What happens when nuclear plants go down for weeks at a time?

[kristopher]

A lot of posts are being made that point to failures of wind turbines. Far from rejecting the premise that such failures are a routine part of wind energy, I agree completely that they are.
However the corresponding truth is that routine failures are pretty common to the nuclear industry also. However, when a wind turbine goes out of service, the grid loses at most a couple of megawatts of capacity and life goes on as usual. But what about nuclear power. The incident below shows that 11 days after a minor technical problem the plant is still only producing a small fraction of its capacity. And last year another small technical failure shut the plant down for 45 days.
The proponents of nuclear power like to claim that nuclear power produces at full capacity more than 90% of the time. However the truth is that the LIFETIME AVERAGE of US nuclear fleet is only 71%. They arrive at their 90%+ number by the novel approach of simply excluding any plants whose failures would impinge on their averages - sort of like a ball player that gets to erase all of his strikeouts.

Another lie they love to promote is that the longevity of their plants. However, they arrive at the long life spans you hear of in exactly the same way. If we look at the global average of the in-service nuclear plants that have shut down permanently, the average lifespan is 22 years.

FRENCHTOWN TWP: Nuclear plant operating at low power due to shutdown

Published: Saturday, April 03, 2010

By Angie Favot

FRENCHTOWN TWP. –– DTE Energy’s Fermi 2 nuclear power plant reactor is operating at low power while work is done on a problem with a device that condenses reactor steam back into water.

Guy Cerullo, DTE Energy spokesman at the plant, said the plant shut down automatically March 25 because of an electrical short in a circuit on the plant’s turbine. Cerullo said the restart process began March 27. “The cause of Thursday’s shutdown was determined to be a fault in a protective circuit that conservatively shut down the turbine,” he said. “Right now, we are operating at 4 percent power while we assess an issue with the main condenser vacuum.” He said that as soon as the issue is resolved, the plant will return to full power, but he didn’t know when that would be. Cerullo said the vacuum is required to convert steam back into water and then to send it back through a closed loop system in the reactor.

The facility was shut down from Sept. 30 to Nov. 12 because of a hydrogen gas leak into the generator’s cooling systems. Cerullo said the two are not related...

http://www.thenewsherald.com/articles/2010/04/03/news/doc4bb774970f640919249272.txt

[dmallind]

1. ermmm...scheduled maintenance ring a bell

Typical downtime for a plant maintenance is about 18 weeks. A lot of it is work, but a good chunk is all the required and manifold redundant safety checks. The reason plants keep working safely.

[kristopher]

3. "Ermmm" this isn't "scheduled maintenance"

...and your post ignores the nuke industry misrepresentations used to secure government funding.

[dmallind]

8. this what? Shutdowns are almost always scheduled maintenance.

Are you really talking specific or generic?

[kristopher]

10. Read the OP...

Not scheduled mx.

[madokie]

2. I learned years ago that if the nuclear power industry says it

you best be taking that with a grain of salt. Surely man can and will do better than build more of these type generating plants.

[Go2Peace]

4. The nuclear "Greenwashing" is the current equivilent of the old Tobacco companies

This is very risky technology that contributes to cancer rates. The propagandize and use the courts and public favor through marketing pushes to disguize the truth. Don't believe it. This is not the future we want to build

[NNadir]

5. Very little. They are designed to do that, and power companies adjust the timing

in connection with expected power needs.

Most nuclear power plants operate for more than a year before requiring refueling. This is why their average capacity utilization energy/(time integrated power) is the highest of any form of electrical generation in the United States (roughly 92%). It's higher than coal (72%) and higher than all of the renewable toys that anti-nukes are always hyping. Wind operates, at best, at between 20% and 30% of capacity utilization, and the plants don't last very long, as I've been noting here, some plants can't make it a two years before being in failure mode.

Anti-nukes generally hate sciences they know nothing about, hence the question. Because nuclear is reliable, most outages are planned, usually for periods when the power demand is low.

It's amusing of course, to hear anti-nukes complain about another point in nuclear energy in which nuclear is vastly superior to the stuff they don't care about. Nuclear need not be perfect to be vastly superior to all the stuff anti-nukes don't care about. It only needs to be vastly superior to all the stuff anti-nukes don't care about, which it is.

