Friday, September 2, 2011

Energy from Thorium's Recent LFTR Cost Thread

A recent Energy fromThorium discussion has once again brought to the fore questions about MSR/LFTR costs. "Zeropoint" kicked off the discussion:
I’m a LFTR n00b so forgive me if this topic has been addressed many times. Also please point me to any threads that already talk about this. I did a search for the other threads but I didn’t see the information I wanted.
I am trying to understand the economics of LFTR.
I am taking my data from the OECD/IEA “Projected Costs of Generating Electricity” report 2010 edition.
Using that data I wanted to know estimate LFTR overnight costs (not taking into account the time value of money) and operating costs.
For the current costs of nuclear, coal and gas plants I went to tables 3.7a, 3.7b, 3.7c in the report. I dropped the two highest and lowest plants as well as any others that looked out of place. I then use the extremely scientific method of averaging them. My results are in the attachment below.
(I cannot reproduce Zeropoint's attachment. Readers will jave to look at the original comment.)
Overnight Costs of LFTR
I figure the theoretical lowest cost for LFTR is the overnight cost of a gas plant since LFTR could be interpreted as a nuclear heat powered gas turbine. Since coal and nuclear have a less efficient conversion efficiency and higher capital costs due to the steam cycle, I am going to make another scientific assumption and assume that the difference between the costs for Coal and Gas is due to the capital cost of the water condenser system, the more efficient conversion, and the efficiency of having the gas turbine factory made which you drop in place. That is a different of $1475. The cost going from the coal plant to the nuclear plant is then the cost of the containment, the nuclear reactor and the redundant systems. That is a difference of $1875. Assuming that the containment is only a third of the 1875, it means that LFTR could be built for
LFTR_Overnight_Cost = 1028 + 1875 * 2/3 = 2258 USD/kWe
I am surprised it is so close the cost of a coal plant. I thought it would be a little higher.
Fuel Cycle
I am going to assume 0.1
Operation & Maintenance
You may be able to have less operators at a LFTR plant than a traditional nuclear plant but the maintenance costs will be higher due to the replacement of the graphite core. I am going to assume that O&M is the same as traditional nuclear, 13.6 MWh.
Construction Time
If LFTR components are factory made and then plumbed up on site then you can probably get the same construction time as a coal plant. Not sure how long the regulatory piece takes. Assume 4-5 years with one of those years being regulatory.
Plant Lifetime
This I have no clue on this. Any guesses?
Decommissioning Costs
Will probably be the same as traditional nuclear at 15%.
Any thoughts?
At this point I would note that Nuclear Green makes a number of assumptions that differ from those offered by "zeropoint". Nuclear Green substitute assumptions are intended too lower MSRs/LFTRs costs, to decrease MSRs/LFTRs construction time, and to increase the speed and number of MSRs/LFTRs deployed before 2050. Thus from the viewpoint of Nuclear Green, Zeropoint makes very conservative assumptions.

"Lars" responded to "Zeropoint,"
It is tough to get a good cost on current nuclear power plants - just look at the variance in actual costs to build existing reactors. The estimates have generally been even to 30% cheaper than LWRs but the estimates are very old.
If LFTR takes off like I think it should then there will be a substantial learning curve advantage to factor in - especially in the all critical time to build.

Here are the reasons we think it should cost significantly less.

Compared to the old LWRs new ones have increased in price dramatically due in large part to redundant engineered safety systems and schedule delaying tactics by opposition intended to drive up the cost. LFTR has its safety due to physics so it is reasonable to hope that the cost will not go up.

There is no high pressure system in LFTR so we don't need the super thick, 600 ton pressure vessel that can only be made in Japan today. In fact, there is reason to think that most of LFTR could be factory assembled and shipped by barge, or truck to the site reducing construction time (similar things are being done for future LWRs too).

Dry or Wet/Dry cooling could allow placing the reactors away from rivers/lakes/oceans generating less opposition and less unique environmental impact reports.

Counter to these nice things is that the reactor is fundamentally different and regulators won't know what to do with it. These days they tend to over-regulate. Some things that were allowed and have been grandfathered in for LWRs likely won't be tolerated for LFTRs.

But - the majority of LFTRs won't be installed in the US or EU so our over-regulation will matter less.

For plant life-time, I'd guess this doesn't matter too much but the general target for modern power plants is 60 years.
Lindsey commented,
(My) gut feel says 2,000 - 3,000/kW for a 400 MW+ sized plant, but my capital cost estimates say for a simple graphite free tank type core $1,350/kWe all in, and that sounds too cheap to me.
Lindsey's too cheap comment is all too familiar to be. Three years ago, I attempted to estimate LFTR costs, using the small factory manufactured LFTR mode; and a number of different assumptions, and kept getting estimates that struck me as far too low.

Lindsey also offered construction time estimates that were far more conservative than those which I believe are both desirable and obtainable.
For standardized designs the equipment manufacture if made to order would be about 10 -18 months depending on hardware used, on-site construction time could be 12 - 18 months, followed by 6 - 12 months commissioning and testing, so all of that together get's you out to 4 years approx from financial investment decision (FID)
Lindsey is thinking in terms of traditional reactor construction methods, but I have reason to believe that research aimed that drastically shortening small reactor field set up times may be underway.

