested in the real world. The problem is it is a refit. The design engineers for the refit had to ask questions that were not clear to refine their models. It is not hard to do. But asking the right questions and changing your assupmtions is hard. It often leads to failure. Thats why there are so many bad products in the world. It comes from an assumption of needs that fails the mark.
     North Korea had a similar level of fail in its rocket test. MaxQ deterioration of the rocket. Other engineers have seen this type of failure as well.
  10. You have chosen to ignore posts from AppDev. Show AppDev's posts

    Hazards from steam generators

    The nuclear power industry is alert to hazards from steam generator failures, partly because of an incident about 20 years ago in Japan. A heat-exchange tube broke apart in a steam generator serving the Mihama unit 2 reactor, operated by Kansai Electric. It is a 0.47 GWe, 2-loop pressurized-water reactor built by Mitsubishi and Westinghouse, commissioned in 1972. At the time, that was the worst nuclear reactor incident that had occured in Japan. [ Karen Lowry Miller and Michael Schroeder, Is nuclear power losing steam in Tokyo?, Business Week, March 18, 1991, at ]

    Although the reactor automatically entered rapid shutdown using its emergency core cooling system, within a short time at least 13,000 gallons of water from the reactor vessel, carrying radioactivity, had surged into vessels and piping that feed generator turbines, partly as steam. About 0.6 Curies of radioactive gases reached the atmosphere, mostly xenon-135. Failures of automated valves could have led to a much worse result, but plant personnel were able to intervene manually. [ Yoshio Kitsunai and Hideo Kobayashi, Leakage of primary coolant at Mihama unit 2 due to failure of a steam generator tube, Failure Knowledge Database, Hatamura Institute (Japan, in English), 2003, at ]

    The Mihama unit 2 incident in 1991 was traced to excessive tube wear, resulting from heat-exchange tubes impacting and rubbing against braces. A major cause of the failure was poorly configured "anti-vibration bars," intended to inhibit out-of-plane vibrations in the U-portions of the tubes. The reactor remained out-of-service for many months to replace the entire steam generator.

    Just recently--so far unnoticed by news organizations and reporters--Mitsubishi described to the U.S. NRC the nature of premature failures in heat-exchange tubes of recently installed replacement steam generators for units 2 and 3 at the San Onofre nuclear power-plant, operated by Southern California Edison. [ Event no. 47833, U.S. Nuclear Regulatory Commission, April 13, 2012, available in a daily digest of reports at ]

    According to Mitsubishi, San Onofre tube failures are concentrated at points where U-portions of tubes are restrained by "anti-vibration bars." Metal fatigue and fretting that erodes metal when tubing impacts or rubs against braces has been a concern for nuclear power-plant steam generators since the 1960s. Once again--as with the 1991 incident involving the same Japanese company's equipment--poor configuration of "anti-vibration bars" has been identified as a potential contributor to failures of steam-generator tubing.

    The Mitsubishi report to NRC also discloses that the replacement steam generators have design margins for tube-plugging of only 778 tubes--about 8 percent--when degradation is found during surveys. Mitsubishi's design changes from the original steam generators at San Onofre, which were built by Combustion Engineering, added 377 tubes to each unit, in areas that had been occupied by large "stay cylinders" in the originals. [ Abby Sewell, Odd deterioration found in both San Onofre reactors, Los Angeles Times, April 13, 2012, at ]

    Without added heat-transfer tubes, the replacement steam generators would have had even smaller margins for tube-plugging--much less than the 20 to 25 percent of typical designs, including the original Combustion Engineering units. It is not clear from information available to the public why margins of the replacement steam generators should be so small. Dimensional limits are mentioned as constraints by design engineers but not explained. The slightly lower thermal conductivity of replacement Inconel 690 vs. original Inconel 600 tubing might be a factor. However, a need to increase tube-plugging margins can help explain why Mitsubishi and Edison opted for a risky approach, omitting the large "stay cylinders" of the original steam generators, adding heat-exchange tubes in their place and changing fluid flow patterns.

  11. You have chosen to ignore posts from topaz978. Show topaz978's posts

    Re: Failures in massive Japanese nuclear components

    It is not "risky" it is in fact likely to cause some person to get fired. The degradation is easily checked as it has been and is simple to assign blame as you have done. So what. Bad parts lead to new contracts and different designs. It is just a peice of gear. The original vendor gets a stiff fine and a serious black note. The issue is that the design changes you are talking about  are a problem inherent in the approval system. Design constraints are built into the system in order to take years for approval for novel design. Small tweaks are allowed without substancial testing, until those prove bad. The risk is not an issue except for the facility buying the component. The fact that so few of the reactors operating around the world have this issue points more to specific actual builders than design faults.
  12. You have chosen to ignore posts from AppDev. Show AppDev's posts

