1. Very informative posting overall, but it looks like a few of the reference links don’t work (for me, at least).

    It would be nice to have a short bio of the author to know his background being simply that he’s a P.E.

    Also, there seems to be an error (wording-wise, at least) in the following paragraph:

    “On occasion units have dropped over 440 MW to operate at 46 percent of full electrical output. On one early 2011 November weekend, according to an IESO Generator Output and Capability Report, one of the units even reduced reactor power to 385 MW and with steam bypass brought the electrical output down to 208 MW, which is around 25 percent of full power. Under these circumstances this is better than the 70 percent dispatchable limit of the CCGTs.”

    Reactor power of 385 MW with electrical output of 208 MW doesn’t make sense. That would be well over 50% efficiency, while nukes are typically right around 33%.

    1. “Being” should have been “beyond” in that second paragraph, and to re-iterate, that was very informative.

      If anyone could comment on any major differences in the controls/dispatching of the ISOs here in the states from what is described in the appendix or whether they’re essentially the same, that would be a nice addition.

    2. Joel:
      The author consistently refers to reactor power in terms of the corresponding electrical generation, i.e., MW(e) rather than in terms of reactor thermal output, MW(th). Thus 385 MW would be about 1200 MW(th), roughly 46% of full reactor power. It’s not clear whether the 208 MW is gross or net electrical output; station load for a Bruce B unit is around 60 MW, suggesting that steam dumping in the order of 120 to 160 MW (electrical equivalent) during the operating period.

  2. The French system is much more complex that this. Yes, they do load changing, in fact they do load “following” with their units, but not all their units do this. It depends on how long the fuel has been in the reactor. They rotate their load following to various online reactors and keep others at base load. In addition they have several GWs of pumped storage in the Alps and, they have a small number of peaking GTs which, apparently, the are building a few more of.

    The new EPR…it appears France is building 2, and perhaps only 2 more…has very good load following capabilities. All of France’s nuclear plants need to be replaced over the next 30 years or so. Likely newer, even better load following ones will have to be built.

    Gen IV reactors of most varieties (and yes, the French are leaders in this area too) are quite good at load following and should follow load as well as a CCGT or even hydro.

    1. If the ‘or so’ is 60 years I would agree David. EDF has ordered a boat load of new SGs so maybe those perfectly good old plants will last a 100 years.

      The new EPR has an equipment hatch that allows replacing reactor vessels & SGs without cutting a hole in the containment.

    2. I’ve heard that the older unit like Fessenheim almost only do base-load, and that it’s the newer unit like the 1300MW N4 ones, that do most of the load following.
      But you are probably right that it also depends on the fuel cycle.

      About the GTs, the truth is that it’s politically easier, and faster/cheaper to build one. The plans of Proglio to build a lot of them are a bit depressing. However in a very discrete way, things may be changing recently, like is shown by this common declaration with UK and 10 other pro-nuclear countries : https://www.gov.uk/government/news/uk-at-forefront-of-european-nuclear-expansion
      I certainly can tell you not a single news outlet said anything about this happening.

  3. I have to apologize…I only read the ‘front page’ of the web site here and didn’t see the details on this question about load following and CANDUs. It’s very good and covers the French grid accurately as far as I can tell.

    I do have some questions/comments, however.

    I’m wondering why the CANDU’s are not allowed to do frequency control? Control is done by speed sensors on the turbine shaft generally, that can detect minute variations in frequency, or ‘speed’ of the system. These contol the governor valves on the turbine itself that regulate the steam. All but hydro uses these valves so why can’t the CANDU as well? I would like to see a fuller explanation.

    There are obvious limits on the rate-of-change for frequency as the running up/down/up to keep the system stable is a complex dance between different generators. But there are dozens of limit relays that detect problems arising from this sort of scheme…to protect the generators primarily from over or underspeed conditions which would wipe out the generators and possibly the main bank transformers.

    I think the PRO-nuclear movement should start to campaign for MORE nuclear to decrease the level and *phase out the use of dangerous polluting natural gas* in the Province.