"Ida-Russkie" commented,
The NRC hearing process for the AREVA eagle rock plant is set to last two years by law. Is a reactor going to be easier to get approval? So, this is one area where there should be some relief after you prove one can be built. the hearing process is to stop the design changes which bankrupted some nuclear plant builds in the past. The environmental impact study took one year to submit.
"Ida-Russkie's" comment clearly assumes a businesses as usual operation for the NRC, however, the potential inherent safety of MSRs may lead to big changes in the regulatory process, and the recycling of coal fired steam plant locations, may drastically change the state permit situation, since permits have already been issued for coal fired power plant sites.

"Cyril R" pointed out one of the reasons for low MSR costs,
different reactor builders are using different numbers of loops. Westinghouse seems to prefer a few big loops for the AP1000. Areva likes to use one more loop. 4 for the EPR, this is 1600/4= 400 MWe per loop compared to 550 MWe for the AP1000. It seems plausible that the loops should be as big as possible for economics. But the LFTR has a remote maintenance requirement. Multiple smaller loops could allow easy modular replacement. The LFTR has less pumping power than a PWR. On the order of 7-8x less I believe, based on the AHTR pumping power requirement of 1.46 kW/MWth and 8 kW/MWth for the EPR, combined with better turbine efficiency, gives around 7-8x less pumping power per kWe. It could be even smaller than that, as the temp drop is larger for LFTR, and this likely more than compensates the heavier fuel salt pumping requirement.
Pumps are important and usually expensive components of reactor design. The EfT discussion makes clear that Westinghouse has chosen to lowe its pump costs by decreasing their number, with each of 2 pumps moving half of the AP-1000 reactor coolant.

"Zeropoint" calls attention to the cost of regulation,
Sorry to dwell on the subject of the NRC which I know all of you love. I wanted to understand the approval process a little better since I didn't have it in my cost estimates.

I just listened to Atomic Rod's podcast #154 "Atomic Round-up With Five Experts" where he mentions that for new nuclear plant designs you have to pay the NRC $200+ an hour for them to analyze and authorize new designs.
Wow! That is going to be a drain on any startup especially for a LFTR based design where the NRC has to be educated on the technology. Is this still the case (the podcast was done a year ago)?

At $250 an hour, one man year costs $500,000 ($250*40hrs*50weeks/yr). Does anyone know how large an NRC team would be to analyze new designs? If the team was 10 people for 10 years, then that is $50mm to get a design approved. That is a pretty good size chunk of change.
Lars estimates that the cost of NRC regulatory approval runs about $50,000,000. With the Cost of LFTR design runint to $500,000,000. Of course in Energy circles $500,000,000 does not amount to a lot. Solyndra, a Solar PV systems manufacturer, has just filed for Chapter 11 bankruptcy. Solyndra received over $500,000,000 in Federal loan guarantees in 2009 in addition to nearly One Billion Bucks from venture capital firms. Well, the chumps who just parted with one and a half billion bucks were warned. Warned by Nuclear Green, and Brave New Climate.

Cyril R, comments
The purpose of the NRC is to maintain the status quo on nuclear. Molten fuel reactors could potentially break the status quo, so the NRC won't license it in our lifetimes. However, they will gladly take all your money to research it to death, and laugh all the way to the bank.

Its like, asking the king to help start a revolution for democracy.

http://ergosphere.blogspot.com/2011/07/quote-without-comment.html

MSRs have the real potential to work, and work much better than LWRs, so the NRC won't help you.
Cyril's link points to a quote from Ugo Bardi
Bureaucracy is a tool to keep the world as it is, not to change it. So, in perfect Tainter-style, the system works hard to avoid innovation, not to promote it. It is almost impossible to be financed to study resource depletion; that would highlight problems that would require changes and that's a no-no. Instead, it is still possible to obtain research grants as long as there is no risk that the results will threaten the status quo. Hydrogen as a fuel is a good example. It is high-tech, fashionable, sophisticated, popular, environmentally friendly, and it doesn't work. This last characteristic makes sure that its development will bring no changes whatsoever.
Several issues touched on in the LFTR cost thread invite Nuclear Green posts, but one by Chemical Engineer Kim L Johnson is down right pregnant. Johnson writes,
I'm a chemical engineer who has been working to develop industrial Fluorides for many years and have assembled lots of online documentation for the benefit of Lftr and the like.

If you have any significant interest in Fluorides, structural materials for Fs and in which forms of whatever elements are best for the Lftr bath & containment, I would be happy to send you lots of link.

