The Economics of Nuclear Power

[LostCosmonaut]

One of the many issues facing nuclear power in the US is economics. Once operational, nuclear energy is quite cheap in terms of $/MWh. However, the startup costs are enormous, which results in the overall cost of power going up. In an environment like today, where there's cheaply available natural gas, nuclear power becomes quite unattractive, especially with all the extra taxes levied on it and public fears.

 

This article has some good info; http://www.world-nuclear.org/information-library/economic-aspects/economics-of-nuclear-power.aspx

 

One interesting aspect which I first heard today is that the inclusion of renewables in the energy grid hurts nuclear energy. Nuclear plants are best run at a steady power level, providing base load. However, renewables such as wind and solar provide wildly varying amounts of power depending on the weather conditions or time or day. As a result, the nuclear plant is forced to vary its power level, running less efficiently. For reasons that I'm not 100% sure of, gas and coal plants can respond to variations in base load better, and so are more economic.

 

There's a couple ways I could think of to get around this. One is mass production of standardized reactor designs, I believe there's already been a lot of work on that front. Another is putting multiple reactors in the same plant; that article I linked says the Chinese estimate you get a 15% cost reduction (in $/MWh) by putting two reactors on the same site instead of separately.

 

I also heard that before building wind and solar was the cool thing to do to score political points, hydro and geothermal were the big renewable energy sources. Because they were economically viable, not because of hippies. Hydro being economically viable is obvious, but I'm more curious about geothermal. It seems like a pretty good energy source to use (where available).

[Belesarius]

One of the big problems I've noticed with nuclear is loss of engineering talent.  You only have to build nuclear plants every so often, so they aren't in constant construction. so you loose your engineering and construction expertise as they move onto new projects.  Which leads to your engineers having to relearn how to fix what the folks before them had already figured out. It's one of the same problems that Toronto has run into with subway construction. This leads to snowballing construction costs which, of course politicians and the media and taxpayers end up screaming about. 

[Collimatrix]

Somewhere I have a chart showing a breakdown of the costs associated with a nuclear reactor.  They're quite different than for other energy sources.

 

Only about a fifth of the cost of a nuclear reactor comes from the fuel, which is actually somewhat surprising because enriched uranium is pretty monstrously expensive stuff.  Must be something about the six orders of magnitude higher energy density.

 

The remaining 80% of the costs are things like maintenance, decommissioning, and construction.

 

So, a highly competitive nuclear reactor design would need to be optimized to reduce those costs, and fuel costs don't matter as much.  Sorry Kirk Sorensen; the amazingly abundant fuel and high burnup of the LFTR probably won't make that big of a difference to the bottom line.  Not when you have to fuck around with two heat exchanger loops filled with molten ThF4.

 

I was provisionally under the impression that nuclear reactors don't deal well with peak load stuff because steam turbines don't like to throttle up and down quickly.  Around here, where nuclear has been abandoned because stupid, coal provides the base load and the peaking plants are all natural gas.  Coal plants use steam turbines, natural gas are largely gas turbines.  AIUI, if you are too aggressive with the throttle on a steam turbine the steam might condense into liquid water drops, which will do hideous things to the turbine blades at operating RPMs.

[Oedipus Wreckx-n-Effect]

I heard a while back that the Carter administration put a law together that killed the refinement of nuclear waste. That without this law, we could refine nuclear waste again and again and reuse said waste until the half life was a much shorter period and there was far less of said waste. Any truth to this?

[Collimatrix]

I heard a while back that the Carter administration put a law together that killed the refinement of nuclear waste. That without this law, we could refine nuclear waste again and again and reuse said waste until the half life was a much shorter period and there was far less of said waste. Any truth to this?

 

Yep.  In 1977 the Carter Administration indefinitely deferred reprocessing of nuclear waste.  A typical commercial power generating reactor like a light water reactor actually fissions relatively little of the fissile material that goes into it.  Let's say that the design needs uranium enriched to 3% ²³⁵U in order to reach criticality.  You feed it uranium enriched to 3.5% (just bullshitting numbers here, but they should be ballpark for light water reactors).  It stays critical until it gets down to 3%, and after that there's not a sufficient concentration of fissile material to keep reacting.

