Are small modular nuclear power plants the solution for remote locations?

[jal]

I found the following
note: Moderators
I would like to know the opinions of experts from here before you move this thread to another forum.
Is this all hyp?
http://www.hyperionpowergeneration.com/
Home page of Hyperion power generation

The Hyperion team is committed to helping make the clean and safe benefits of nuclear power — benefits that could assist in solving the worst of our planet’s problems — available in even the most remote locations. We hope you will enjoy learning about Hyperion through our web site!

Perfect for moderately-sized projects, Hyperion produces only 25 MWe — enough to provide electricity for about 20,000 average American sized homes or its industrial equivalent. Ganged or teamed together, the modules can produce even more consistent energy for larger projects.

Invented at the famed Los Alamos National Laboratory, Hyperion small modular power reactors make all the benefits of safe, clean nuclear power available for remote locations. For both industrial and community applications, Hyperion offers reliable energy with no greenhouse gas emissions. Hyperion power is also cheaper than fossil fuels and, when you consider the cost of land and materials, watt to watt, Hyperion’s innovative energy technology is even more affordable than many developing “alternative” energy technologies.

I have just read the thread Small nuclear power plants. that showed up after posting this thread.
jal

 

[Astronuc]

It's not hype. Hyperion and others are serious about small, compact and self-contained nuclear systems. This is a commercial approach to what the Navy does for nuclear reactors in submarines and surface ships - more or less - in the sense that the core is not refueled periodically, e.g. annual, sesquiannual, or biannual basis. Instead the core operates for a life-time without refueling, after which it is removed and replaced. Traditionally NPP's have been designed for 40 years, with the initial capital costs retired by that time. Now plants are having their lives extended to 60 yrs and there is a new effort to push 80 years, although O&M costs will likely increase, or new capital costs incurred as some large components are replaced.

 

[mheslep]

A search turned up this thread, and I see recently Hyperion identified a buyer for unit, so I'll append here, as I have a feasibility interest for portable reactors.

Apparently the Hyperion unit design is 25-30 MW(e), 70MW(t). My question: how is it possible to effectively reject the 40-45 MW(t) of heat without a large water source, or absent that an extremely large heat exchanger. A portable reactor can have neither.
Edit: This design uses a Brayton cycle.

http://technorati.com/posts/HK3w8tp1oLKudb99ak4pTORJxp%2Ba_rsUr_0TPMtmQQs%3D"

 

[Astronuc]

They could possibly do a Brayton cycle, or a combined Brayton-Rankine cycle.

And perhaps they plant to sell the rejected heat as 'process heat' or district heating.


Some nuclear and conventional plants use dry forced-convection cooling systems.

Using geothermal heat sink may be an alternative.

 

[mheslep]

They could possibly do a Brayton cycle, or a combined Brayton-Rankine cycle.

And perhaps they plant to sell the rejected heat as 'process heat' or district heating.Some nuclear and conventional plants use dry forced-convection cooling systems.

Using geothermal heat sink may be an alternative.

Yes http://en.wikipedia.org/wiki/Hyperion_Power_Generation#Design_description".

Edit: From http://www.world-nuclear.org/info/inf33.html" r

The Hyperion Power Module (HPM) is a small self-regulating hydrogen-moderated and potassium-cooled reactor producing 70 MWt /25 MWe fuelled by powdered uranium hydride. It operates at about 550C and is designed to operate for 5 - 10 years before being returned to the factory for refuelling. It is about 1.5 metres wide and 2 metres high, so easily portable, it is sealed and has no moving parts. Hyperion Power Generation has had preliminary discussions with the Nuclear Regulatory Commission and a US design certification application is possible in 2012, when the company plans to begin manufacturing the plants in New Mexico. The design is licensed from the DOE Los Alamos laboratory there. The company reported sales interest from Eastern Europe in August 2008, at $27 million per unit.