Sadly however, I can no longer post important details Freely. Serious Foreign competition could very well, in a few years' time, leave the US so far behind in our own Fluoride-Energy tech we'd never recover economically.
Fortunately, the very successful efforts of TEA & other "T Com" leaders -- fruit we shall soon see in the Senate & elsewhere -- should shortly enable guys like you, (hopefully) Lars, and many others to develop Thorium-enabled technologies full time (*just* getting off the phone with this effort's champion) !
MSR/LFTR endeavors are being launched. Some are so clouded in secrecy that their very existence may be unknown to me at this point. Lots of things are going on that I can only guess at. Part of this secrecy is due to the entry of the Chinese into the MSR/LFTR field, and part is due to proprietary concerns.

The EfT discussion suggested that the Nuclear Green view that MSR/LFTR technology can lower nuclear costs, is shared by engineers and scientists who are aware of that technology. It also suggests that other members of the EfT community may be more wed to business as usual assumption than Nuclear Green is. However, the discussion points to concerns about Federal regulation as an barrier to technological advancement that require further attention.

The EfT LFTR cost thread makes conservative assumptions, but still suggest that LFTR costs may be so low that at least one discussion used the words too low to make sure we knew he is sane. At the very least, it can be concluded from the EfT discussion that MSR and LFTR technologies may be a road to lowering the cost of nuclear power.

8 comments:

gallopingcamel said...

Kirk Sorensen has said that MSRs such as the LFTR are not going to be cheap.

It is not at all obvious to me why this should be so, especially when one realises that LFTRs are scalable so one could mass produce small units (e.g. 100 MWe) in factories and deliver them to site on trucks.

Kirk himself suggested building MSRs into ships that could travel to wherever they were needed. Remember that we could produce two 10,000+ "Liberty Ships" each day during WWII and that was only a small part of our manufacturing capacity.

Charles Barton said...

gallopingcamel, when you name your company after an expensive salt formula, you are not giving priority to low cost. David LeBlanc, pm th other hand, does think about lowering costs.

Anon said...

Cheap is relative. I think it a reasonable expectation that a Molten Salt Reactor can come in under a coal plant for life cycle costs and likely capital costs even in areas with the cheapest coal.

But even that is still going to be pretty expensive (they will cost millions at the very least).

gallopingcamel said...

Charles,
How much does FLIBE cost and how much does one need per MWe of generating capacity?

Charles Barton said...

ORNL-TM-2606-12 offers a discussion of salt costs.
http://www.energyfromthorium.com/pdf/ORNL-TM-2006-12.pdf

Part of the Discussion was based on a 1971 document

In 1971, LiF-BeF2 was estimated to cost $26.30 per kg, while NaF-KF-ZrF4 cost $4.60 per kg. The 2006 price of Derived fluorides ($/kg-fluoride) was $63.54 per kg of LiF and $117 per kg of BeF2. This does not tell the whole of the cost story. It is nor clear if the 2006 ORNL study included the cost of separating Li-6 from Li-7., but in 1971 that raised the LiF costs to $120 per kg. In contrast the cost of NaF ran in 2002 to $1.32 owe kg. It is calculated that if LiF-BeF2 salt is used, the cost of the salt will be 10% of the total capital costs. Quite obviously if NaF can be substituted for BeF2, you are going to save a lot of money.

gallopingcamel said...

Charles,
Thanks for that very interesting information. Probably you can point me to a paper that explains the pros and cons for the salts that you mention.

My experience with Be is not a happy one; I still have serious concerns about its toxicity as well as its cost. Presumably NaF has some drawbacks or Kirk would have called his company "NAF" instead of "FLIBE".

Looking at the amount of salt required per MWe, I imagine that the core geometry is the critical issue with spherical geometry leading to the smallest possible core volume and cylindrical geometry (LeBlanc) being the next best.

Charles Barton said...

GC, ORNL/TM-2006/12 (Assessment of Candidate Molten Salt Coolants for the Advanced High-Temperature Reactor (AHTR)) would probably be a good starting point. Yiy can fund it in Kirk Sorensen's document archive. It also contains an extensive bibliography, that would probably be helpful.

Typong in the words Molten Salt Chemistry brings up 137 pages of entries on an Information Bridge Search.

gallopingcamel said...

Thanks for the ORNL link that I have saved for future reference.

It shows that salts based on Lithium, Beryllium and Fluorine have the advantage of low thermal neutron absorption cross-sections. However, the paper related to materials for use in cooling loops so there was no discussion of solubility issues that are important in LFTRs. How much Thorium can be dissolved in NaF compared to Flibe or other salts?

My apologies for being such an ignoramus but my field is electro-optics rather than nuclear physics:
http://www.bdidatalynk.com/PeterMorcombe.html

I am a trained "Radiation Worker" with experience of managing personnel protection systems relating to free electron lasers and high intensity gamma ray sources.

Followers

Blog Archive

Some neat videos

Nuclear Advocacy Webring
Ring Owner: Nuclear is Our Future Site: Nuclear is Our Future
Free Site Ring from Bravenet Free Site Ring from Bravenet Free Site Ring from Bravenet Free Site Ring from Bravenet Free Site Ring from Bravenet
Get Your Free Web Ring
by Bravenet.com
Dr. Joe Bonometti speaking on thorium/LFTR technology at Georgia Tech David LeBlanc on LFTR/MSR technology Robert Hargraves on AIM High