 

That's a hell of a lot of ²³⁵U still left in the spent fuel.

 

On top of that, there's a significant amount of ²³⁸U that got bred into ²³⁹Pu (all reactors breed ²³⁹Pu, but breeder reactors do it faster).  So fuel reprocessing can get at that stuff too.

 

The reasoning behind the cessation of fuel reprocessing is that ²³⁹Pu is the preferred material for nuclear weapons (although it generates electricity just fine too), and there were concerns that moving large amounts of reprocessed fuel around would make it easier for plutonium to go missing for illicit bomb making.  This is complete horseshit and ignores the well-documented fact that plutonium reprocessed from spent reactor fuel is hopelessly contaminated with ²⁴⁰Pu, an isotope that has a very high spontaneous fission rate.  Even a fairly modest amount of ²⁴⁰Pu contamination makes the plutonium unsuitable as a bomb material, and enriching the plutonium to remove the ²⁴⁰Pu is extremely expensive, since the mass ratio between it and the desired isotope (²³⁹Pu) is so close.  With laser enrichment this would not be the case, but laser enrichment of actinides doesn't exist yet.  So when the decision was made it was complete horseshit.

But you try to explain any of this to a hippie and their eyes just roll back.  This is why you should never attempt to explain things to hippies.  Just punch them instead.

 

The reason this could allow waste with much more manageable half-life is a little more involved, and would require a more involved fuel cycle that I don't think anyone has seriously tried.

 

You have two sorts of radioactive crap in spent fuel.  There are fission fragments, which are the things that the uranium that fissioned split into, and you have long-lived actinides, which are various radioisotopes that the un-fissioned uranium got turned into while bathing in the neutron flux of the reactor core.  Most of the long-lived actinides are plutonium isotopes, but you also get some exotic americium, curium and neptunium isotopes, along with various other weird shit.  Generally speaking, the fission fragments have half-lives in the microseconds to decades range.  But long-lived actinides can stick around for years, or tens of years, or tens of thousands of years.  Much more annoying.

 

A fast neutron reactor can break down long-lived actinides into the less persistent fission fragments.  However, that would require a large number of commercial fast fission reactors, and only a few nations have even dabbled with that concept.  The Soviets and the French got the furthest, but it looks like the Chinese will pick up at some point in the next ten years.  I'm not entirely sure what sort of fuel cycle they intend to pursue (according to Mech, they're a little bit touchy about their nuclear stuff), but their building program for the future includes a large number of fast-neutron reactors.  Last I checked these were supposed to be license-built versions of the Soviet/Russian BN-800 design, but the article linked above mentions a gas-cooled design that I'm not familiar with.

[Ulric]

From my discussions with a person who is involved wit the management of a local natural gas burning power plant, the coal fired plants in the area are used to take up the base load, while the natural gas plants can throttle to cover fluctuations in the wind and solar production. This single plant burns an absolutely hilarious amount of gas. IIRC, he said that in one month they burned more gas than North Dakota burned in a year. From my understanding, currently you could only replace the coal and oil fired plants with nuclear. You would have to maintain some variety of quick response ability to cover fluctuations in grid loading during the day. I imagine that this would still be the case even if you got rid of wind and solar all together, but you would have a more predictable change in grid loading.

 

It's a real shame about that natural gas plant. It used to be the only commercially operated nuclear reactor that used helium as a working fluid. Maybe we can find a way to use hippies as a renewable energy source. They might actually provide some benefit to society, then.

[Collimatrix]

The British AGCR nuke plants use CO₂ for coolant.  I wonder if they are more responsive to throttle inputs, and whether they could be used for peaking.

[LostCosmonaut]

That would seem to be a good idea, using CO2 or helium as the working fluid would eliminate the problem of having phase changes in your turbine causing issues.

 

PWRs look like they would be easier to adapt to this, since they have a separate steam loops for cooling and the turbines. You could still use water for cooling the reactor, and use something else to run the turbines.

 

Then again, CO2, helium, and other gases have a lower heat capacity, so you might not be able to remove enough heat from the coolant to make it viable without ridiculous mass flows.