    Also, it seems odd that with Ontario’s (and neighboring Quebec) massive hydro infrastructure that they haven’t started a long-term program of converting them to pump storage which would likely solve a lot of their problems.

    1. In Quebec’s case, over half the hydro plants are run-of-the-rive, and the others suffer reservoir inventory issues during parts of the year. This manifests itself in Spring and Fall with a large surplus of water and low power demand, and low head conditions in Summer and Winter coupled with high demand for electricity. Thus when the reservoirs would need pumping up the most, there is the least spare power to do it.

      Ontario’s hydro plants are much smaller than those in Quebec to start off with, and I expect that the same issues would hold there.

    2. The regulator’s expectation certainly used to be, and probably still is that the licensed control room operator is in firm command of the plant output at all times. Having a CANDU respond automatically to requests for power increases/decreases commanded by a remote third-party operator in an IESO control centre would not be viewed as a prudent mode of operation. Besides, there are some operating conditions at a CANDU power plant – on-line refueling or safety system testing for example – that are best done with a very stable reactor power level.

  4. This begs the question are Candus better than LWRs? Does anybody know how different the amount of waste produced per unit of energy is comparing one of these to a LWR. The Candu’s use Heavy Water allowing them to run on unenriched uranium. The Candus can also use Thorium. So it seems fuel preparation is less of a political concern to enviromentalists.
    Candus can also run on Thorium. Another question is Candus could be sold to large American population centers and the NRC would approve. Why doesn’t the US want to try building a Heavy Water reactor of their own? Also recent interest in the Thorium MSR in a deal between Australia and the Czech Republic makes the US appear overly cautious and behind the times. These are touchy topics and I mean no disrespect. We all know the issues are complex.

    1. No offense intended, but no US utility is going to build a CANDU reactor – ever. There is no comparison between the longevity and lifetime reliability of a PWR and a CANDU plant. I’ve worked extensively in both and can tell you that a 40+ year old PWR generally runs very well, while a 40 year old CANDU generally does not.

      As a point, I would challenge you to look up the capacity factor of either of the two (operational) units at Pickering A and compare them to the Ginna nuclear plant on the other side of Lake Ontario. Ginna is somewhat older and has *many* more operating hours (since it didn’t have the extended shutdowns that the Pickering A units had both due to retube in the mid-1980s and the 1997 layup).

      The story is similar at Bruce A – a required multi-billion dollar heart transplant at mid-life isn’t a compelling selling feature for a reactor technology.

      I have one comment regarding the discussion of load following above. It is my recollection that long-term operation of steam dump is not particularly good for the health of a condenser since the steam conditions are radically different. CANDU is at somewhat of a disadvantage in that you cannot as easily maneuver the reactor thermal power output to match the turbine (due to the need to prevent Xe-induced shutdown), requiring far more steam dump operation. Maneuvering a PWR or BWR is far simpler over the majority of a fuel cycle. On this I speak from experience, having been the lead reactor engineer at a PWR plant for over 10 years. Typically we required no steam dump operation for maneuvering, and only for a short time during load rejection transients.

      As to your question about the waste produced per unit of energy – CANDU loses here. The discharge burnup of CANDU fuel is much lower. While that means that it is technically not as bad since it has fewer fission products, it is still plenty bad. A typical LWR fuel runs to between 4-5 times higher burnup. That means a much lower volume of high-level waste. The spent fuel pool that I was previously responsible for had all fuel from 40 years operation (minus 40 assemblies) and was much smaller – to my eye at least – than the pool at a single-unit plant like Point Lepreau (which I don’t think could hold all of the fuel from 40 years of operation at high capacity factor).

      Don’t get me wrong – I started my career in CANDU plants. The engineer in me loves them. I’ve worked extensively at several of them. I sincerely hope that Bruce A 1/2 are highly successful when they come back (I was working on Unit 1 and walked through Unit 2 on the sad day it shut down in October 1995). It’s just that US utilities aren’t “behind the times” in not adopting CANDU or other unproven technologies – they’re just making wise investment decisions.

