The Nuclear Power Thread

[Astronuc]

Archive of Previous Symposia
http://www.world-nuclear.org/sym/subindex.htm

The papers are fairly general and deal with the industry, trends, fuel cycle issues, waste and other related topics.

 

[Astronuc]

If one is studying nuclear engineering, or is planning to do so, or planning a career in nuclear engineering, then this is relevant.

http://www.ne.doe.gov/pdfFiles/rpt_NEFutureRequiredRDCapabilities_Sep2008.pdf

Lot's of other good reports here - http://www.ne.doe.gov/


An overview of new and proposed NPPs
http://www.world-nuclear.org/info/inf08.html

 

[oldsloguy]

I would like to add to this. Current thermal-spectrum reactors use MAINLY U-235 in the power production. U-235 is 0.7% of the natural content of uranium on earth. In fact, at high burnup, SOME U-238 (the 99.3% remaining if we neglect some traces) is converted into Pu-239 and is burned up ; about 30% of the energy that is extracted in a reactor comes from this Pu-burning, and 70% comes from the original U-235 burning.

So that means that currently, we use effectively ONE PERCENT of the energetic content of the uranium that has been extracted.

In a fast reactor, we can use ALL of it, because U-238, through conversion in Pu-239, can become a nuclear fuel. We can use all the U-238 that we already DUG UP, and partially discarded (in the "enrichment" of uranium, which is nothing else but removing 3/4 of the U-238 from the original ore), and MOST of the "burned fuel" which consists mainly of passive U-238.

So, by switching to fast breeders, we can extract in principle ONE HUNDRED TIMES MORE ENERGY from the EXISTING waste than we already extracted. In principle without any more uranium input. Just by using the "waste" correctly.

If some powerplants have been working for 30 years, this means, in principle, that we can extract the same power for another 3000 years, just by using its "waste".

Yes, and you have not even mentioned thorium which I understand is 3 to 5 times more plentiful in the Earth's crust than uranium. Is there some reason that you did not mention thorium or were you just addressing uranium issues only!

 

[oldsloguy]

Sodium makes people nervous because of its reactivity with water, but all the other properties of sodium are OK, which makes it less of a problem than people think. For instance, a liquid sodium reactor is NOT under pressure, which relieves a lot of safety, materials and mechanics issues. In that respect, a liquid sodium reactor is "easier" than a LWR which is under high pressure. Also, one can, as with the IFR, use a "buffer bath" of sodium to make the reactor entirely passively safe. The only true engineering challenge is to keep the water out in all circumstances.

I’m not very familiar with hot liquid sodium, but you seem really comfortable with the idea of handling hot liquid sodium in and accident which might expose it to air. Does it not burn very vigorously or is it easily controlled?

 

[oldsloguy]

I believe that the probability of a nuclear accident associated with Nuclear power is low... unfortunately the cost is high. Utility is the product of the probability * the cost. There is a good reason to be cautious about Nuclear power.

I remember reading somewhere in the past that the cost of cleaning and decommissioning TMI was about 900 million dollars and that the initial construction cost was about 4 billion dollars. I'm not sure if those cost were adjusted for inflation over the intervening time differential between them or not, but I don't think it really matters for the purposes of this discussion.

It seems to me that TMI is about the worst possible accident that can happen to a modern LWR. Am I correct in that or can anyone reasonably postulate a worse accident?

If so, agreed that the risk is low, but when you have over 100 commercial plants operating 30+ years each and the worst case accident, which occurs only once over that period, is ¼ the value of one plant, how can you argue that “unfortunately the cost is high”?

 

[joelupchurch]

I remember reading somewhere in the past that the cost of cleaning and decommissioning TMI was about 900 million dollars and that the initial construction cost was about 4 billion dollars. I'm not sure if those cost were adjusted for inflation over the intervening time differential between them or not, but I don't think it really matters for the purposes of this discussion.

It seems to me that TMI is about the worst possible accident that can happen to a modern LWR. Am I correct in that or can anyone reasonably postulate a worse accident?