[Collimatrix]

The lower heat capacity also reduces the thermal efficiency of the turbine.  At least, in theory it does.  AGCRs actually manage a better turbine thermal efficiency than most PWRs.  But for some reason they have lower burnup.  How in the hell they ended up with better thermal efficiency using CO₂ but lower burnup using a graphite moderator, I'm not entirely sure.  Nuclear reactor engineering is complicated sometimes.

 

To be perfectly honest, I'm not clear what prevents nuclear plants from being good at load-following.  My guess is that it's a limitation of steam turbines, but it may also be a limitation of the reactor core itself.  I've done a bit of research, and the issue of ¹³⁵Xe neutron poisoning has come up.  It is interesting to note, however, that naval reactors apparently react more readily to throttle input than gas-fired steam turbines:

 

 

Speaking of the Enterprise, she left Bainbridge behind just as she did to many ships during the Vietnam War. When she was launching planes, she accelerated very quickly and kept at high speed for hours on end. Those techniques made her look much faster than she actually was. To really understand this, you have to be along side Enterprise (or a Nimitz) when they accelerate. It is impressive. The Bainbridge could out accelerate the Big E easily, but no conventional steam-powered ship has a chance. You see, you just can't wing the throttles open in a tin can like you can in a "nuke." Heat input is too low. Steam pressure falls off, you lose critical heat, the boilers depressurize and cool down, and the steam bubble collapses… nastily. You have to increase speed slowly on a conventional critical steam plant. You have to build up heat (actually heat flow), and maintain temperature and pressure as you slowly accelerate in a tin can.

 

Nuclear power plants simply don't have those limitations. When Big E had to launch on short order, she just ran away from her escorts. Conventional carriers just couldn't do that! Therein lays the seed of deception and myth. Enterprise looked fast, because fast destroyers couldn't catch her! By the time they got up to speed and began closing distance, Big E was back down to what appeared to be normal speed (though she was still at her maximum speed). Sailors that didn't know better (we can go 34 knots, and Big E just ran away from us… we couldn't catch her until she slowed down!), thought that Big E had to be able to achieve speeds of 36-40 knots to do the things that they all saw with their own eyes. In fact, her throttle-men were not limited by fire rates, fuel pumps, or critical boiler conditions. Steam generator temperature was controlled by a reactor, and it could change heat rate in a heart beat.

 

But naval reactors are quite different than most power-producing reactors, and may be optimized for good throttle response.  So I'm not really sure what's going on here.  The fact that most naval reactors are fast-neutron types and most civil powerplants are thermal neutron types does spring immediately to mind, however.

 

Is this more evidence that fast neutron reactors are the master race?  Probably!  Is this more evidence that using low enriched uranium on naval reactors is an idiotic idea pushed by hippies and other wrecker scum?  Probably!

[Oedipus Wreckx-n-Effect]

Cp of CO2 gas ranges from 0.8 to around  1 Kj/Kg*mol.

 

 

Helium gas does better, sitting at around 5.09 Kj/Kg*mol. 

http://catalog.conveyorspneumatic.com/Asset/FLS%20Specific%20Heat%20Capacities%20of%20Gases.pdf

 

Helium could definitely work, but it's expensive and we're running out of helium on Earth.

 

Also, I just asked a professor friend of mine why the hell Helium gas has such a high heat capacity. 

 

He says, "Well, it's sort of trippy and it has something to do with its chemistry."

 

Thaaaaanks Doc. 

 

My issue with using a gas would come with the heat transfer coefficients being rather low compared to that of a liquid. 

[Collimatrix]

Cp of CO2 gas ranges from 0.8 to around  1 Kj/Kg*mol.

 

 

Helium gas does better, sitting at around 5.09 Kj/Kg*mol. 

http://catalog.conveyorspneumatic.com/Asset/FLS%20Specific%20Heat%20Capacities%20of%20Gases.pdf

 

Helium could definitely work, but it's expensive and we're running out of helium on Earth.

 

Also, I just asked a professor friend of mine why the hell Helium gas has such a high heat capacity. 

 

He says, "Well, it's sort of trippy and it has something to do with its chemistry."

 

Thaaaaanks Doc. 

