The Pickering hearings – some last comments on evacuation, GE-Hitachi and Orangeville

The treatment of PSA (Probabilistic Safety Assessment) was much better at the Pickering hearings than at the Darlington hearings of last December. In earlier posts I criticized the cavalier way that OPG and CNSC staff tossed around PSA probabilities as representing accident probabilities. I’m happy to say that there was much less of that this time. There was still confusion about limits, goals, objectives and other associated terminology but I had the impression that the problems were more in communications than in fact. I was also heartened to hear CNSC staff acknowledge that unknown unknowns (Black Swans) were the main lesson of Fukushima – “expect the unexpected”. Being a very conservative organization OPG still tends to cling to the dubious use of PSA (or PRA) for overall accident frequency prediction over and above its legitimate value in organizing and focusing safety related issues.

Evacuation plans were taken seriously at these hearings. Six months earlier at the Darlington hearings the fundamental questions concerned the existence of a plan and who was in charge. At the Pickering hearings much credit should go to intervenors from the Canadian Environmental Law Association (CELA) who presented detailed work they had done on evacuation issues. This was useful and contributed both to the hearing process and also on focusing those responsible for detailed planning on the nuclear aspects. Emergency plans are now designed to accommodate generic accidents of all types but there will be problems specific to nuclear emergencies that might not be covered. There was significant drilling down into specific aspects such as the distribution of KI pills, traffic patterns in terms of radioactive deposition, and suitability of building types for sheltering from radioactivity. The latter are issues specific to nuclear accidents and currently not well covered in the broader plans of the various emergency organizations.

Because the Pickering reactors are about 30 km from downtown Toronto this topic was very sobering indeed. A highlight of the hearings for me was an intervention by a person who talked about her experiences as an evacuee during hurricane Hugo in 1989. This put a human face on what might otherwise be considered a theoretical discussion. The consequences of a “doomsday” type accident at Pickering would be unthinkable and I believe most people both within and outside the nuclear industry would agree that it’s time to close down the Pickering reactors. The only issue is when.

In contrast to these worthwhile interventions, a few intervenors complained about the GE-Hitachi nuclear fuel operation in Toronto. Aside from being largely irrelevant to the topic of the hearings, it seemed silly to complain about it. First of all, the facility is completely benign and thoroughly inspected with negligible possibility of any accident. Second, it’s been there for about fifty years and anyone moving in since then who objects to it clearly didn’t do their due diligence in discovering its existence before they bought houses in the area. Boiled down they are essentially admitting that they, their real estate agents and lawyers were asleep when they bought. It’s like those people who move in near an airport and then whine about the noise. These intervenors not only are complaining about a non-existent danger but also by doing so they lower neighborhood property values including their own. In my opinion this is just dumb and I have no patience with them.

Arnie Gundersen, former reactor operator turned prominent US anti-nuke, provided comic relief as he argued in a genial way against granting a renewal. He made disparaging comments about CANDU reactors being “an evolutionary dead end” meaning that with the ACR-1000 dead there will not be any more new versions of CANDU after the EC tweak of the CANDU 6. He’s probably correct but it’s still not easy to hear. His main point was that the Pickering reactors were among the oldest still operating in the world. True Pickering units 1 and 4 started in 1971 and 1973 but he was still wrong. They were refurbished (twice for unit 1) and returned to service in 2005 and 2003 respectively and thus, are in better shape than units 5 to 8 which came on line 1983-6 but have not been refurbished.

The fact that a waste disposal site will be needed to store the large volume of radioactive non-fuel bits and pieces arising from refurbishing and later decommissioning Ontario’s reactors was is not a surprise but the ham-fisted way it came out at the Pickering hearings certainly was a surprise. Someone, I believe in an OPG document, casually opined that this site should ideally be equidistant from the reactor stations and just by looking at a map came up with a site near Orangeville. If you don’t believe in coincidences, that town used to be (still is?) the location of an Ontario Hydro training facility and maybe OPG has some land available there for a waste site. Therefore, I would tend to ignore any subsequent back pedalling on this site by OPG.

What a way to introduce Ontario’s third nuclear waste storage site! I assume it would probably be a DGR (Deep Geological Repository). The first waste site is a DGR now under study by a CNSC-appointed committee. In it OPG will bury low and intermediate level waste from reactor operations at the Bruce site. The reason for the location is simply that the Bruce power station was located on Lake Huron for reactor cooling water and OPG owns a lot of land there (now leased by Bruce Power) at an approved nuclear site. Geological justifications were later found to fit these business considerations. Originally OPG planned to build this DGR under the lake but many in Canada and the US got very excited about potential radioactive contamination of the Great Lakes – the source of drinking water for 40 million people. Now they plan to build it near but not under the lake.

The second nuclear waste site is the long-term DGR for high level used nuclear fuel that some communities around Bruce among others in Canada have expressed preliminary interest in hosting. Once again this has provoked the contamination of the Great Lakes issue. Predictably the two DGRs are being confused deliberately by critics pushing the idea that the first will become the second. With Orangeville on the table, it now appears we have three DGRs in play.

I had thought that Bruce Power’s inept handling of its plan to ship contaminated steam generators for recycling was the leading nuclear waste public relations fiasco in the past few years. However, I’ve changed my mind. The casual Orangeville site reveal by OPG at the Pickering hearings was bungling at an even greater level turning Canada’s nuclear waste disposal efforts into a true three ring circus.