    Small changes with large effects

    From the 1960s onward, there were two distinct approaches to steam generators for pressurized water reactors made in the U.S. Those from Babcock & Wilcox have straight, vertical tubes between upper and lower plenums, while those from Westinghouse and Combustion Engineering have U-shaped tubes between two chambers of a lower plenum. The Babcock & Wilcox design is more resistant to vibrations but more expensive to manufacture, requiring a pair of thick bulkheads known as "tubesheets," instead of only one, plus twice as many joints between tubes and "tubesheets." Babcock & Wilcox built only 7 U.S. reactors with its design. The other 65 "second-generation" pressurized water reactors in the U.S. all used U-tube steam generators, as do most such reactors worldwide. [ Kenneth Chuck Wade, Steam generator degradation and its impact on continued operation of pressurized water reactors in the United States, Electric Power Monthly, August, 1995:ix-xxi, U.S. Energy Information Administration, available at ]

    The main approach to controlling vibrations in steam generators has been bracing. So-called "support plates" run crosswise to tubes and restrain them along straight lengths. So-called "anti-vibration bars" run between layers of tubing in the U-portions, to block bending out of the plane of the U: the most flexible mode of distortion with the largest potential vibrations. A so-called "stay cylinder" in the Combustion Engineering design stiffens and supports its "tubesheet," probably inhibiting some lower-frequency axial vibrations. Tube vibrations involving bending of U-portions within the plane of the U, rippling within both straight- and U-portions of tubes, and impact and rubbing against braces remain the main concerns for tube wear.

    Replacement steam generators designed and built by Mitsubishi in Japan for the San Onofre nuclear plant operated by Southern California Edison have been described as involving several design changes from the original Combustion Engineering units. Although dimensioned drawings have not been made available to the public, the number of tubes added by Mitsubishi in an area that would have been occupied by a "stay cylinder" indicates that its outer diameter was around about two feet--large enough for substantial effects on flow patterns, with the "stay cylinder" replaced by heat-exchange tubes. Changes made in feedwater distributors and in flow rates may also affect vibration excitation.

    For assemblies with thousands of U-tubes, exact calculations of vibrations are far too complex, even with supercomputers. Software amateurs, Web-oriented workers and members of the general public sometimes assume that just because there is software to analyze something, it must produce credible results. Developers of professional software are well aware of difficult issues, including practical situations with intrinsically chaotic behaviors or with complexity that radically exceeds capabilities of current or foreseeable computers to calculate solutions with well-bounded errors. For such situations, software uses approximations, and interpreting results involves art, judgement, unknowns and potentials for major mistakes.

    Approximate algorithms for steam generator vibrations have been refined since the 1960s. A few of the key references follow:
    [ Y.N. Chen, Flow-induced vibration and noise in tube-bank heat exchangers due to von Karman streets, ASME Journal of Industrial Engineering 90:134-146, 1968 ]
    [ M.J. Pettigrew, Y. Sylvestre and A.O. Campagna, Vibration analysis of heat exchanger and steam generator designs, Nuclear Engineering and Design 48(1):97-115, 1978, available at ]
    [ Ki-wahn Ryu, Chi-yong Park and Huinam Rhee, Effects of support structure changes on flow-induced vibration characteristics of steam generator tubes, Nuclear Engineering and Technology 42(1):97-108, 2010, at ]

    Ryu, Park and Rhee recently showed how even small configuration changes in tube bracing can result in large effects on tube wear. It is also known that U-tube assemblies can develop chaotic, non-harmonic vibrations. [H.J. Connors, Flow-induced vibration and wear of steam generator tubes, Nuclear Technology 55:311-331, 1982] Mechanical engineering researchers have reported that current vibration safety codes do not accurately estimate some natural vibration modes. [ Minshan Liu, Tong Liu and Qiwu D'ong, Study on factors influencing dynamic properties of steam generator U-tubes, Journal of Pressure Equipment and Systems 6(1):92-96, 2008, available at ] [ASME Boiler and Pressure Vessel Code, Section III, Division 1, (Nuclear) Appendix N-1300, Flow-induced vibration of tubes and tube banks]

    At San Onofre, there probably will have been changes in design details not mentioned in the descriptions available to the public. An intensive review of the Mitsubishi designs, almost surely underway, could find some apparently minor changes from the Combustion Engineering designs that produce tube wear at many times the rates once expected.

  13. You have chosen to ignore posts from AppDev. Show AppDev's posts

    Preparations for a long outage

    In signs that problems at San Onofre may be very hard to solve, Southern California Edison has been making preparations for a long outage. The latest is an agreement with the U.S. Navy to cut summer power demands. [ Michael R. Blood, Associated Press, Navy nears power deal to help avoid California blackouts, WTOP (Washington, DC), April 30, 2012, at ]

    San Onofre represents nearly a third of the San Diego-area generating capacity. The area is at the tail of power transmissions from the Washington and Oregon hydroelectric generators and the coal-fired power-plants at Four Corners and elsewhere in the Southwest. There is no practical way to import enough reserve power to make up the area's shortfall for next summer's maximum potential demand.