[Statistical]

6. Complete lie. Capacity factor is based on all outages (planned or otherwise).

Capacity factor is very simple to calculate. It simply is power delivered divided by theoretical max over a period of time. Usually the time period is one year.

A theoretically perfect plant would run 24 hours a day at peak output forever. In a year a theoretical perfect plants generation would be 24 x 365 x nameplate rating. For 1 GW reactor that is 1 * 24 * 365 = 8760 GWh = 8.76 billion kWh.

Actual generation (over one year) divided by theoretical max is capacity factor.

By virtue of the calculation it includes any slowdowns and outages.

If a 1 GW reactor ouputs 8 billion kWh in a year than it has a capacity factor of 91.3% (8/8.76).

[joshcryer]

7. He's averaging over *lifetime* while at the same time...

...being two faced and saying "modern wind has a higher capacity factor."

If you were to add in all of those wind turbines over the life of wind that ever existed, uh, yeah, the capacity factor would be low (since all of those 60s wind farms are now no longer in existence).

[kristopher]

12. Nuclear power is a mature industry

And the statistics are nuclear industry generated. They have achieve 90%+ for a short time only and there is no evidence from anywhere that this can be maintained. Wind power increases in capacity factor, on the other hand, have resulted from only two things - better siting and wind predictions are one and the other is higher towers for the turbines. With the push for more wind in the past decade the entire country's wind potential has been re-evaluated and a great deal of refined modeling has identified much better wind resources. The situation with wind (especially offshore wind) would be comparable to finding uranium in much higher natural concentrations than has previously been located. The move to 80-100meter towers has gotten the blades into wind that has less drag and is therefore stronger. This means that increases in capacity factor at previous locations is also improved.

So unless the towers shrink there isn't much to decrease the CF for wind.

However the real gains in wind isn't increased capacity factor so much as production per unit (or EROI if you like). The larger turbines produce 5X, 10X even 20X earlier models at the same location. Capacity factor increase is only a marginal contributor to the increased production.

[joshcryer]

15. BTW, a decade is not "a short time only."

[kristopher]

18. That's like saying the weather last year proves climate change is false...

Taking a snip from a larger set of data in order to obtain funding is misrepresentation. There is no reason to expect that the lifetime average of new plants will be 90% plus and every reason to think it will be substantially lower.

Not least among those reasons is going to be the cost of electricity from nuclear compared to renewables. Who are they going to sell their overpriced power to? The forecast by investment firms is for load factors in areas like 56%.



Says Citigroup:

Therefore under very conservative assumptions on renewables, we can reliably expect an extra 330TWh of electricity to be generated by 2020, leaving a shortfall of 16TWh to be made up by either energy efficiency or new nuclear.

There are currently 10GW of nuclear capacity under construction/development, including the UK proposed plants that should be on operation by 2020. If we assume that energy efficiency will not contribute, that would imply a load factor for the plants of 18%. Looking at the entire available nuclear fleet that would imply a load factor of just 76%. We do believe though that steps towards energy efficiency will also be taken, thus the impact on load factors could be larger.

Under a scenario of the renewables target being fully delivered then the load factor for nuclear would fall to 56%.

[joshcryer]

29. *Wrong.* Technology gets better. A car made in 1905 might last 5 years. A car made now...

...might last 15 (if properly maintained).

[joshcryer]

30. Short term trends are bad with climate change because you need very long term trends...

...to derive the data.

[joshcryer]

31. Complete strawman and no comparable in any way whatsoever.

[joshcryer]

33. BTW, citing bankers anywhere else on DU would get you in deep trouble.

[joshcryer]

34. You must provide evidence that capacity factor is decreasing.

[joshcryer]

35. Like the models showing wind will have power output drops as wind variablity decreases due to AGW.

[kristopher]

38. It took you 6 posts to write nothing?

You are special, aren't you?

Taking a snip from a larger set of data in order to obtain funding is misrepresentation. There is no reason to expect that the lifetime average of new plants will be 90% plus and every reason to think it will be substantially lower.

Not least among those reasons is going to be the cost of electricity from nuclear compared to renewables. Who are they going to sell their overpriced power to? The forecast by investment firms is for load factors in areas like 56%.



Says Citigroup:

Therefore under very conservative assumptions on renewables, we can reliably expect an extra 330TWh of electricity to be generated by 2020, leaving a shortfall of 16TWh to be made up by either energy efficiency or new nuclear.

