Saturday, January 14, 2017

Moderation Inmodica et Neclegentia Moderationis Eaedem Sunt

Amici, Americani, Compatriotae,

The last two days we discussed the constriction of excessive regulation and how that stifles innovation and growth in the nuclear power industry, preventing the replacement of fossil fuel energy with a clean, non-polluting and a far safer form of energy production. But there is a flip side to this. While excessive regulation never serves public health and safety, equally so nuclear companies playing fast and loose with current regulation doesn't either, hence the title of this blog post: "Excessive Regulation and Heedlessness of Regulation Are the Same." Three examples will be given:

  • The Areva Priorization Module for Digital Instrumentation and Controls in the Evolutionary Power Reactor (EPR)
  • The Mitsubishi Software Engineering Plans for Digital Instrumentation and Controls in the Advanced Pressurized Power Reactor
  • The GE-Hitachi Vibration Calculations for the Steam Dryers in the Economically Simplified Boiling Water Reactor.

EXAMPLE 1

Areva, a French company, was planning to build its new EPR - an advanced pressurized water reactor with active safety systems - in the United States. It had planned to use an electronic module in its instrumentation and controls to prioritize signals form safety-related systems over those from non-safety-related systems. This module used Programmable Logic Device (PLD) technology which in turn employs Boolean logic programmed onto an integrated circuit chip using Hardware Description Language (HDL) coding. The US Nuclear Regulatory Commission (NRC) regards such technology as software simply because of the HDL code component. The NRC dislikes the use of software to prioritize safety over non-safety signals. The reason why is because of the checkered history that exists in various industries, not just nuclear, for using software to block non-safety and make safety processing a priority. While the technology is completely different, and likewise the example, back in the late 1980s Atomic Energy of Canada Limited (AECL) produced a machine called Therac-25 for cancer radiation therapy. Certain hardware interlocks were replaced with software ones. These failed to operate as intended. Patients were overdosed with radiation and died. You can read more here: An Investigation of the Therac-25 Accidents. There are more examples in the medical, aircraft, rail road, petrochemical, and nuclear energy industries. But in all these the lesson is clear: don't rely on software to prioritize safety or provide safety interlocks. So over the protestations of Areva ("it works in France," "we're great because 80% of our electricity is nuclear," "trust us because we know what we're doing," "our regulations are better than yours," etc.) the NRC rejected the prioritization module with this: Summary of NRC Staff Review of AVR-42 Priority Module Topical Report. Areva subsequently withdrew the topical report and a little later suspended its plans for an EPR in the United States.

EXAMPLE 2

Mitsubishi, a Japanese company, was planning to build its new APWR - like Areva's EPR, an advanced pressurized water reactor with active safety systems - in the United States. It had planned to use a completely digital, software controlled safety system. It based its design on work that had been done under the governance of Japan's Nuclear Safety Commission (which was discredited by allegations of collusion with TEPCO and by the results of the accident at Fukushima Daiichi, and it is now replaced by Japan's Nuclear Regulation Authority). The NRC issued a very damning assessment of Mitsubishi's software engineering plans for APWR digital instrumentation and controls. Like the French Areva, Mitsubishi gave the usual excuses: "It works in Japan," "Trust us because we are experts," etc. The bottom line, however, remains the same: Mitsubishi is indefinitely delaying its plans for an APWR in the US.

ASIDE: what the reader may not realize is that Mitsubishi was involved in another debacle. It provided replacement steam generators for the San Onofre Nuclear Generating Station (SONGS - a two unit site) in California back in the 2000s. Those steam generators developed tube leaks: primary water - reactor coolant - at 2235 psig leaked into the secondary side. The cause was vibration and subsequent tube rubbing at the upper quadrants of the steam u-tubes. Reactor coolant by definition is slightly radioactive. That made the steam side slightly radioactive. When the steam exhausted from the turbines into the condensers, air ejectors which maintain the condensers at a vacuum (for increased plant thermal efficiency) exhausted this radioactivity out the steam jet air ejector exhausts into the environment. Now the amount of radioactivity released was minuscule and non-injurious to either plant or animal life. But that fact of the matter is that it occurred, and did so on the eco-wacko left coast where liberals go nuts at the first mention of the word radioactive. All this which resulted from u-tube vibration could have been avoided had software QA been exercised with regard to the software-generated calculations which were done to analyze and predict u-tube vibration in high steam flow conditions at 100% power. But Mitsubishi screwed up [and so did Southern California Edison (SoCal), the plant owners]. Now SoCal had a plan to repair the tubes, add stiffeners to prevent vibration and rubbing, and restart one of the two units at 80% power. That plan was submitted to the NRC. But all the anti-nuke wackos for which California is so famous were in an uproar and the NRC dragged its feet to the point where SoCal without revenue from an operating plant could not keep staff employed on site for a promised restart that the NRC might never permit anyways. Thus, Democrat Governor Jerry Brown (that dropout from the latter 70s and early 80s - just remember Linda Ronstadt his concubine and the phrase Governor Moonbeam, and you got the picture) and the leftist progressives won. SONGS is now shutdown and to be decommissioned, and Mitsubishi is in large measure to blame for playing fast and loose. So of course when it came to Mitsubishi's digital I&C software engineering plans for APWR, the NRC was going to be wary. Do you, dear reader, get the point? Don't mess around with regulatory compliance. END ASIDE.