      1. While CANDUs will never be built in the US, I would hardly typify them as “unproven technology.” I’ll note too that discharged fuel from LWR could continue to be burned in a CANDU squeezing even more energy out of a given amount of fuel with the DUPIC (Direct Use of spent PWR fuel In CANDU) fuel cycle. As well while LWRs produce a lower volume of high-level waste per MW, both waste and tailings from enrichment must be added to that inventory, and the SWRs subtracted from the efficiency. Just because this happens elsewhere, doesn’t mean it can be forgotten.

        Finally CANDU isn’t a single design, and improvements are ongoing. The newer versions are somewhat improved than the Bruce A which went critical in 1976.

        1. DUPIC is a concept and a fairly impractical one last time I checked. I remember going to a talk on it in the mid-1990s and I don’t think it’s gotten anywhere in the interim. There is no way it will be economic in the conceivable future to put irradiated PWR waste in a CANDU reactor given the difficulty of handling it and making it into new fuel. There is no denying the fuel cycle flexibility of the CANDU, but again in the context of investment decisions being made today this is somewhat irrelevant.

          You are right that waste from tailings needs to be considered, but it’s not quite the same problem to deal with as high-level fuel waste. I’d take 10x the fuel heavy metal weight in tailings over 5x that amount in high-level waste.

          Do you mean Separative Work Units? Of course with several gas centrifuge plants coming on line in the next few years the energy intensity of enrichment is going WAY down. Let’s not forget the very significant capital cost of the heavy water in a CANDU plant which is greater than $0.5B per reactor unit. I’m not so versed in CANDU fuel cycle economics, but I can say that just the cost to refurbish a CANDU at mid-life might cover the fuel costs of a 1000 MWe PWR for 40 years (I’m basing this on a PWR fuel cost ~$80M/18 months, and a CANDU refurbishment cost of about $2.4B/unit based on the latest Bruce A estimates I’ve seen).

          Honestly – this isn’t the best place to debate the relative economics of different reactor technologies. I’ll leave that to the utilities that buy them to do exhaustive studies of the economics of each reactor technology. I won’t list the number of reactors currently under construction by reactor type here.

          You’re right – technologies evolve. AECL has done some incredible things over the years (I still marvel at the engineering in the SLAR tool used to reposition pressure tube garter springs). I’m not sure – does the ACR-1000 require retubing? As a utility customer I’d be nervous if so – it’s just too big of an uncertainty given past history.

          Lastly, remember that LWR technology has evolved as well. For example, the passive technology available in the modern LWR designs is pretty impressive, especially in a post-Fukushima world.

          Back to the original point though. The original poster asked about whether CANDU was better than LWR technology. CANDU may have a niche role to play in nuclear going forward, but it’s really not the premiere technology that some believe it to be, and that seems to be borne out in the current marketplace.

        2. And just to clarify – I wasn’t saying CANDU was unproven. The original poster also talked of Thorium molten-salt reactors, which I would consider a commercially unproven technology. Interesting yes, but not something a utility is going to consider in my lifetime.

        3. Sorry for the late reply, the e-mail was misplaced.

          The DUPIC fuel cycle. In unit 1 of the Qinshan Phase III plant in China, there has been a demonstration using fuel bundles with RepU from PWRs blended with depleted uranium to give natural uranium equivalent (NUE) fuel with 0.71% U-2356. It behaved the same as natural uranium fuel. Subject to supply from reprocessing plants, a full core of natural U equivalent (NUE) is envisaged. Following design, licensing, etc, full core implementation in both China’s CANDU reactors is envisaged by the end of 2013.

          Canada, and South Korea, which hosts four CANDU units as well as many PWRs, have been working on a bilateral research program to develop DUPIC, and the Korean Atomic Energy Research Institute (KAERI) has been implementing a comprehensive development program since 1992. KAERI believes that although it is too early to commercialise the DUPIC fuel cycle, the key technologies are in place, and they are ready to move forward.