There are currently 10GW of nuclear capacity under construction/development, including the UK proposed plants that should be on operation by 2020. If we assume that energy efficiency will not contribute, that would imply a load factor for the plants of 18%. Looking at the entire available nuclear fleet that would imply a load factor of just 76%. We do believe though that steps towards energy efficiency will also be taken, thus the impact on load factors could be larger.

Under a scenario of the renewables target being fully delivered then the load factor for nuclear would fall to 56%.

[joshcryer]

39. What was written wasn't buried by your attempts to overshadow others posts here.

[Statistical]

20. Only in anti nukker mind is a sample set consisting of 104 reactors for 87,600 hours "small"

104 reactors over a period of 87,600 hours for a combined total operating hours of 9,110,400 operating hours a "short time".


In that "short time" nuclear reactors generated "only" 8400 billion kWh! 8.4 TRILLION kWh.
The total wind power produced in the US since DOE began tracking it is about 300 billion kWh so in this "short time" nuclear energy has delivered almost 30x as much energy as wind has EVER.

Even more interesting is that in the period of time 1998-2008 the number of reactors stayed static however peak output rose from 97.1 GW to 100.3 GW. Due to efficiency improvements peak output rose by about 3%. That is roughly equal to 3 "free" nuclear reactors.

Of course it is only a "small" sample!

[kristopher]

26. When 87,600 hours is part of a 350,000 hour data set it is a "short time".

You use the same deceptive type of data trimming when you claim that the new nuclear fleet will have a lifespan of 60 years. The actual average global lifespan of the 119 reactors that have been placed into service and then decommissioned is 22 years.

[joshcryer]

32. It is fucking 1/4th their lifetime. How is that small?

Is .25 small? Because that's close to the capacity factor of wind. ;)

[Statistical]

36. So when EPA says Prius is 50mpg compared to average car is 32mpg are they lying.

I mean they are only comparing it to cars built this year. They should go back and compare it to cars as far back as 70 years right?

The reality is a 40 year set of data is only applicable if you conclude that
a) No technology improvements have occurs in last 40 years to indicate that NEW reactors will perform similarly to most recent data.

b) New reactors will go back to 9month refuel cycle. Yup earliest reactors refueled every 9 months. Well that right there capped potential capacity at 89% (11% was spent refueling the reactor). Over last 40 years that has improved from 9 months, to 12 months to 15 months, to 18 months and now "most" reactors are on 24 month refuel cycle. The difference between 24 month and 9 month refuel cycle is 6% of potential capacity right there. All new reactor designs are compatible with high burnup to achieve 24 month refueling cycles.

c) No process improvements. Nuclear reactors are simply run better than they were 40 years ago. The cost of power has come down (in real terms) so margins are smaller. A reactor could be profitable with 80% to 85% capacity factor in 1980s can't anymore. Number of unplanned outages has dropped every year for last 20 years (directly related to the continual rise in capacity factor).

So the reality is new reactors aren't being built in a vacuum. They aren't starting from scratch. They will be building on the experiences, knowledge, and improvements of the last 40 years.


Claiming we should use data that is 40 years old is just as disingenuous as claiming wind turbines are only 100-200kw each since that is all that was available 40 years ago.

[kristopher]

37. inappropriate analogy

The same reactors that are now delivering 90% started at 50%. It took thirty plus years to tweak and twist and push them against safety limits to the range they are now operating in.

So to use your Prius analogy we would have to posit a Prius that got 26mpg the first 3 years, 36mpg the second three years, and 46mpg the final three years of its life. We would then claim that this performance in the final stage is the only predictive number that the buyer should consider.

[joshcryer]

40. No, reactor upgrades are effectively rebuilds.

http://paksnuclearpowerplant.com/capacity-upgrade

It'd be like upgrading the battery pack on a Prius. For the longest time I didn't understand why nuclear had the most R&D of all power by and large, it turns out that nuclear reactors are basically rebuilt over time.

[Statistical]

41. Except newer reactor go to 90%+ capacity much faster.

South Texas Project 1 & 2 for example are two of the "youngest" reactors built. Both of them exceeded 90% capacity less than 4 years after first criticality and have been 90%+ for over a decade now.