If so, agreed that the risk is low, but when you have over 100 commercial plants operating 30+ years each and the worst case accident, which occurs only once over that period, is ¼ the value of one plant, how can you argue that “unfortunately the cost is high”?

The construction cost on TMI-2 was 800 million in 1978, which is 2.5 billion in 2007 dollars. Here is a table with construction costs of various reactors adjusted to 2007 dollars:
http://depletedcranium.com/why-i-hate-the-nrc/#more-2748"

Also you should note that the Probabilistic Risk Assessment on the new Westinghouse Ap1000 is hundred times less likely to have a core meltdown than a 2nd generation plant.
http://www.asmeconferences.org/ICONE16/pdfs/NewPlantsBeBuilt.pdf"
The PRA starts on page 23.

 

[oldsloguy]

The construction cost on TMI-2 was 800 million in 1978, which is 2.5 billion in 2007 dollars. Here is a table with construction costs of various reactors adjusted to 2007 dollars:
http://depletedcranium.com/why-i-hate-the-nrc/#more-2748"

Also you should note that the Probabilistic Risk Assessment on the new Westinghouse Ap1000 is hundred times less likely to have a core meltdown than a 2nd generation plant.
http://www.asmeconferences.org/ICONE16/pdfs/NewPlantsBeBuilt.pdf"
The PRA starts on page 23.

Thanks, that is an interesting. TMI-I cost the 400 million dollars and TMI-2 800 million dollars. So, doing the correction more accurately would yield:

Assumptions, using 2007 $:
Cost TMI-2 = 2544 million $
Cost of clean up = $973, over 12 years, use 1985 for adjustment
http://www.ans.org/pi/resources/sptopics/tmi/cleanup.html
Inflation adjustment from 1985 = 1.93
http://www.usinflationcalculator.com/

Clean-up of TMI-2 as a fraction of plant cost = 973*1.93/2544 = 0.74

So, restating my earlier post:

If so, agreed that the risk is low, but when you have over 100 commercial plants operating 30+ years each and the worst case accident, which occurs only once over that period, is 3/4 the value of one plant, how can you argue that “unfortunately the cost is high”?

And as you point out, and my gut feeling is, even that small overhead loss is way over stated.

 

[vanesch]

I’m not very familiar with hot liquid sodium, but you seem really comfortable with the idea of handling hot liquid sodium in and accident which might expose it to air. Does it not burn very vigorously or is it easily controlled?

I think it is the main worry: the confinement has to be double. In a thermal reactor, you want the stuff not to get out in any case, and in a sodium-cooled reactor, on top of that, you don't want water or air to get in in any case. That's why people look into other types of coolant such as liquid lead or gas. But most experience is nevertheless with sodium (and yes, there have been minor problems with it). I think it is the main challenge in the design of a fast reactor. But it is not necessarily so terribly more difficult than a PWR, because there's no pressure.

 

[mheslep]

Special section on in the Sept 8 edition titled "The New Nukes"
http://online.wsj.com/wsjgate?subURI=%2Farticle%2FSB10001424052970204409904574350342705855178-email.html&nonsubURI=%2Farticle_email%2FSB10001424052970204409904574350342705855178-lMyQjAxMDA5MDAwODEwNDgyWj.html"

The article is lengthy, covering many of the topics up thread. To start, I wanted to summarize the various cost figures cited through the article:

• Gen III plants in general, i.e. all designs: $4k to $6.5k per kw
• Small, modular nuclear, i.e. Hyperion or B&W: $3.5k to $5k per kw, add $50 to $100m licensing costs per site.
• Gen IV Ge-Hatachi Prism design: $10k per kw (small size ~300 MW)

Summary of experts quoted in the article:
Revis James, EPRI
Ronaldo Szilard, Idaho National labs
Amir Shakarami, Exelon VP
Tom Cochrane, NRDC

 

[joelupchurch]

Special section on in the Sept 8 edition titled "The New Nukes"
http://online.wsj.com/wsjgate?subURI=%2Farticle%2FSB10001424052970204409904574350342705855178-email.html&nonsubURI=%2Farticle_email%2FSB10001424052970204409904574350342705855178-lMyQjAxMDA5MDAwODEwNDgyWj.html"

