This week, a redacted version of a root cause analysis performed by Mitsubishi Heavy Industries of the San Onfore Unit 2 and Unit 3 replacement steam generators was released publicly for the first time. The report reads more like an apology than an analysis.
Southern California Edison, the utility which operates the San Onofre nuclear power station, had a strict list of requirements for MHI during the design and manufacturing process, specifically created in order that the utility might justify not allowing the Nuclear Regulatory Commission to approve the designs and regulate the replacement process, by using the justifications allowed in 10 C.F.R 50.59 to claim that it was a “like for like” exchange. Most of these restrictions limited the external physical changes to a minimum by ordering MHI to design replacement steam generators which would fit within the same physical space, rather than inhibit multiple changes from being made to internal components. There is an age-old axiom about what kinds of beauty are only skin deep.
The report quickly shoots down this argument, and details a list of changes that were made which the utility felt still fit within the 50.59 process.
- SCE imposed physical and other constraints on RSG design characteristics in order to assure compliance with 50.59. (RSGs must fit within same space)
- Increased heat transfer surface area 11% from 105,000 ft2 to 116,100 ft2.
- Higher void fraction (maximum steam quality)
- Added more tubes
- Tube designs substituted Iconel 690 for Iconel 600. (Iconel 690 has a thermal conductivity approximately 10% less than that of Iconel 600.)
- Increased U-bend radius
- Added more AVBs
- Required more stringent tube-to-AVB gap requirement
- Distances between AVB tube supports are shorter
The San Onofre replacement steam generators had more tube vibration margin than comparison replacement steam generators at the Fort Calhoun Nuclear Generating Station, also designed by MHI, which only experienced a small number of tube wear occurrences. But, MHI and SCE decided to change the very alloy the tubes were made out of, for one which is more prone to fluid elastic stability, but was thought to be more resistant to corrosion. The tubes at San Onofre are longer, have thicker walls, and are stiffer, than those in the Fort Calhoun generators. The new replacement design also increased the heat transfer surface area, which increased the steam quality in the generators. On top of that, they elongated and thinned the retainer bars that are installed around the tubes, which caused an unexpected increase of amplitude vibration contact between tubes and bars causing even more wear.
Mitsubishi than revealed that in-plane instability like this had never been experienced in the nuclear industry prior to San Onofre, and that none of the predictive models were even capable of including it in their analysis. Even more, none of the 12 AVB supports were found to be restraining this tube motion. Because they did not even know about in-plane fluid elastic instability at time of design or manufacture, let alone preventing it, they never thought to question if with those changes, they would need more contact force to prevent vibration increases.
Mitsubishi records two main root causes, and a redacted list of contributing causes. One root cause and all contributing causes are associated with the “decision-making” and “resources” processes. The other root cause and two of the above mentioned contributing causes are associated with standard “work practices”. Essentially, MHI states that the root causes all relate back to a lack of resources, a decision-making process which wasn’t properly regulated or guided, and assumptions made based on common “work practices”.
It is akin to admitting; first off, we didn’t even know that these particular types of problems existed, we know the general problems which plague our decisions are based on the fact that we do this all the time, we are always limited by a lack of resources, we tend to assume it is no big deal, this is standard operating procedure; except for this time no one stopped us, because no one thought it could go wrong. Maybe if we hadn’t spent so much time trying to get around the rules, had we only required the analyses which would’ve evaluated a little more, if only we hadn’t assumed these were acceptable industry practices.
The questions which still need answering are focused if the utility hadn’t pre-determined it would use the 50.59 process to justify the design, would Mitsubishi have acted differently, and are all parties aware of the full implication and repercussions for their actions?
Mitsubishi admitted that there were other design changes which could’ve been made, but would’ve had “unacceptable outcomes”, and failed to list what these changes would’ve been. They did however take particular care to point out that other replacement steam generators which had been in operation for longer, did not have the same wear problems witnessed at San Onofre, which is worthy of note.
In answer to the second question is a resoundingly no. This appears to be a blatant attempt to manipulate the system to achieve an acceptable outcome and avoid unacceptable alternatives, which undermines the very faith in the system itself; for who could have faith in something which does not do what it claims to do, and has no ability to ensure that it does in the first place?
For most people, their perception of the law is based less on what they read, as much as they are by precedent and the way it is carried out before them. What actions can we expect, what hope in the future, when so distracted and led astray the children are, by the examples of the leaders of today?