The uranium hydride fuel in this is a powder and incorporates the hydrogen moderator. However, above 550C the UH3 of the fuel dissociates and the resulting decrease in moderator density reduces the core reactivity. As it cools, hydrogen is reabsorbed, increasing the core reactivity. All this is without much temperature change since the main energy gain or loss is involved in phase change. Outside the core is storage for hydrogen at a controlled temperature, which determines the hydrogen pressure in the whole system and the precise core temperature (increasing with higher pressure). The whole system thus becomes self-regulating, and is inherently safe. Enrichment level starts off at about 5% and over five years is burned down to about 3%, with criticality being maintained by steadily reducing the proportion of deuterium in the hydrogen.

 

[Astronuc]

The Wikipedia article and the press surely provide lofty praise and expectations for the Hyerpion modular power unit, but I'd imagine that it's going to take a couple of years to get through the NRC process.

I'd be interested in testing of the UH3 concept in their core configuration. It sounds like there is a hydrogen feed and bleed system (H2, D2), which also must deal with tritium. I am also curious how they handle the noble radioisotopes of Xe and Kr, and the volatiles such as Cs, I and Br. And I have to wonder about Xe-transients/swings.

 

[mheslep]

The Wikipedia article and the press surely provide lofty praise and expectations for the Hyerpion modular power unit, but I'd imagine that it's going to take a couple of years to get through the NRC process. ...

Yes here's NRC's web status on them
http://www.nrc.gov/reactors/advanced/hyperion.html

Status/Other Info: A conceptual design. NRC has had limited interactions with Hyperion and is awaiting further design work before scheduling pre-application meetings.

I saw elsewhere a public NRC comment that was disappointing. It read, paraphrasing, that, well, the Hyperion design was unusual, different, so they were never going to touch it. So I expect that their sales will all be international, or to the military.

 

[Astronuc]

To pass muster with the NRC, I would expect that Hyperion would have to build a demonstrable module, and put in a reactor like ATR to demonstrate the concept. It is radially different than current reactor designs.

The NRC has difficulty with just the conventional Gen 3+ LWRs, so I imagine they'd be reluctant to take on the Hyperion UH₃ system at the moment.

 

[mheslep]

I'd be interested in testing of the UH3 concept in their core configuration. It sounds like there is a hydrogen feed and bleed system (H2, D2), which also must deal with tritium. I am also curious how they handle the noble radioisotopes of Xe and Kr, and the volatiles such as Cs, I and Br. And I have to wonder about Xe-transients/swings.

Sorry I don't follow the tritium source. Aside from the standard in 0.01% tritium production in general fission, how does the introduction of H2 or D2 increase that in any way? Same goes for the other radionulclides. Why would you expect they would be handled in a different waste process than is used with removal in PWR?

 

[Astronuc]

Sorry I don't follow the tritium source. Aside from the standard in 0.01% tritium production in general fission, how does the introduction of H2 or D2 increase that in any way? Same goes for the other radionulclides. Why would you expect they would be handled in a different waste process than is used with removal in PWR?

d + n => t.

In conventional LWR fuel, the Xe, Kr, Cs, I mostly accumulate in the UO2 matrix. Some of the gases and volatiles (~1% - ~20% under normal operation) find their way to the void volume. The void volume and fuel are surrounded by cladding, which ideally retains its hermiticity.

The UH₃ can volatize, and the H can react with other metals like Zr (also a fission and decay product). Part of the NRC's requirement would be a demonstration of the fuel behavior to maximum design burnup, and a demonstration that Hyperion scientists/engineers can predict the thermo-mechanical and neutronic/reactivity behavior of the UH₃ fuel system.

 

[mheslep]

d + n => t.

Arg, yes of course, missed that, though the D cross section is small.

In conventional LWR fuel, the Xe, Kr, Cs, I mostly accumulate in the UO2 matrix. Some of the gases and volatiles (~1% - ~20% under normal operation) find their way to the void volume. The void volume and fuel are surrounded by cladding, which ideally retains its hermiticity.

The UH₃ can volatize, and the H can react with other metals like Zr (also a fission and decay product). Part of the NRC's requirement would be a demonstration of the fuel behavior to maximum design burnup, and a demonstration that Hyperion scientists/engineers can predict the thermo-mechanical and neutronic/reactivity behavior of the UH₃ fuel system.