The article is lengthy, covering many of the topics up thread. To start, I wanted to summarize the various cost figures cited through the article:

• Gen III plants in general, i.e. all designs: $4k to $6.5k per kw
• Small, modular nuclear, i.e. Hyperion or B&W: $3.5k to $5k per kw, add $50 to $100m licensing costs per site.
• Gen IV Ge-Hatachi Prism design: $10k per kw (small size ~300 MW)

Summary of experts quoted in the article:
Revis James, EPRI
Ronaldo Szilard, Idaho National labs
Amir Shakarami, Exelon VP
Tom Cochrane, NRDC

I read the article and it was pretty good except for the comments by Tom Cochrane. My biggest disagreement would be with implicit assumption that US construction costs are facts of nature rather than products of our regulatory environment. The Chinese are building AP1000 reactors for about $2K per KWh. The World Nuclear Association has better information. http://www.world-nuclear.org/info/inf02.html"

I have been very pleased with the construction updates I've seen from Sanmen. I was inclined to write off the modular design stuff as Westinghouse marketing hype, but the actual results are impressive. As I recall, one of the pictures I saw was the whole control room being lifted in place as a single module. By the time we start building our AP1000s, we will be dealing with a proven design and not have to deal with FOAK issues.

 

[Astronuc]

I read the article and it was pretty good except for the comments by Tom Cochrane. My biggest disagreement would be with implicit assumption that US construction costs are facts of nature rather than products of our regulatory environment. The Chinese are building AP1000 reactors for about $2K per KWh. The World Nuclear Association has better information. http://www.world-nuclear.org/info/inf02.html"

The Chinese didn't have to pay the development costs that Westinghouse did. Westinghouse sold an essentially off-the-shelf design at a relatively huge discount. The Chinese did however buy the first 4 which are now under various stages of construction.

I have been very pleased with the construction updates I've seen from Sanmen. I was inclined to write off the modular design stuff as Westinghouse marketing hype, but the actual results are impressive. As I recall, one of the pictures I saw was the whole control room being lifted in place as a single module. By the time we start building our AP1000s, we will be dealing with a proven design and not have to deal with FOAK issues.

Modular construction is relatively new. Designs like the AP1000 were on the drawing boards before modular construction techniques has matured.

The cost of concrete and steel is a big factor in current capital costs, as well as labor, as is interest. The Chinese government would certainly be more inclined to subsidize NPPs than would the US government.

 

[mheslep]

The Chinese didn't have to pay the development costs that Westinghouse did. Westinghouse sold an essentially off-the-shelf design at a relatively huge discount...

Any idea why Westinghouse would do that? To what end? Are you suggesting that the several AP1000 sites on the NRC proposal list would enjoy Chinese construction costs?

 

[Astronuc]

Any idea why Westinghouse would do that? To what end? Are you suggesting that the several AP1000 sites on the NRC proposal list would enjoy Chinese construction costs?

Money. Short term gain. After W assumed CBS, the nuclear (WN) part got sold to BNFL, who in turn sold WN to a partnership with Toshiba (majority) and Shaw (minority), and some others, IIRC.

The nuclear industry is very competitive, but it is very expensive and the margins are often thin.

 

[mheslep]

Money. Short term gain. ...

I'm referring to your statement "at a relatively huge discount", implying they did it for any other reason but short term money. So 1. Why the huge discount? 2. Can the US get the same deal?

 

[Astronuc]

I'm referring to your statement "at a relatively huge discount", implying they did it for any other reason but short term money. So 1. Why the huge discount? 2. Can the US get the same deal?

From a relative who negotiates gas contracts in China, the Chinese are tough negotiators - and the market is competitive. W competes with AREVA and others in the Chinese market. The W deal also involved technology transfer.

The US market is different. I don't see the same deals being done in the US, because W, AREVA and Mitsuibishi are the primary PWR suppliers - and they can't afford to lose money here.