The level of expertise and performance at newest reactors continue to lead the fleet. In 2009 STP achieved something no other nuclear plant has ever achieved. 5 back to back refueling cycles with no outages (other than planned refueling cycle) for 5 refeuling cycles (10 combined reactor years) the only time the reactor was ever offline was when refueling.

http://www.stpnoc.com/PrRel%2009-12%20B2B%20run%205%20F.doc (word document).


A huge amount of the gains in capacity has been the use of burnable neutron poisons in reactor fuel to achieve higher burnup. Same fuel lasts much longer now, reactors get more energy per ton of fuel, and that means less refueling outages. A refueling outage is 0 output for a month and delays can and do happen. During refueling outage there are a large number of people inside containment. Lots of heavy machinery involves, massive fuel assemblies, and high tonnage cranes being used. Even a simple mechanical accident can delays startup by months.

As a result the industry has pushed heavily to reduce number of refueling outages to do that burnup has gone from 20Gwd/MTU (Gigawatt days of energy per metric ton of uranium) to about 60GWd/MTU. The saving in fuel costs is negligible however the saving in terms of refueling outages is massive.

The fueling "overhead" dropped from about 11% of capacity to 4% of capacity. Every new reactors starts from day one with that "free 7%".

[kristopher]

42. That still doesn't justify the claims being made nor ...

does it address the economic problems expected to impact the capacity factor of future nuclear plants.

The fact is there is not even a hint of certainty behind the inflated predictions of load factors and longevity being hawked by nuclear industry circus sideshow barkers.

You can make a case that this is what you believe will happen, but the routine blanket, unequivocal statements that the US nuclear fleet has a capacity factor above 90% is deceptive at best and flat out lying at worst.

[joshcryer]

43. It does disprove your poor statistic. I hope it is not posted in the future, as it is dishonest.

[kristopher]

44. "My" poor statistic? "Dishonest"?

The number is from Herbst & Hopley's book with the coy antinuclear title of "Nuclear Energy Now: Why the Time Has Come for the World's Most Misunderstood Energy Source".

Perhaps you should email them and let them know that their data is comprised of poor statistics and is dishonest.

If the claims of the nuclear industry were true or had a strong foundation, then there would be plenty of investment money lined up to build the plants. The fact that the money isn't there is a testimonial to the fact that the people who are being asked to lay money on the line DO NOT BELIEVE the claims being made.

It really is that simple.

[joshcryer]

45. New plants are going to be built.

Sorry.

[Dead_Parrot]

19. The first wind turbine was built in 1888

How long, exactly, before it become as mature as nuclear? Another century or two?

[kristopher]

9. US nuclear power industry has a 71% capacity factor.

If one assumes perfect plant components, routine refueling/maintenance, and flawless performance, at best reactors can achieve very-short-term, 90% load factors (1). During the first 30 years of US-commercial-fission experience (beginning in the 1950s), proponents say nuclear-load-factor averages were 50% (2). With more reactors than other nations, the US has 104 plants. Nuclear proponents say their lifetime-load-factor average is 71% (1). UK load factors are similar (3). Only 7 global reactors (1.7% of 414) — mostly those with lax design/standards/enforcement in developing nations—have ever eliminated original ‘‘bugs,’’ then later achieved short-term, 90% load factors (3). Although reactor vendors claim a 79%, global-average-load factor, this figure excludes early-retirement (poorly performing) plants and reactors’ early years of operation (3;1;4).

Rather than 71 or 79%, however, most nuclear-cost studies, like the WNA and MIT analyses assume 85–95% nuclear-load factors, e.g.(5;6;7;8;9;10;11;12;13) — a lifetime-fleet average never achieved by any nation. When the pro-nuclear MIT (see later discussion) and US Nuclear Regulatory Commission studies recently reported US-nuclear-load factors of "about 90%," they admitted this figure covered only the last 5 years, included no new plants, and ignored lifetime-average data and early-shutdown reactors (13), all of which reveal the correct, lifetime-load average to about 70%.