Pushing the Limits of CANDU Pressure Tubes

The basic issue at the Pickering hearings was the request of OPG to operate the four Pickering B reactors beyond the previously assumed lifetime of their pressure tubes. The lifetime was set in the past to be 210,000 hours (more accurately Effective Full Power Hours) and running Pickering B to 2018/2020 would according to OPG push the lifetime to 247,000 hours. With 8,760 hour/year and a capacity factor of 0.8 we get roughly 7,200 hour/year. At this rate 216,000 hours is about 30 years and 37,000 more hours amounts to an extra 5 years. With the exception of reactor number 8 which will reach the limit around 2015, it looks to me that the other three Pickering B reactors will be at the 210,000 hour limit in 2014.

This issue has wider implications and indeed could be the most critical issue for Canada’s CANDU program. For example, the six Bruce Power reactors will need refurbishment soon if the resources are available. Once again rough calculations indicate that some of them are already within a year or two of 210,000 hours in part due to their very high capacity factors in recent years. So what will the CNSC do when confronted with the same issue at Bruce? Cancelling the Bruce licenses just isn’t feasible and so one can safely assume that sometime in the period 2014-15 we will see ten reactors (four Pickering B, two Bruce A and four Bruce B) all operating with pressure tubes past the 210,000 hours mark.

It is already well-known and observed phenomenon that CANDU pressure tubes deform as they age. This is due to an effect known as radiation-induced ‘creep’, the latter meaning the metal becomes like a very slow flowing plastic driven by the heat generated by fission in the uranium fuel. An American colleague of mine used to call CANDU reactors “the world’s largest creep experiments”. Creep arises on the atomic level from the displacement of metal atoms by high energy neutrons (before they are slowed down by the moderator) and by the embedding (ingress) of hydrogen isotope atoms in the metal forming hydrides which further degrade the metal’s macroscopic properties.

In operating CANDUs various measures are taken to control the dimensional changes of the aging pressure tubes, for example, the ends of expanding tubes may be selectively clamped to encourage uniform growth in both directions. The tubes also sag and swell as they age changing the geometry of the nuclear assembly thus affecting both its thermal and neutronic aspects. How much of this degradation a reactor can take before it becomes unsafe to operate? This is a very complex issue and a great deal of R&D has been done on this subject since the inception of the CANDU concept. The reason was that in order for a natural uranium/heavy water reactor to work (achieve criticality) its metal structure has to be minimized to avoid excessive neutron losses. Therefore, the pressure tubes had to be thin and it was acknowledged from the outset that creep would be a problem in CANDUs.

The history of CANDU pressure tubes was bumpy at the beginning. In 1974 leaks were discovered near the rolled end joints in some tubes in Pickering A reactors. This problem was attributed to delayed hydride cracking which arose from how the joints had been fabricated and was corrected by replacing the faulty tubes. In August of 1983, a pressure tube ruptured in Pickering A reactor number 2, some twelve years after it began operation. There was no leak-before-break in this case which caused dismay since it was always touted as a safety feature. This failure was attributed to friction damage caused by spacers around the tube and also to the alloy the tube was made from. This accident required the re-tubing of first two Pickering A reactors with tubes made of a better alloy. (AECL also lost the royalty it had on the electricity produced which was the start of its unremitting financial woes.) This event resulted in even more R&D on pressure tube degradation from which a target limit of 210,000 hours was set.

So what has happened to change this limit? What is the scientific basis for extending the lifetime of the pressure tubes? To me this is the critical question. In the first place it is reassuring that not much negative has happened in the pressure tube field for many years which would indicate that at least 210,000 hours is a reasonable limit. However, in my opinion OPG hasn’t done a very good job of informing the public of the science behind their request to go to 247,000 hours. The Pickering hearing transcript offers little in the way of evidence that a change is justified. OPG talked about a research program started in 2009 that produced some eighteen reports submitted to the CNSC. I haven’t been able to find copies of these reports on the CNSC or OPG website. Comments made about this program at the hearings didn’t convince me of its validity.

More significantly I can’t find any mention of this work in international journals such as the Journal of Nuclear Materials or Metal Physics. Maybe I’m not looking in the right places? A result with such high impact on reactor operation and safety should be peer reviewed by refereeing for publication in a high quality publication and preferably presented at widely attended international conferences. To me it is ironic that the CNSC stressed the need for peer reviewed publication of the epidemiology work showing no increased cancers near nuclear stations but not for the pressure tube program, the main topic of the hearings. By peer review I don’t mean sending the results to a few foreign experts for their opinion – this isn’t anything like proper peer review. I’m afraid that the CNSC too often does that because it buys into the nuclear industry’s unhealthy obsession about propriety information to the detriment of any realistic quality control over the science involved.

In the final analysis it boils down to the old British expression “suck and see” meaning inspections at reactor outages to see how the tubes are doing; the reactors have indeed become “creep experiments”. Sudden (without leak-before-break) failure of any one pressure tube could be handled by the existing safety systems as the 1983 accident showed. I suspect that’s the basis of the CNSC granting OPG a limited license extension to operate reactors up to the 210,000 hour limit. However, I doubt one would want to continue to operate any of the Pickering B reactors after one tube failed. What would that do to the continued operation of the six at Bruce? The nuclear enterprise in Ontario has entered perilous waters where the failure of just one pressure tube could sink it.