Besides the US DOE (Uncle Sam) is supposed to kick in some subsidies (direct and indirect).

 

[joelupchurch]

The cost of concrete and steel is a big factor in current capital costs, as well as labor, as is interest. The Chinese government would certainly be more inclined to subsidize NPPs than would the US government.

Actually Westinghouse is also claiming the AP1000 uses a lot less concrete and steel than other designs also. They are claiming 100,000 cubic meters of concrete compared to 520,000 for the Sizewell B reactor. I assume Sizewell B must be a bad design, but they are showing a very small footprint even compared to their own 2nd generation plants. (Page 30-31)

http://amgroupes.fr/admin/compte_rendus/195_compte_rendu.pdf"

The reason that the Chinese got a good deal on the AP1000 is that they ordered a 100 of them. The most any US utility company ordered is 2. I've suggested on my blog that Congress change the charter of the TVA so they can build nuclear power plants anywhere in the United States. Maybe they could get some economies of scale also.

BTW on the question of what government is providing loan guarantees for these reactors, the answer is the United States.

"[URL
/info/inf63.html [/URL]

The US, French and Russian governments were reported to be giving firm support as finance and support arrangements were put in place. The US Export-Import bank approved $5 billion in loan guarantees for the Westinghouse bid

 

[mheslep]

Actually Westinghouse is also claiming the AP1000 uses a lot less concrete and steel than other designs also. They are claiming 100,000 cubic meters of concrete compared to 520,000 for the Sizewell B reactor. I assume Sizewell B must be a bad design, but they are showing a very small footprint even compared to their own 2nd generation plants. (Page 30-31)

http://amgroupes.fr/admin/compte_rendus/195_compte_rendu.pdf"

The reason that the Chinese got a good deal on the AP1000 is that they ordered a 100 of them. The most any US utility company ordered is 2.

I don't think 'ordered' is the correct word, as the implies a contract to build, and I'm unaware of a final go ahead on any US nuclear units. Design and services contracts have no doubt been placed.

Where did you get the number 100 for China?

 

[Astronuc]

The reason that the Chinese got a good deal on the AP1000 is that they ordered a 100 of them. The most any US utility company ordered is 2. I've suggested on my blog that Congress change the charter of the TVA so they can build nuclear power plants anywhere in the United States. Maybe they could get some economies of scale also.

Please cite sources. Checking the World Nuclear link - China has plans for about 100 1 GWe units - not all of which are AP1000. And China negotiated a technology transfer program, so after the first 4, CNNC (and perhaps CGNPC and China Power Investment Corporation (CPI)) will likely build most of them. They are also contemplating CPR-1000 and EPRs.

Sizewell B was based on standard 4-loop Westinghouse design like Wolf Creek or Callaway. The essentially used the requirements of the time, given the limited experience (less than 20 years.) Now with 40+ years experience, the plant and reactor designs can be optimized.

 

[joelupchurch]

I don't think 'ordered' is the correct word, as the implies a contract to build, and I'm unaware of a final go ahead on any US nuclear units. Design and services contracts have no doubt been placed.

Where did you get the number 100 for China?

That is what it says in Wikipedia. http://en.wikipedia.org/wiki/AP1000"

I checked the reference and the Westinghouse CEO did say that:
http://www.pittsburghlive.com/x/pittsburghtrib/s_575073.html"

I also found an article repeating what Wikipedia said about the AP1000 being the standard for Inland nuclear power plants.
http://www.neimagazine.com/story.asp?storyCode=2053048"

I would interpret that to mean that they will use something else for plants that don't require a cooling tower.

US AP1000 Orders.

http://www.nucpros.com/?q=node/4313"
http://www.world-nuclear-news.org/newsarticle.aspx?id=24250"

 

[Astronuc]

The AP-1000 is a simplified design with a simplified ECCS. It will require cooling towers in some areas where the site would not have adequate river or ocean volumes.