(1) Herbst and Hopley 2007
(2) Sweet 2006
(3) Thomas 2005
(4) International Atomic Energy Agency (IAEA) 2007
(5) World Nuclear Association (WNA) 2005
(6) Scully Capital Services Inc. 2002
(7) Du and Parsons 2009
(8) PB Power 2006
(9) Royal Academy of Engineering 2004
(10) Canadian Energy Research Institute (CERI) 2004;
(11) Tarjanne and Luostarinen 2002;
(12) OXERA 2005;
(13) Deutsch et al. 2009

Climate Change, Nuclear Economics, and Conflicts of Interest
Kristin Shrader-Frechette
Journal of Science and Engineering Ethics

DOI 10.1007/s11948-009-9181-y

[Statistical]

11. Car mileage improves, productivity improves, efficiency improves, and yes capacity factors improve.

Generally speaking we have learned a lot after 40+ years of experience in ANY SECTOR and that includes nuclear power.

Over last decade capacity factor for US nuclear reactors has been 90%+. This is easily verified.

We have amount of power generated by each plant and we know the theoretical max (nameplate rating x time). Capacity factor is very complex or hard to calculate.
.
Annual Capacity Factor = Annual energy output / Annual theoretical max

In 2009 that was 92%. US nuclear reactors delivered 92% of the theoretical energy for the year. This achievement is even more surprising when you consider the average age of US reactors is now 37 years old.

Despite this authors claims MANY reactors in the US achieve 100% uptime other than refeuling outage. With current burnup allowing refueling every 24 months and refueling outages lasting 1 to 1.5 months this allows a potential capacity factor of 94% to 96%.

Some reactors as examples:
http://www.eia.doe.gov/cneaf/nuclear/page/at_a_glance/reactors/southtexas.html

South Texas 1 & 2 combined delivered 21,493 million kWh.
Reporting capacity factor is 96%.

The reactors have a nameplate rating of 2560 MW (1,280 MW each). 2560 MW * 24 * 365 = 22,425 mil kWh.
So a "perfect" plant would deliver 22,425 mil kWh annually. They delivered 21,493 which is a capacity factor of 96% (21,493 / 22,425).


http://www.eia.doe.gov/cneaf/nuclear/page/at_a_glance/reactors/turkeypoint.html
Reactors often will have 100% capacity factor in the "off year" (most reactors refuel 1/3 of fuel every 24 months).
Turkey Point #3 for example had 101% capacity factor for the year. The extra 1% is due to fact that NRC authorized reactors to run at up to 102% of thermal output due to limited fine-tuning of power output in some reactors. Now #4 had lower than average capacity factor of 86% however they still averaged out to 94% (above national average). Utilities attempt to stagger refueling outages one reactor refuels at the midpoint of the other reactors fuel cycle.

Given the amount of energy generated by each plant is available as public record you claim can basically be boiled down to "don't trust your eyes".
The reality is nuclear energy capacity factor is higher than any other form of power and has steadily risen in the United States and around the world.

[kristopher]

13. Over the last decade...

But it took you 40 years to get there and there is NO assurance nor belief that the next round of plants will commence operations with anything like that kind of number.

The capacity factor for the nuclear industry in the United States is 71%.

[joshcryer]

16. It's more that demand pushed technological advancement.

You'll note with any power graph that nuclear improved its output as well as all other technologies over the same period of time.

You cite modern turbines for this, but you somehow think that with nuclear it shouldn't apply and you should average over the lifetime of the technology.

If anyone did that I'd laugh in their face, and rightly so.

Modern nuclear has a 90%+ capacity factor.

[kristopher]

21. No, it doesn't.

For the past 5 years nuclear in the US has managed to achieve 90%+ capacity factor.

Again for the slow learners - nuclear power's performance historically and worldwide contradicts the assertion that a future fleet can be assumed to operate at a 90%+ capacity factor.

To say that they will is simply a sales tactic designed to enhance the economic analysis of nuclear as it attempts to get via crooked politics what it can't get on its economic merits. The investors won't touch it because it is a pis poor financial risk and they are not being fooled by the hype.

[joshcryer]

22. Making stuff up.

Typical for you.

[kristopher]

24. .

If one assumes perfect plant components, routine refueling/maintenance, and flawless performance, at best reactors can achieve very-short-term, 90% load factors (1). During the first 30 years of US-commercial-fission experience (beginning in the 1950s), proponents say nuclear-load-factor averages were 50% (2). With more reactors than other nations, the US has 104 plants. Nuclear proponents say their lifetime-load-factor average is 71% (1). UK load factors are similar (3). Only 7 global reactors (1.7% of 414) — mostly those with lax design/standards/enforcement in developing nations—have ever eliminated original ‘‘bugs,’’ then later achieved short-term, 90% load factors (3). Although reactor vendors claim a 79%, global-average-load factor, this figure excludes early-retirement (poorly performing) plants and reactors’ early years of operation (3;1;4).