 

My issue with using a gas would come with the heat transfer coefficients being rather low compared to that of a liquid. 

 

Helium has been used before, look up the Saint Vrain reactor, which used a helium coolant connected to steam turbines via heat exchanger.  In addition to having bizarrely high specific heat capacity, helium is non-corrosive,doesn't turn radioactive under neutron bombardment, and hardly moderates neutrons at all in fast reactors (it seems like it should because it has such a low atomic mass, but it doesn't because reasons^TM) so it's a pretty good reactor coolant.

 

Pity about it being so expensive and leaking all the damn time.

[Oedipus Wreckx-n-Effect]

I'm sure it is a good reactor coolant choice. However, after reading a few reports about FSV's reactor, it sounds like the failures were a little bit design and a lot human fuck-ups. 

 

https://en.wikipedia.org/wiki/Fort_St._Vrain_Generating_Station

 

If you want to read one of the worst written summaries of their nuclear reactor design, check out the wiki page. 

[LostCosmonaut]

The way I was thinking about was flipping around FSV, using a water cooled reactor (to piggyback off existing expertise with PWRs) and a helium secondary going through the turbine.

[Oedipus Wreckx-n-Effect]

That could work pretty well. 

 

Also, am I the only one who thinks that the motto of the NRC paints nuclear power in a terrible light?

 

http://www.nrc.gov/reading-rm/doc-collections/nuregs/contract/cr6839/

 

"Protecting People and the Environment"

[LostCosmonaut]

Here's a list of gases by specific heat; http://www.engineeringtoolbox.com/specific-heat-capacity-gases-d_159.html

 

Hydrogen actually has the highest of anything listed, followed by helium and N2O4. Fun!

 

Edit: natural gas is pretty decent, obviously we should use fission reactors as preheaters for colossal natural gas plants.

 

I should point out that in order to heat your working fluid to maximum temp quickly, you'd want a low heat capacity, I'm just worried about the capabiltiy of something like Xenon to remove enough heat from the water loop the keep the reactor within an acceptable temperature range.

[Collimatrix]

I rather like the concept Rod Adams put forward; nitrogen cooled core spinning COTS gas turbine blades.  Gas turbines are designed to operate in air, air is mostly nitrogen, therefore COTS turbines will work with pure nitrogen.  And work the first time.  And be cheap.  There are no heat exchangers, in order to keep costs down.  The design sacrifices some cycle efficiency, but who gives a fuck it's nuclear and your energy density is nuts so screw efficiency, and in exchange is supposed to be much cheaper to build and to operate.  There is some small problem with nitrogen sucking up neutrons and becoming radioactive ¹⁴C, but Adams claims this can be scrubbed out fairly easily.

 

As a fun mental exercise, try to come up with a fast reactor design that uses a coolant that doesn't explode on contact with air or water, isn't toxic, can be run directly from the core to the turbines, and isn't absurdly expensive.  I can't think of any.

 

Lost, does the specific heat for N2O4 include the additional capacity supplied by the reversible decomposition to NO2?

[LostCosmonaut]

I don't believe it does.

 

Argon seems like a reasonable option, considering it's widely available and nonreactive. The specific heat ratio is a lot different from air, though, which keeps you from using a COTS turbine.

[Collimatrix]

I don't believe it does.

 

Argon seems like a reasonable option, considering it's widely available and nonreactive. The specific heat ratio is a lot different from air, though, which keeps you from using a COTS turbine.

 

I had a similar thought about argon, but I'll be damned if I can find anything about its neutron absorption cross section.  A big part of what makes ⁴He so attractive is its bizarrely low neutron interaction cross section.  I think this is a function of its extremely high binding energy per nucleon, but it could be the result of some other magical property of the atom.  ⁴He has a great number of magical properties.

 

I do find a few results claiming that the BN-1200 fast neutron reactor is sodium and argon cooled.

[LostCosmonaut]

Very detailed report from MIT on the use of nuclear in low carbon energy infrastructure; http://energy.mit.edu/research/future-nuclear-energy-carbon-constrained-world/

 

TL;DR Nuclear becomes more attractive the more you reduce your emissions goals (in terms of kgCO2/kW)

[Alzoc]

A bit of necro here but to pick up on the "why nuclear power is bad at load following" they aren't particularly bad at it and naval reactor demonstrate that just fine as @Collimatrix pointed out (although they have a bit of an advantage because of their small size and hence low thermal inertia).