Well I believe the idea is that since they have need no control rod mechanicals, their containment mechanicals are much simplified.

 

[Astronuc]

I was just talking with a colleague about the NRC's thoughts on the Hyperion system. Apparently, Hyperion has talked with the NRC, but has only provide 'design on paper' details. The NRC is not going to review it unless the folks at Hyperion provide experimental data, and demonstrate that they can simulate the physics of the reactor and fuel.

Hyperion, like the Toshiba4S and BWXT's design, are 'low capacity' systems. Given that the NRC's experience is LWR technology, they are not about to seriously considered alternative systems without a lot of experimental and analytical work. The NRC would also need to hire staff members with experience in such technology (that may be a kind of Catch 22).

Apparently there is also an ANL/LLNL design.

 

[mheslep]

I was just talking with a colleague about the NRC's thoughts on the Hyperion system. Apparently, Hyperion has talked with the NRC, but has only provide 'design on paper' details. The NRC is not going to review it unless the folks at Hyperion provide experimental data, and demonstrate that they can simulate the physics of the reactor and fuel.

Yes the NRC states as much https://www.physicsforums.com/showpost.php?p=2273870&postcount=7"

Hyperion, like the Toshiba4S and BWXT's design, are 'low capacity' systems. Given that the NRC's experience is LWR technology, they are not about to seriously considered alternative systems without a lot of experimental and analytical work. The NRC would also need to hire staff members with experience in such technology (that may be a kind of Catch 22)...

That's the trap: The NRC doesn't want to be forced to look at this technology. The best face on this is that it puts NRC out there where they have to do more work to cover themselves politically, and can't simply point to years or PWR data to show due diligence. <speculation>The nastier side is that they draw a lot of staff from Big Power (Westinghouse and the operators), have grown comfortable with the multimillion dollar fees they can charge Big Power for approval, and are just overly cosy with them. So to cover for all this, they demand expensive documentation with the sub-text that they won't act on it even if submitted, i.e., go away, and Big Power happens to be just fine with that too</speculation>.

 

[gmax137]

...The NRC is not going to review it unless the folks at Hyperion provide experimental data, and demonstrate that they can simulate the physics of the reactor and fuel.

Well what else would we expect? That NRC provide the experimental data and NRC develop calculational methodologies? Remember, everyone got their wish back in the 1970's and the NRC was split from AEC - to ensure that the regulators were not synonymous with the promoters. It is not the NRC's job/function to develop new technology.

And as far as 'big power = W' , I suspect Westinghouse isn't involved because they don't develop products for markets that don't exist yet. hmmm that's not very clear, is it? WEC's customers (the power companies) are not interested in anything that isn't licensed. And if the customers aren't interested, why would WEC be interested? Another Catch-22.

Also, the power companies are looking at 1100 MWe plants as "too small" so I don't know who is going to buy one of these little things. Finally, W and Areva make their money by selling the reactor fuel, so a device that requires refueling every 10 years isn't high on their list of things to look into. The naval reactors are a different story, 'big power' was interested in them to (1) get into the technology, and (2) bragging rights/national pride in the 1950s - 60s.

Sorry if this has a ranting tone, it wasn't meant to be

 

[Astronuc]

Well what else would we expect? That NRC provide the experimental data and NRC develop calculational methodologies?

Um no. I was responding the hype by Hyperion. Yes - they've talked to the NRC. But they only have a design on paper. The NRC will not evaluate it until there is experimental evidence that the concept works - i.e. it exhibits predictable behavior, particular with respect to reactivity control. It is ludicrous for Hyperion to expect otherwise.

Remember, everyone got their wish back in the 1970's and the NRC was split from AEC - to ensure that the regulators were not synonymous with the promoters. It is not the NRC's job/function to develop new technology.

The AEC did developmental research, and DOE still does, e.g. Gen IV. The Hyperion concept comes out of LANL. Does LANL have experimental data? If so, then Hyperion would have to submit those to the NRC before being seriously considered.