See also
Westinghouse, Shaw to provide four reactors to China
http://www.neimagazine.com/story.asp?storyCode=2046380

from the Candris article:

Last year the company beat out French rival Areva to win a $5.3 billion contract to build four AP1000s in China. Although Westinghouse will transfer the technology to Chinese licensees over the next few years, Candris said, it will build several additional plants with partner The Shaw Group, of Baton Rouge, La.

http://www.pittsburghlive.com/x/pittsburghtrib/s_575073.html

The AP1000 has a core size of 157 17x17 assemblies with a 14-ft (4.27 m) active fuel length (core height) with a thermal output of ~3400 MWt, which is about the same thermal output from a standard 4-loop 193-assembly core with a core height of 12 ft (3.66 m) before they were uprated. The fuel rods in the AP-1000 uses a 9.5 mm cladding OD.

 

[mheslep]

I checked the reference and the Westinghouse CEO did say that:
http://www.pittsburghlive.com/x/pittsburghtrib/s_575073.html" ..

I only check this reference - it says China wants 100, not bought. I've never gotten a volume discount for wanting anything (so far)

Regards the US plants, even Vogle in Georgia is bit early to call an order. The utility and the state utility commission have given the go ahead, but the NRC doesn't give final signoff until 2011, if then. So ground breaking is several years away, if if happens.
http://www.nrc.gov/reactors/new-reactors/col/vogtle/review-schedule.html
I hope it does go ahead, but I expect this administration won't let it happen - not in 2011 anyway.

 

[joelupchurch]

I only check this reference - it says China wants 100, not bought. I've never gotten a volume discount for wanting anything (so far)

Regards the US plants, even Vogle in Georgia is bit early to call an order. The utility and the state utility commission have given the go ahead, but the NRC doesn't give final signoff until 2011, if then. So ground breaking is several years away, if if happens.
http://www.nrc.gov/reactors/new-reactors/col/vogtle/review-schedule.html
I hope it does go ahead, but I expect this administration won't let it happen - not in 2011 anyway.

If the Chinese build 100 AP1000 plants, I'm not sure if Westinghouse cares if they see much money from them. There is a whole international supply chain being built for the AP1000. This gives them a lot of leverage when they are bidding against Areva and GE. That's assuming that Westinghouse only licensed the design for construction in China.

The second point is probably valid. The NRC just announced a delay in licensing the AP1000 because Westinghouse hadn't answered all their questions about the sump design.

http://greeninc.blogs.nytimes.com/2009/09/09/a-nuclear-renaissance-stumbles-forward/"

 

[Equate]

Am I Reading This Right?

DOE Has $40M for Design and Development of Next Gen Nuclear Plants

The $40 million funding announcement made today will support phase one activities including the development of cost-shared conceptual designs, cost and schedule estimates and a business plan for integrating Phase 2 activities. The data gathered in Phase 1 will be used to determine if Phase 2 should continue. Applications for receiving funds from the $40 million are due by November 16 and the DOE expects to make two awards in February 2010 with each supporting a unique reactor concept.

A demonstration plant is expected to be produced by 2021.

Source: http://www.dailytech.com/article.aspx?newsid=16304

$40 million to be split in half for two separate research projects? They are spending billions to bail out dubious investors and then they are giving out jump change for the development of NGNPs?

That's a slap in the face...

 

[tmyer2107]

I was wondering what everyone's thoughts were on gas turbine modular helium reactors (GT-MHR). I recently came across the technology and found it very interesting but wasn't able to find many details on it. Some details can be found here, http://gt-mhr.ga.com/" . How does this technology compare to AP1000 reactors?

 

[Astronuc]

Some info on GA's MHR.

http://web.gat.com/pubs-ext/MISCONF06/A25381.pdf


Some info on GT-MHR:
http://ocw.mit.edu/NR/rdonlyres/0863E2E6-0F70-4843-B2C0-541EC6CD8F59/0/gtmhr.pdf

http://www.world-nuclear.org/sym/1997/labar.htm

http://txspace.tamu.edu/bitstream/handle/1969/1531/etd-tamu-2004C-NUEN-Cocheme.pdf?sequence=1 (particulars on core and fuel are on pages 38-46)


