Heavy water versus light water in Ontario nuclear war: CANDU denouement

A few years ago I told Steve Paikin in a TV interview that when it comes down to a choice of nuclear reactors for Ontario, I tend to cheer the same way I do in international hockey tournaments: for the home team. I feel the same way today. Distill the choice between the two nuclear technological contenders in Ontario today—the CANDU EC6 and Westinghouse AP1000—to the essence of the decision, and your choice is between Canadian and foreign technology. You can watch the interview, which took place in 2008, here:

Of course there are important technological differences between the EC6 and AP1000. The former is 700 megawatts, the latter 1,200. The CANDU runs on natural (unenriched) uranium, is heavy water moderated, and features 380 individually pressurized, horizontally arranged fuel channels; the AP1000 runs on enriched uranium, is light (ordinary) water moderated, and features a single vertically arranged fuel assembly inside a single pressure vessel.

These major technological differences give rise to many other implications regarding the day-to-day and lifetime operation of their respective power plants. These implications form the basis for arguments in favour of either of these reactors, and I must say there are good arguments both ways. But I won’t go into that here. I will just stick to the implications of either technological choice on the supply chain for reactor components.

The point is, both these machines work. Both are—and have proven over decades of continual operation to be—far more dependable and robust than any other kind of non-nuclear thermal (steam) power plant. Either would perform as expected in Ontario’s power system: either would provide decades worth of affordable, reliable, air-pollution-free electricity.

The main difference is, CANDU components are for the most part made in Canada, and the major components of the AP1000, including and especially the heavy forged pressure vessel, would come from another country.

So, all things being equal, which machine should Ontario choose?

If Ontario were to choose the CANDU, then a lot of people in Canadian firms would get a lot of work. If we choose the Westinghouse machine, less Canadians would get work. It is that simple.

I would be happy to hear an argument that puts the AP1000’s technological and operational characteristics so superior to the EC6 that they outweigh the jobs factor. But I have yet to hear such an argument. The CANDU is a pretty solid machine.

0 0 vote
Article Rating
Notify of
Newest Most Voted
Inline Feedbacks
View all comments
7 years ago

The ACR’s dependency on *two* isotope separation infrastructures, one of which won’t be in Canada any time soon, bothered me. I’m glad the EC6 is now the frontrunner.

7 years ago
Reply to  G.R.L. Cowan

I’d like to see a bunch of EC6’s in the USA, for two reasons:

1.  If the papers I found are correct, they can get about 18000 MW-days/ton out of re-sintered, spent LWR fuel (DUPIC).  This adds about 40% to the energy yield from the LWR fuel stream without generating another gram of SNF.
2.  The final SNF would have more plutonium, ideal for S-PRISM feeds.  Again, not one gram of additional SNF!

Peter H. Roth
7 years ago
Reply to  Engineer-Poet

The heavy water reactor produces more fissile Plutonium then the light water reactor. Many folks are concerned that Plutonium is a direct bomb making isotope whereas Uranium must be enriched to almost 90% U235, which is a much more complex process.
Making heavy water is also a complex process, however my Chem. Engr. research process, many years ago, we invented a very efficient process for making Deuterium from ammonia synthesis gas.
The govt. confiscated our thesis documents, but we got full credit for the work. I guess at the time it was considered too sensitive while we were in a “Cold War ” situation.

Steve Foster
7 years ago

Steve: what about the question of security of fuel supply. Canada doesn’t possess enrichment facilities, so fuel for the AP1000 would presumably have to come from the US. What does that mean in terms of security of supply? Is it possible that supply could be threatened, or subject to arbitrary price increases depending on US politics or market dynamics (nice reactor you got there, be a shame if you couldn’t refuel it…) Who can guarantee stable US policy regarding nuclear fuels, enriched U, and foreign governments for the next 50 years? There is a certain attractiveness to owning the whole technology and fuel cycle.

7 years ago

The Candu 6E is an ancient Generation 3 product from a infamously discredited company now an international pariah with links to dictator Gaddafi. There are no chance of sales outside of Canada. We understand that the Gaddafi’s want to sell the machines for twice what AECL was building them for.