As for 1100 MWe being too small, various US utilities have selected the AP1000 (~3415 MWt/1200 MWe (gross)/1115 MWe (net)), as opposed to the EPR (4590 MWt / 1700 MWe) or APWR (4450 MWt / 1700 MWe). Some utilities are concerned about putting all their 'eggs in one basket' at a given station, although most Gen3+ plants are likely to be built at existing sites.

The ABWR is rated 3926 MWt / 1350 MWe (gross) and the ESBWR is projected to provide 4500 MWt / 1590 MWe (gross).

So far US utilities seem to favor PWR technology.

 

[mheslep]

Well what else would we expect? That NRC provide the experimental data and NRC develop calculational methodologies? Remember, everyone got their wish back in the 1970's and the NRC was split from AEC - to ensure that the regulators were not synonymous with the promoters. It is not the NRC's job/function to develop new technology.

Nor is it their job to impede new technology .

Also, the power companies are looking at 1100 MWe plants as "too small" so I don't know who is going to buy one of these little things.

Who says they're 'too small'? I think you have that exactly backwards. What they are is too expensive at $4 to $8 per W(e) installed. Hyperion is quoting their 25MW(e) reactor at $1/W(e). The question is more likely who would bye an AP1000? So far no one in the US.

Finally, W and Areva make their money by selling the reactor fuel, so a device that requires refueling every 10 years isn't high on their list of things to look into.

Well then thankfully we have the like of Hyperion, Babcock and Wilcox, etc, to address this market.

 

[gmax137]

Nor is it their job to impede new technology .

In this context, their job is to review and approve (or not) license applications. The job of the applicant is to provide a complete application. That seems pretty clear to me; I don't think you can blame the NRC for lack of progress by Hyperion, or LANL, or DOE, or anyone else.

Who says they're 'too small'? I think you have that exactly backwards. What they are is too expensive at $4 to $8 per W(e) installed.

That's a good point.

Hyperion is quoting their 25MW(e) reactor at $1/W(e).

Can you help me understand this? I can't figure out what is included in that cost. What's confusing me is the combination of "MWe," "brayton cycle," and "no moving parts." How do they propose to convert the heat into electric power w/o moving parts? Or, are the moving parts an additional cost (for those customers who want MWe no MWth)? I haven't seen this explained in the Hyperion website.

 

[Astronuc]

Hyperion is quoting their 25MW(e) reactor at $1/W(e).

Can you help me understand this? I can't figure out what is included in that cost. What's confusing me is the combination of "MWe," "brayton cycle," and "no moving parts." How do they propose to convert the heat into electric power w/o moving parts? Or, are the moving parts an additional cost (for those customers who want MWe no MWth)? I haven't seen this explained in the Hyperion website.

This is a key point. Does the cost projected by Hyperion cover just the modular unit? If so then what does the rest of the plant cost? It will have to be secure!

A 25MW(e) reactor will cost $25M at $1/W(e), but will that require a $25 million facility or a $100 M facility or more in which to locate it?

And what happens with the back end? Will Hyperion take back the unit? Where will it be stored? Will it be reprocessed? If so, where? What about ultimate disposal? What about liability?

And Hyperion still has to demonstrate that the technology will perform as claimed, and reliably for the design lifetime of 5 years or so.

 

[mheslep]

...Can you help me understand this? I can't figure out what is included in that cost. What's confusing me is the combination of "MWe," "brayton cycle," and "no moving parts." How do they propose to convert the heat into electric power w/o moving parts? Or, are the moving parts an additional cost (for those customers who want MWe no MWth)? I haven't seen this explained in the Hyperion website.

I had not seen the no moving parts description, but if true that must apply only to the reactor chamber, emphasizing that the design uses no mechanical control rods.

 

[mheslep]

This is a key point. Does the cost projected by Hyperion cover just the modular unit? If so then what does the rest of the plant cost? It will have to be secure!

Almost the cost is just the modular unit, as site costs will be variable and Hyperion could not quote them.