          So dismissing DUPIC as impractical may be premature.

          The cost to refurbish a CANDU are more a question of politics than economics, at least in Canada proper as governments have tended to see these as make-work projects. The referb of Korea’s Wolsong-1 was both cost effective and reasonably quick.

          As for the market, to that the degree that it is not manipulated by the larger players and their political supporters, consider that for some years CANDUs were the only commercial stations being built period.

          I don’t think CANDUs are the be all and end all of reactors, but they still have a place, as India’s plans to start marketing their CANDU 3 knock-off show, in places that may not want to find themselves beholden to the US or Russia for access to enriched fuel.

  5. If you reprocess your fuel, the recovered plutonium with thorium can give a long life fuel. The Canadians will not, of course, publicize it as it may have bearing on uranium sales.

    1. Reprocessing CANDU spent fuel is just not economically feasible, as new uranium is too cheap. This is true of reprocessing LWR spent fuel however it is more likely that this will break-even long before CANDU used fuel.

      For reprocessing to become a credible alternative within the next 50 years in Canada, we would
      soon need to find a much cheaper process and one specially tailored to low burnup CANDU fuel.

      CANDUs can, and have in test batches, burned thorium fuel bundles, which can be introduced without modification to the reactor. Again, there is no economic incentive in Canada to do so.

  6. CANDUs are not approved by the NRC. Something about their positive void co-efficient of reactivity. There is/was some took of their new model, which uses a lightwater moderator I think, and filling out the apps for it. I think it cost $200 million to file for approval. I wouldn’t waste my time in the U.S. with all the hurdles and uncertainty around. See the Arne Grundersen entry above this one.

    1. CANDs are never going to be approved in the States as long as Canada won’t approve LWRs for use in Canada. This is a trade issue, and has little to do with the technical excuses given by both sides.

  7. Ontario’s Long Term Energy plan (link: http://www.mei.gov.on.ca/en/pdf/MEI_LTEP_en.pdf)is a plan for a disaster. By deciding to limit nuclear generation to 50% and precipitously shut down all coal-fired generation when the potential to expand hydroelectric generation (in Southern Ontario at least)is limited, severely restricts the availability of electricty which is needed to sustain a growing industrial economy.

    Only under the conditions of such a pre-planned energy crisis would anyone be talking about windmills and solar panels which require a huge physical and monetary investment to build and generate very little electricity.

    At a time when the Canadian government has just sold AECL’s nuclear reactor division to SNC-Lavalin/CANDU Energy, the author’s article indicates that the potential, and need, for nuclear power in Ontario (let alone in the rest of Canada) is much greater than current policies indicate.

    For example, if Ontario were to plan on generating the 205.5 Terawatt hours per year that Ontario’s Long Term Energy Plan calls for by 2030 by depending upon nuclear power for 75% of the generation, a very rough calculation suggests that Ontario would need not only to refurbish the existing rectors and build 2 new Darlington plants (total of about 14 GWe) but build another 13-14 GWe nuclear. While the numbers are rough this suggests about 75% nuclear, 20% hydroelectric and 5% natural gas generation of the 205.5 Terwatt hours per year that Ontario is said to require by 2030.

    I would like the opportunity to talk to and perhaps interview the author for 21st Century Science and Technology magazine.


  8. Just saw this recent piece on the CBC web site.


    The comments below the article are more interesting than the article itself where a business man (who is part of the Canadian marketing team) revealed that Britain is sitting on a huge stockpile of plutonium. As others have posted elsewhere here, CANDU reactors can burn any fuel including spent fuel from light-water and breeder technologies. This is a win-win way for Britain to move forward.

    1. They are also seriously considering PRISM for the exact same usage.
      But CANDU has indeed the advantage of being already proven.

Comments are closed.

Recent Comments from our Readers

  1. Avatar
  2. Avatar
  3. Avatar
  4. Avatar
  5. Avatar

Similar Posts