The Ap1000 is the best Gen III+ machine out there with many sales pending and 8 under construction with those in China at the same price the last Candu’s built there were sold for.

With spinoffs at least the Canada nuclear industry could benefit from the AP-1000 future. There is no future with the Candu.

7 years ago
Reply to  seth

How hard would it be to add passive cooling and other Gen III+ features to CANDU?  It doesn’t look all that hard to me.  The features which allow the AP-1000 to cool passively don’t appear to be in the RPV or other major components, so they could be tacked onto CANDU in a similar way.

Something as simple as a natural-draft dry cooling tower on top of the reactor, with weight-operated valves from the steam generators to the cooling coils in the tower, would do it.  The weights would be held up by solenoids.  Lose power, cooling valves open, done.

7 years ago
Reply to  Engineer-Poet

Because a lot of heavy water doesn’t ruin the neutronics, CANDU has always had a lot of heavy water in the calandria. Its heat capacity amounts to a lot of passive cooling. As I recall, it has its own cooling system that removes the heat deposited by neutrons and gamma rays, and keeps it down to 80°C. Jeremy Whitlock could tell you more, or has already done so in his Canadian Nuclear FAQ.

7 years ago
Reply to  G.R.L. Cowan

I’ve been over the CANDU presentations, and there is a failure mode where loss of cooling eventually leads to melting of the pressure tubes.  It would be better to make the plant walk-away safe.  If you can do that by dumping steam to natural-draft condensers so the water recirculates indefinitely, you’ve eliminated a failure mode and a source of public worry.

I admit to being curious about other design features, but I’d have to ask an old AECL engineer or three to find out why certain decisions were made.

An Interested Observer
7 years ago

The choice of which nuclear technology will best serve Ontario’s future energy needs must be based in my view principally on considerations of expected life-cycle performance and life-cycle costs. Basing this very important decision on protectionist policies towards indigenous industries is an unsound approach.

Firstly, about 70% of the equipment in any NPP is independent of the technology used in the nuclear reactor; thus, choosing an APR100 will still allow ample opportunity for Canadian based industry to participate in the supply and delivery of much of the plant equipment. Some manufacturers of specialized CANDU components such as fuel channel assemblies (end fittings, closure plugs, shield plugs, liner tubes, garter springs, etc., etc.) will be disadvantaged of course, but to warp the decision making process to protect these minority interests could disadvantage many generations of future Ontario electricity consumers.

Secondly, the construction and installation of a new build project will account for about 50% of the capital cost. Here again, the choice of reactor technology will not significantly change the scope, nature and cost of this work. One would expect that Canadian suppliers of goods and services to large engineering projects should be able to compete and win.

I’m not anti-CANDU; if an EC6 is the right machine that will be wonderful. But that choice must be based on a dispassionate examination of many factors over the lifetime of the plant – operating economics, simplicity of maintenance and operation, reliability, safety – are just a few.

Buy the best machine that will get the job done in the best way for the maximum benefit of the taxpayers; not for the benefit of a few manufacturers of unique hardware.

7 years ago

I agree with the article on the best path forward. CANDU is not ancient: several hundred $M have been spent on development of the next generation EC6. With on-power refueling, and lower cost uranium operation, combined with safe heavy water moderation…why would you consider the AP1000 LWR? There are so many safety features in the current CANDU design that you CAN walk-away shutdown safely.

Eric L. Hanson
7 years ago

Two questions.

1. Which model of CANDU reactor uses slightly enriched uranium of 1-2%?

2. What are the relative costs? You get a 700 MWe CANDU for x billion dollars and a 1200 MWe Westinghouse for Y billion dollars. What are x and y?

7 years ago
Reply to  Steve Aplin

What’s the advantage of using SEU, since there is no downtime for refueling?

It sounds like DUPIC would be as good as SEU, and eliminate one spent-fuel disposal problem and cut another one by 60% or so.

An Interested Observer
7 years ago
Reply to  Engineer-Poet

There is an excellent discussion on the use of SEU in CANDU reactors at http://www.nuclearfaq.ca/brat_fuel.htm

7 years ago

Very good introduction to the issues there.  Thanks.