A 25MW(e) reactor will cost $25M at $1/W(e), but will that require a $25 million facility or a $100 M facility or more in which to locate it?

The simple graphics shown so far indicate they believe a simple hole in the ground with some concrete lining would suffice. No doubt that's true for nominal operation, but then what about seismic events, the protection against which drive costs so high elsewhere? I venture they would say there reactor is inherently fail safe - if the reactor is damaged the hydrogen escapes and all reaction except the normal decay just stops.

And what happens with the back end? Will Hyperion take back the unit? Where will it be stored? Will it be reprocessed? If so, where? What about ultimate disposal? What about liability?

This they do address to some degree: refueling is accomplished by shipping back some kind of core to the factory.

And Hyperion still has to demonstrate that the technology will perform as claimed, and reliably for the design lifetime of 5 years or so.

As this technology came out of Los Alamos, there must have been some prototyping done there.

It occurs to me a small portable low cost portable reactor such as this with siting complications would be idea for one customer in particular: the Army/Airforce. They looked at small reactors decades ago (Army Nuclear Power Program) and decided the state of the art at the time was too much trouble. But now this, a $1/W(e) reactor that could be flown in on one C-130 and installed at an ~ airbase in a hole dug in a couple days, and continue to run without refueling for five years? And, the main hurdle for which, the NRC, does not apply to them.

 

[mheslep]

Pertaining to the earlier NRC review, I rediscovered this from them:

So for the next two years, the NRC "will need to limit interactions with the designers of small power reactors to occasional meetings or other non-resource-intensive activities," the federal agency said in a letter to Babcock & Wilcox on May 27. "As such, any requested work on the mPower reactor design that goes beyond these limitations will be placed on hold."

http://www.cleveland.com/ohio-utilities/index.ssf/2009/06/new_nuclearl.html

That seems woefully misguided at the least, and at the worst has the appearance of corrupt collusion with the large plant vendors, so that they can keep out the competition and protect those colossal investments.

 

[joelupchurch]

So far US utilities seem to favor PWR technology.

They don't really have any choice. They can't build a design the NRC hasn't approved.

I'm surprised anyone even tries to build new reactor designs in the United States, since the regulatory environment is so unfavorable. As it stands now, China can have a reactor built and operating in less time than it takes for us to do the paperwork.

China is actually building production pebble bed reactors and they haven't even got on the queue yet at the NRC.

http://www.world-nuclear.org/info/inf63.html"

 

[mheslep]

They don't really have any choice. They can't build a design the NRC hasn't approved.

I'm surprised anyone even tries to build new reactor designs in the United States, since the regulatory environment is so unfavorable. As it stands now, China can have a reactor built and operating in less time than it takes for us to do the paperwork.

China is actually building production pebble bed reactors and they haven't even got on the queue yet at the NRC.
...

Certainly I'd like to see more transparency and speed in the US regulatory environment, though I wouldn't point to the Chinese example alone and waive our system away. The Soviets before Chernobyl likely didn't let paperwork slow them down either. Also the Chinese simply take their more corrupt bad actors out and shoot them, a somewhat sharper incentive than paperwork to not cut corners in operations.

 

[joelupchurch]

Certainly I'd like to see more transparency and speed in the US regulatory environment, though I wouldn't point to the Chinese example alone and waive our system away. The Soviets before Chernobyl likely didn't let paperwork slow them down either. Also the Chinese simply take their more corrupt bad actors out and shoot them, a somewhat sharper incentive than paperwork to not cut corners in operations.

I hardly think a comparison of the current Chinese government to the old Soviet government is valid. The last time I looked eight out of nine standing members of the Chinese politburo are former engineers.

I have no idea why it takes the NRC several years to issue a site license. In every case, the licenses are for plant designs that the NRC has already approved and in many cases at sites that already have operating nuclear power plants. What is there to check?

All the important oversight would happen during construction when you can verify that the reactor is being built to specification, but you can't do that until construction starts.