Operational Parameters on the Gen IV Gas-Cooled Fast Reactor

Reactor power                      600 MWth
Net plant efficiency               48% (direct cycle helium)
                                    
Coolant inlet/outlet temperature   490°C/850°C
  and pressure                     90 bar
Average power density              100 MWth/m3
Reference fuel compound            UPuC/SiC (70/30%)
                                    with about 20% Pu content
Volume fraction,Fuel/Gas/SiC       50/40/10%
Conversion ratio                   Self-sufficient
Burnup, Damage                     5% FIMA; 60 dpa

Ref: http://gif.inel.gov/roadmap/pdfs/gen_iv_roadmap.pdf
5% FIMA is about 51-52 GWd/tHM, which is comparable to discharge burnups in modern LWRs.

Fast reactors can achieve much higher burnup ~100 GWd/tHM, or about 10% FIMA.

The GT-MHR unit would produce 288 MWe based on 0.48 * 600 MWt.


In contrast, the AP-1000 produces ~3400 MWt (~1100 MWe) using 157 fuel assemblies. The fuel is a 14-ft (4.27 m) 17x17 fuel design, which is typical of STNP or French REP1300 or N4 reactor fuel designs.

The AP-1000 produces approximately the same power (~3411 MWt, without uprate) as a standard Westinghouse 4-loop 17x17 plant with a 12-ft core of 193 assemblies.

 

[mheslep]

I was wondering what everyone's thoughts were on gas turbine modular helium reactors (GT-MHR). I recently came across the technology and found it very interesting but wasn't able to find many details on it. Some details can be found here, http://gt-mhr.ga.com/" . How does this technology compare to AP1000 reactors?

Reading the GA material suggests the major difference is simply the He cooling instead of water, which provides:
o high temperature operation and a Brayton cycle
o no corrosion potential from the coolant
o no phase change (water - steam - water) complications, reducing the size/cost of the balance of plant.
I can't find a reference, but I expect the fuel is still enriched U. Graphite encased.

 

[mike232]

Just to add this, I don't know if it is true but I heard that France was working on a way to reactivate, and reuse the nuclear waste that is produced by the nclear plants.

 

[mheslep]

Just to add this, I don't know if it is true but I heard that France was working on a way to reactivate, and reuse the nuclear waste that is produced by the nclear plants.

Not working-on, but have-been-doing - as a major operation, the largest in the world since 1976.
http://en.wikipedia.org/wiki/COGEMA_La_Hague_site

http://en.wikipedia.org/wiki/Nuclear_reprocessing

 

[Astronuc]

Forbes - Where The Jobs Are: Nuclear Plant Work

Atomic power is coming back, and so are the jobs to make it happen.

http://www.forbes.com/2010/02/12/nuclear-power-jobs-leadership-careers-employment.html

America hasn't built a new nuclear plant in three decades. That's about to change. With 28 license applications pending at the Nuclear Regulatory Commission and construction likely to begin on at least four plants within the next five years, companies like Westinghouse, General Electric, Bechtel, Areva, URS and the Shaw Group are staffing up, according to Edward Quinn, a past president of the American Nuclear Society, a nuclear power industry group.

The jobs aren't just for nuclear engineers. In fact, only 5% to 10% of the employees who work on a plant hold a nuclear engineering degree, says Quinn. Those specialists work on the core, which uses nuclear technology. There are plenty of other types of engineers employed at nuclear plants and the companies that service, design and construct them, including civil, mechanical and electrical engineers. Beyond engineers, there are a slew of jobs for workers in the construction trades, from welders to grinders. There's also plenty to do for electricians, pipe fitters, iron workers, carpenters and boiler makers.

. . . .

 

[mheslep]

Forbes - Where The Jobs Are: Nuclear Plant Work

Atomic power is coming back, and so are the jobs to make it happen.

http://www.forbes.com/2010/02/12/nuclear-power-jobs-leadership-careers-employment.html

Yes pending at the Nuclear Regulatory Commission, for several years now, and pending, and pending. I've yet to see anything indicating the NRC's Jaczko is going to actually give a go ahead on any of them.