|Movie poster for Chaplin's|
The NRC team also determined that Southern California Edison provided the NRC with all the information required under existing regulations about proposed design changes to its steam generators prior to replacing them in 2010 and 2011.
In effect the NRC said SCE did not mislead the agency about the designs, manufacturing, and installation changes to the steam generators by the manufacturer.
In an effort to identify the causes, the NRC said SCE brought in a large number of outside industry experts, consultants, and steam generator manufacturers, including Westinghouse and AREVA, to perform thermal-hydraulic and flow induced vibration modeling and analysis.
SCE identified “the most probable causes of the tube-to-tube wear as a combination of higher than predicted thermal/hydraulic conditions and changes in the manufacturing of the Unit 3 steam generators, a conclusion with which the NRC team agreed.”
“The changes in the manufacturing resulted in less contact forces between anti-vibration bars and the tubes. The combination of these causes allowed excessive vibration to occur.”
|Function of the steam generator in a nuclear power station|
Image: Japan Nuclear Power Training Center
Anti-nuclear groups unconvinced by NRC report
Anti-nuclear groups shifted their stance on the plant somewhat in response saying Southern California Edison is paying out millions in replacement fuel costs and rate payers throughout Los Angeles and San Diego may yet see brownouts or blackouts if an extended heat wave overwhelms the now underpowered grid.
Friends of the Earth (FOE) repeated its accusation that SCE deliberately misled the NRC about the changes to the steam generators including the number of tubes and changes to support structures. The groups has issued several reports about the steam generators based on engineering documents it says it obtained from the plant leaked by utility employees.
FOE said in its report that equipment design and manufacturing changes “created a large risk of tube failure” at SONGS.
Separately, Daniel Hirsch, a member of the faculty at University of California Santa Cruz, who is a long-time critic of the nuclear industry, told the Associated Press July 19 “the agency (NRC) is attempting to exonerate itself from blame” for the problems with the steam generators.
He said the NRC claimed there was no wrong doing by the utility or the regulatory agency despite the fact that the “faulty generators caused unprecedented tube damage (early wear) and raised the risk of a serious accident (release of radiation by tube leaks).”
|Cut away diagram of a Westinghouse steam generator|
The units at SONGS were manufactured by Mitsubishi.
For illustration purposes only.
In a conference call with nuclear bloggers July 20, Southern California Edison (SCE) spokesperson Jenifer Manfre said repairs on the steam generator for Unit 2 would likely be completed before Unit 3. However, she said there is no timeline for restart of the twin 1,100 MW reactors and she predicted they would be out of service for the rest of the summer.
Manfre said that Southern California Edison agrees with the NRC that faulty computer modeling carried out by Mitsubishi was inadequate.
It raises questions about computer model verification and whether the utility had a responsibility to verify Mitsubishi’s numbers before it gave the go ahead to manufacture the steam generator based on the model results.
The NRC said in its report (pg.51) in the restrained language of its engineering staff, “Because of the limited information provided, the team could not determine the validity of the benchmarking of FIT-III.”
“Overall, the team determined that the validation and verification of the FIT-III code did not present overwhelming evidence that this code has been adequately benchmarked.”
The shortfall in model verification was that "adequate validation and qualification was performed to show compliance to software requirements and that the code could predict flow-induced vibration."
What it apparently did not do is prove that the model produced accurate results. It is one thing to say that your model can simulate flow in a steam generator. It is entirely another thing to say with certainty that it simulates the flow with precision relative to the reality of the conditions inside a real operating unit.
What a difference a model makes
Another of the challenges faced by SCE at SONGS is that the utility must understand why there are differences in tube wear inside the steam generators installed at Unit 2 versus those at Unit 3. SCE and several other nuclear engineering firms helping the utility understand the problems believe the causes are related to changes from Unit 2 to Unit 3 in design and in manufacturing of the support structures and the tubes themselves.
The critical piece is the computer model used by Mitsubishi underestimated steam flow and other variables, such as vibration wear caused by tube movement and the adequacy of support structures, by a factor of two-to-three times according to the NRC report.
The NRC calculated steam flowed through a tube at 5.2m/sec. By comparison, the Mitsubishi model indicated a flow in the same tube at 2.5 m/sec. It is a major error.
For its part Mitsubishi has not issued a statement on the NRC report and has not made itself available to the media. None of the wire service coverage, nor from major newspapers, include a statement from the company.
New lamps for old
Manfre said that three other suppliers of steam generators are working with SCE to develop a new model. The firms include Areva, B&W, and Westinghouse. The new model will help SCE understand what is really happening inside the steam generators and also to specify cost effective repairs.
The new model will be run with different variables including steam velocity, temperature, amount of moisture in the steam, and power level of the reactor. For instance, a lower power level of the reactor results in slower velocity of steam and wetter steam.
Buried deep in the NRC report (pg 51) is this conclusion.