 

[oldsloguy]

Apparently there is also an ANL/LLNL design.

https://www.llnl.gov/str/JulAug04/Smith.html

 

[mheslep]

https://www.llnl.gov/str/JulAug04/Smith.html

Not much detail on the design there. I'm especially curious, as before, how the system is cooled externally, and what is the nature of the connection to an steam turbine and generator.

Also:

the reactor will be about 15 meters tall by 3 meters wide and will not weigh more than 500 tons—small and lightweight enough to be transported on a ship and by a heavy-haul transport truck.

A 500 ton truck load? Surely they mean rail only?

 

[gmax137]

A 500 ton truck load? Surely they mean rail only?

I watched Ansaldo's heavy haul contractors bring the replacement steam generators into Palo Verde - each SG is about 1.5 million pounds dry. Calling those rigs "trucks" is kind of a stretch, but they are surely road-going, not railroad. From Italy by sea to Mexico then over roads to the plant, west of Phoenix.

See also "schnaebel car"

 

[mheslep]

I watched Ansaldo's heavy haul contractors bring the replacement steam generators into Palo Verde - each SG is about 1.5 million pounds dry. Calling those rigs "trucks" is kind of a stretch, but they are surely road-going, not railroad. From Italy by sea to Mexico then over roads to the plant, west of Phoenix...

Well that would do it.

I've been by Palo Verde a few times. Any idea if it's 4GW is the largest nuclear capacity in the world? Also, what does P.V. use for a water source outside Phoenix? I saw those large open shallow pools, couldn't make sense of them.

 

[gmax137]

Any idea if it's 4GW is the largest nuclear capacity in the world?

I'm pretty sure it's the highest capacity site in the US, but it's not the highest "in the world." The Ulchin and Yeonggwang sites in South Korea each have six reactors, giving each site close to 6,000 MWe total capacity. People talk about France's nuclear program, but most don't seem to realize what the Koreans have done. Also, they have the capability to fabricate the heavy pressure vessels & steam generators (eg, at Doosan).

Also, what does P.V. use for a water source outside Phoenix? I saw those large open shallow pools, couldn't make sense of them.

The ultimate heat sink (cools the condensers) is forced-draft cooling towers that use Phoenix's waste water (yes, treated sewage). The large pools you saw from the road are holding pools for the water from Phoenix. It comes all the way out to the site in a long pipeline. The safety related heat sink (for cooling the reactor) is a separate system of 'spray ponds.'

 

[joelupchurch]

I happened to come across an article recently discussing the SNAP 10A, which is the only fission reactor that the United States ever launched in space.

http://www.etec.energy.gov/History/Major-Operations/SNAP-Overview.html"

What grabbed me was a picture of the reactor core:

http://www.etec.energy.gov/History/Major-Operations/MajorOpsImages/S8DR_Core_assembly_31Jul64THUMB.jpg

The reactor only put out 500 watts, but it was severely limited since it had to dissipate all the heat by radiation. I suspect an air cooled design with the same core could have put out lot more power.

 

[Astronuc]

TEPCO’s Kashiwazaki-Kariwa nuclear station, which consists of seven (7) large BWRs, is the largest nuclear power generation facility in the world, licensed for 8,200 MWe (when all 7 are up and running).

 

[mheslep]

TEPCO’s Kashiwazaki-Kariwa nuclear station, which consists of seven (7) large BWRs, is the largest nuclear power generation facility in the world, licensed for 8,200 MWe (when all 7 are up and running).

Thanks Astronuc. That's a lot of power in one place. I note http://nuclearstreet.com/blogs/nuclear_power_news/archive/2009/05/08/xtepco.aspx" :

All seven nuclear generators at Kashiwazaki-Kariwa nuclear power plant have been shut since a magnitude 6.8 quake hit the region in July 16, 2007. The least damaged No.7 unit would be the first of the seven to be restarted.

No doubt they are back up and running now, but to lose 8,200 MWe instantly from their grid like that must have been difficult. The threat of such a loss argues heavily against mammoth installations in my view, and in favor of smaller distributed power sources like the ones mentioned in this thread, especially for seismically active countries.