    Others have argued that premature failures of tubing in the newly installed, replacement steam generators, designed and built by Mitsubishi in Japan, represent avoidable errors by the manufacturer. Possibly so, but possibly not. Like, for example, aerodynamic drag produced by aircraft fillets, fairings and winglets--the vibrations in U-tube steam generators are long known but still challenging technical problems.

    Since its mid-April release of information--still unnoticed by U.S. news organizations and reporters--NRC has provided no further data on failure rates and no information on causes at San Onofre. [ Event no. 47833, U.S. Nuclear Regulatory Commission, April 13, 2012, available in a daily digest of reports at ]

    Given the unusual circumstances and severity of the problems, NRC is likely to be interested in a full survey, not just partial measurement, of the more than 37 thousand tubes in the four steam generators. Depending on the amount of instrumentation available, that could take half a year or more. It could also take that long or longer to determine causes and to design and implement remedies for the problems.

  14. You have chosen to ignore posts from AppDev. Show AppDev's posts

    Pop-fly caught in mid-air

    Anxious Southern California Edison managers of the San Onofre nuclear power-plant have already let slip that the outage from damaged steam generators may cost over $100 million, and they are not yet figuring the cost of replacement power during summer months of high demand. Last week they lofted a pop-fly into shallow left field: maybe they can get the plant going again at reduced output--maybe 50 to 80 percent, let's say. [ Mark Chediak, San Onofre nuclear plant may be online by June, SC Edison says, Bloomberg News, May 3, 2012, at ]

    The ball was promptly intercepted by the NRC shortstop: chairman Jaczko. "No dice," according to the chairman, saying, "We will take whatever time is necessary to conduct a thorough safety review." [ Michael R. Blood, Associated Press, NRC chair says there is no timetable to restart California nuclear plant, ABC News, May 8, 2012, at ]

    Reading between the lines, it's fairly obvious that neither the engineers at Mitsubishi, who designed and built the replacement steam generators in Japan, nor those at Southern California Edison, who approved many departures from the design of the original units, understand the causes of early failures that blew holes in heat-exchange tubing. If they did, they would be saying more than simply counting tubes known so far either to have failed or to be near failure.

  15. You have chosen to ignore posts from AppDev. Show AppDev's posts

    NRC says failures caused by bad design

    Anticipated by observers, including this thread, the replacement steam generators at the Southern California Edison nuclear power-plant in San Onofre were likely placed at risk of failure by bad design rather than by bad materials or workmanship, as the regional NRC administration recently confirmed. [ Michael R. Blood, Associated Press, Federal regulators say design led to nuclear plant problems, Boston Globe, June 19, 2012, at ]

    So what to do about that? To start with, in its typical evasive ways NRC is looking into fining SCE, for misrepresenting the new steam generators as "in kind" replacements when their design actually had significant differences from that of the originals. Such a penalty would certainly be challenged and seems unlikely to fly, since NRC staff previously let slip they had been told about design changes. More likely, SCE will be allowed to restart and test at least one reactor and will be required to derate the generating capacities, if it can show that vibrations subside at lower output.

    The AP reporter does what he can wth the story about Mitsubishi steam generators but obviously doesn't know how surprisingly difficult a problem is posed by analysis of vibrations in large units. It was "botched," he says. Yes, indeed, it was. So was the analysis of gusset plates for the I-35W highway bridge in Minneapolis, by the world-class firm of Sverdrup & Parcel, which knew how do do the latter job but failed to do it correctly.

    The likely conundrum is for vibrations to be reduced but not eliminated. How much less is enough? With such a question, NRC would be stuck. The agency's long-term failure to invest very much in independent research on nuclear safety leaves it unprepared in many areas to issue regulations based on reliable knowledge rather than on guesswork.

  16. You have chosen to ignore posts from AppDev. Show AppDev's posts

    Cutback in power-plant workforce

    Southern California Edison has announced that it is cutting back staff by about one-third at the San Onofre nuclear power-plant near San Diego. Both reactors there have been shut down since early this year because of failures in replacement steam generators, built by Mitsubishi in Japan. According to the LA Times, some monitoring organizations read the cutback to mean that reactor unit 3, which has the worst damage, is going to be mothballed. [ Abby Sewell, Edison to lay off 730 workers at San Onofre, Los Angeles Times, August 21, 2012, at,0,119444.story ]

    San Onofre currently has a workforce of over 2,200--more than most nuclear power-plants. The only plant of similar size in New England--Millstone in Waterford, CT--has half as many workers running a plant that also has two operating pressurized-water reactors in the same range of power outputs. It looks possible that Southern California Edison is taking advantage of the extended outage to adjust its staffing and continues to anticipate restarting both reactors. [ John Sheehan, Will Millstone power station shut down?, Waterford Patch (CT), April 9, 2011, at ]

    With the highest power demands of the year winding down, southern California has managed to scrape by with the use of an older power-plant that was being decommissioned and has had no unusual outages. The success in managing to do without San Onofre means that the analysis of what went wrong with the steam generators and how to correct the problems is not so likely to be compromised by pressures as some had feared.