EXAMPLE 3

GE-Hitachi, a joint venture of the US General Electric Company and the Japanese Hitachi Company, have developed two different advanced reactors: the Advanced Boiling Water Reactor (ABWR) with internal reactor water recirculation pumps, and the Economically Simplified Boiling Water Reactor (ESBWR) which is all natural circulation and with completely passive safety systems. In the design of a typical boiling water reactor, control rods enter the core from the bottom, and steam separators and dryers are at the top above core internals. Since water boils in the core, steam ascends out the top. The separators remove moisture droplets which inn turn fall back into the downcomer region for subsequent recirculation into the core, and the dryers perform the last the drying out process before steam is routed out the main steam lines penetrating the reactor pressure vessel (RPV) to the high pressure turbine. "GE Hitachi allegedly made false statements to the NRC and Department of Energy about a component of the advanced nuclear Economic Simplified Boiling-Water Reactor (ESBWR) known as the steam dryer." GE Hitachi subsequently agreed to pay a 2.7 million dollar fine because of this (without actually admitting any wrong doing). The US Department of Justice statement is here. Guess how many GEH ESBWRs are being built in the United States (or anywhere in the world for that matter)? NONE. ZERO. ZIP POINT SQUAT. Now that may change if Dominion's North Anna project starts up again some day. I certainly hope it does. The GEH ESBWR plant is really a great design, and its passive safety features for a large scale (1600 MWe) reactor are unparalleled. But the less here is clear: obey the regulation, PERIOD.

CONCLUSION

In each of these examples (and there are others, like KEPCO's APR1400) the companies involved played fast and loose, acting as though they get in interpret the regulation, and behaving as if their design is so safe that regulatory scrutiny is an afterthought. Well here is the truth of the matter: for better or for worse the US NRC makes the regulation. It is the arbiter and decider and interpreter of said regulation. It is the height of arrogance for a company to act as the interpreter, deciding what is right and what is wrong. Doing so is just like a Christian deciding he gets to personally and individually interpret Sacred Writ, contrary to what St. Peter says in verse 20 of chapter 1 in his first epistle: "Knowing this first, that no prophecy of the scripture is of any private interpretation." The Magisterium of the Church (the Pope, the Ecumenical Patriarch of Constantinople, and the Bishops with them) determined at the Council of Rome (A.D. 382) and Third Council of Carthage (A.D. 397) what Books would be in the Bible and what would NOT be. Books like the Shepherd of Hermas and the Didache while valuable and salutary to read were rejected, and Books like James, Jude and Revelation while argued about were accepted. The Magisterium determined that under the inspiration of the Holy Spirit. Therefore, the Magisterium is the authentic interpreter, NOT the reader. In the same way no company (like Areva or Mitsubishi or GEH) gets to interpret the regulation. Rather, unflinching obedience and verbatim compliance is required and demanded. It's that way for the Christian for the safety immortal souls. And it's that way for the nuclear energy professional for the safety of mortal bodies.

Now unlike Sacred Writ - and here is where the analogy breaks down - regulation isn't perfect. That's why in a free Republic (NOT a Democracy - you do NOT get a vote) we have a process for changing regulation by petitioning the regulatory directly, by going to Congress which oversees the regulatory body, and by appealing to the President. Regulatory constriction should be addressed at that level and not by simply ignoring or reinterpreting it to one's personal profit and satisfaction. If the NRC is wrong in being strangulating - and it is - then equally so are the nuclear companies wrong for playing fast and loose. This is a two way street and both sides are equally responsible, and should be held accountable by both the voter and the consumer (who are often one and the same).

No comments:

Post a Comment