“Both the NRC and Mitsubishi ATHOS results were reasonably consistent and strongly suggested that high velocities coupled with high void fraction were primary causal factors in the tube fluid-elastic instability and the excessive wear patterns observed in the Unit 3 steam generators.”
Plugging tubes for wear
The threshold for tube wear that requires a tube to be plugged is 35% Manfree said. It is this level of wear with only a few years of operation in equipment expected to last two decades that has precipitated the crisis at SONGS. The risk of tube leaks and releases of radioactivity into the open air is what caused plant operators to shut down Unit 3 last January. Unit 2 was already shut down for maintenance.
Manfre shared statistics on tube plugged so far which is the most effective way to prevent leaks in tubes that have more than 35% wear.
In Unit 2 just six tubes have more that 35% wear, but in Unit 3 SCE found 381 tubes with more than 35% wear. SCE’s engineer plugged these tubes.
In Unit 2 an additional 504 tubes have been plugged, based on wear problems, as a preventative measure. In Unit 3 another 406 tubes have been plugged as a preventative measure.
Statistically speaking, the number of tubes affected is relatively small for both steam generators. Each steam generator has over 9,700 tubes.
Requirements for reactor restart
In terms of restarting the reactors, SCE must prove to the NRC that it understands the safety significance of the tube wear problem, what can be, and has been, done to address the issue, and that the reactors can be operated safely. The NRC in its report listed 10 issues that SCE must address.
and which require additional follow-up:
(1) further review of the adequacy of the plant’s post trip/transient procedure;
(2) review of the adequacy of acoustical alarms used to identify loose parts in steam generators;
(3) evaluation of steam generator retainer bar design for vibration impacts;
(4) evaluation of and control of the Unit 3 divider plate repair;
(5) Unit 3 steam generator shipping requirements;
(6) lack of tube bundle support for steam generators during shipment;
(7) evaluation and disposition of accelerometer data used to measure unusual movement of steam generator shipping packages;
(8) review of the process used by the NRC to approve the plant’s steam generator replacement;
(9) control of manufacturing differences; and
(10) adequacy of Mitsubishi Heavy Industries’ computer simulation modeling.
SCE’s formal response to the NRC report
In a press statement issued July 19, SC said that safety not schedule would drive the process to restart the reactors.
"We are committed to continuing to work with the NRC on the steam generator issues and will continue to use conservative decision making as we work on repairs and planning for the future," said Senior Vice President and Chief Nuclear Officer Pete Dietrich. "The number one priority is the safety of the public and our employees."
“Additionally, the commission report states that the steam generator design changes were appropriately reviewed in accordance with the 10 CFR 50.59 requirements, which govern design changes between original and replacement steam generators, noting that the changes at SONGS are common in replacement steam generators today."
"The report also stated that, with the exception of the wear found at the tube retainer bar locations, the wear related to support structures is similar to wear found at other replacement steam generators after one cycle of operation.”
SCE said it continues to work on its response to the commission’s Confirmatory Action Letter, the formal step in the process of restarting Units 2 and 3. The company will not restart the units until the utility and the commission are satisfied it is safe to do so.
- SCE Background fact sheet on steam generators
- SCE Fact sheet on NRC Confirmatory Action Letter (CAL)
The 2,150 MW plant is owned by Southern California Edison with a 78% equity stake. San Diego Gas & Electric owns another 20% equity stake. The utility serves customers throughout southern California including the cities of Los Angeles and San Diego.
On an average day SCE provides power to nearly 14 million people in 180 cities in 50,000 square miles of service area, encompassing 11 counties in central, coastal and Southern California.
Commercial industrial and nonprofit customers include 5,000 large businesses and 280,000 small businesses
To deliver that power, it takes:
· 16 utility interconnections
· 4,990 transmission and distribution circuits
· 425 transmission and distribution crews
· More than 15,500 employees with 24 x 7 coverage
Postscript - Who will pay?
I asked SONGS spokesperson Jenifer Manfre about accountability for the costs of the problems with the steam generators, but she declined to address the issue. As a practical matter, I see the following moving pieces. In terms of financial issues for the parent firm, Edison International NYSE:EIX here are some questions.
- Fuel replacement costs for the duration of the outage of the two reactors
- Regulatory inspection and repair costs for the steam generators
- Whether insurance and to what degree it will cover fuel replacement costs or the repairs
- Financial liability per contract with Mitsubishi said to be capped at $160 million
- Whether the California pubic service commission will accept and approve a rate case to recover costs not addressed by insurance
- Lost revenue from future operations if the reactors have to run at a lower power level
And insider buyers exceeded insider sales for the last quarter and for the last 12 months. The stock is trading at $46.01 or less than $1.00 against its 52 week high: L:$32.64 H:$46.81
In short, stockholders of Edison International do not appear to have been hurt by the extended outage at SONGS or anticipation the utility will be hit with unacceptable costs.
# # #