Rather than 71 or 79%, however, most nuclear-cost studies, like the WNA and MIT analyses assume 85–95% nuclear-load factors, e.g.(5;6;7;8;9;10;11;12;13) — a lifetime-fleet average never achieved by any nation. When the pro-nuclear MIT (see later discussion) and US Nuclear Regulatory Commission studies recently reported US-nuclear-load factors of "about 90%," they admitted this figure covered only the last 5 years, included no new plants, and ignored lifetime-average data and early-shutdown reactors (13), all of which reveal the correct, lifetime-load average to about 70%.

(1) Herbst and Hopley 2007
(2) Sweet 2006
(3) Thomas 2005
(4) International Atomic Energy Agency (IAEA) 2007
(5) World Nuclear Association (WNA) 2005
(6) Scully Capital Services Inc. 2002
(7) Du and Parsons 2009
(8) PB Power 2006
(9) Royal Academy of Engineering 2004
(10) Canadian Energy Research Institute (CERI) 2004;
(11) Tarjanne and Luostarinen 2002;
(12) OXERA 2005;
(13) Deutsch et al. 2009

Climate Change, Nuclear Economics, and Conflicts of Interest
Kristin Shrader-Frechette
Journal of Science and Engineering Ethics

DOI 10.1007/s11948-009-9181-y

[kristopher]

25. .

If one assumes perfect plant components, routine refueling/maintenance, and flawless performance, at best reactors can achieve very-short-term, 90% load factors (1). During the first 30 years of US-commercial-fission experience (beginning in the 1950s), proponents say nuclear-load-factor averages were 50% (2). With more reactors than other nations, the US has 104 plants. Nuclear proponents say their lifetime-load-factor average is 71% (1). UK load factors are similar (3). Only 7 global reactors (1.7% of 414) — mostly those with lax design/standards/enforcement in developing nations—have ever eliminated original ‘‘bugs,’’ then later achieved short-term, 90% load factors (3). Although reactor vendors claim a 79%, global-average-load factor, this figure excludes early-retirement (poorly performing) plants and reactors’ early years of operation (3;1;4).

Rather than 71 or 79%, however, most nuclear-cost studies, like the WNA and MIT analyses assume 85–95% nuclear-load factors, e.g.(5;6;7;8;9;10;11;12;13) — a lifetime-fleet average never achieved by any nation. When the pro-nuclear MIT (see later discussion) and US Nuclear Regulatory Commission studies recently reported US-nuclear-load factors of "about 90%," they admitted this figure covered only the last 5 years, included no new plants, and ignored lifetime-average data and early-shutdown reactors (13), all of which reveal the correct, lifetime-load average to about 70%.

(1) Herbst and Hopley 2007
(2) Sweet 2006
(3) Thomas 2005
(4) International Atomic Energy Agency (IAEA) 2007
(5) World Nuclear Association (WNA) 2005
(6) Scully Capital Services Inc. 2002
(7) Du and Parsons 2009
(8) PB Power 2006
(9) Royal Academy of Engineering 2004
(10) Canadian Energy Research Institute (CERI) 2004;
(11) Tarjanne and Luostarinen 2002;
(12) OXERA 2005;
(13) Deutsch et al. 2009

Climate Change, Nuclear Economics, and Conflicts of Interest
Kristin Shrader-Frechette
Journal of Science and Engineering Ethics

DOI 10.1007/s11948-009-9181-y

[joshcryer]

27. I think you meant to post that to Neaderland.

[joshcryer]

28. Applying this metric to wind would be lulzy.

[Nederland]

17. Using the same rules

What is the capacity factor of wind over the last 40 years? I'm curious...

[joshcryer]

23. You didn't see the weaseling?

He'll copy and paste it in a second. I already suggested that be done.

[joshcryer]

14. Wait for him to cite modern wind and how it has a better capacity factor.

And then try to take a statistic for old reactors and apply it to new ones.

It's a fun little game he plays.