 

The main reason is that it's not cost efficient and that it put some stress on the equipment , and since equipment qualified for use in nuclear environment tend to be on the expensive side you avoid it if you can. Whenever possible you want your NPP to run on base load.

 

Another reason is that it's a non-trivial thing, it require a fine monitoring, understanding and control over the neutron flux and power level in the core (understand you need to be capable to map/simulate what's happening in the core in 3D and in real time).

 

Since we have 75% of our electricity generated by NPP here our plants needed to be able to do load following to a greater extent than in most other country and EDF have extensive experience with it. But it require secondary control rods for fine tuning and a lot of calculation before hand to define different mode of operation.

 

Spoiler

France in particular has a long history of utilizing aggressive load following with their PWRs, which are capable of (and used for) both primary and secondary frequency control in addition to load following. French PWRs use "grey" and/or "black" control rods in order to maneuver power more rapidly than chemical shim control or conventional control rods allow. These reactors have the capability to regularly vary their output between 30–100% of rated power, to maneuver power up or down by 2–5%/minute during load following activities, and to participate in primary and secondary frequency control at ±2–3% (primary frequency control) and ±3–5% (secondary frequency control, ≥5% for N4 reactors in Mode X). Depending on the exact design and operating mode, their ability to handle low power operation or fast ramping may be partially limited during the very late stages of the fuel cycle.^[10]

 

And if I trust this document, apparently you guys also have legal limitations with load following on top of the technical aspects:

 

Spoiler

In some countries, there are explicit regulatory limitations on manoeuvring in the automatic mode.

For example, according to the US Code of Federal Regulations (10 CFR Part 50),

 

“the licensee may not permit the manipulation of the controls of any facility by anyone who is not a licensed operator...”

and

“Apparatus and mechanisms other than controls, the operation of which may affect the reactivity or power level of a reactor shall be manipulated only with the knowledge and consent of an operator or senior operator licensed pursuant to part 55 of this chapter present at the controls”.

 

Although this does not prohibit power load variations controlled by the operator (if justified from the technical and economic viewpoints), maneuvering in automatic mode is not authorized by current regulations in the United States.

 

Long version on the subject:

 

Technical and Economic aspects of load following with Nuclear Power Plants

[LostCosmonaut]

7 hours ago, Alzoc said:

 

And if I trust this document, apparently you guys also have legal limitations with load following on top of the technical aspects:

 

  Hide contents

In some countries, there are explicit regulatory limitations on manoeuvring in the automatic mode.

For example, according to the US Code of Federal Regulations (10 CFR Part 50),

 

“the licensee may not permit the manipulation of the controls of any facility by anyone who is not a licensed operator...”

and

“Apparatus and mechanisms other than controls, the operation of which may affect the reactivity or power level of a reactor shall be manipulated only with the knowledge and consent of an operator or senior operator licensed pursuant to part 55 of this chapter present at the controls”.

 

Although this does not prohibit power load variations controlled by the operator (if justified from the technical and economic viewpoints), maneuvering in automatic mode is not authorized by current regulations in the United States.

 

 

 

This seems like something that should be revisited now since automatic systems are significantly more advanced than when the regulation was presumably first written (a cursory search doesn't reveal when that section of 10 CFR 50 was enacted). Appropriately tested software monitored by an operator should be sufficient for  civilian applications.

[Alzoc]

13 minutes ago, LostCosmonaut said:

 

This seems like something that should be revisited now since automatic systems are significantly more advanced than when the regulation was presumably first written (a cursory search doesn't reveal when that section of 10 CFR 50 was enacted). Appropriately tested software monitored by an operator should be sufficient for  civilian applications.

 

I guess so.

You probably have a chance that it'll happen at some point since the NRC isn't as stubborn than our safety authority.

I mean we still have very low activity waste category and they just don't want to delete it.

 

In practice it means that anything that went into a controlled area will be disposed as waste as long as it has as much a 1 Bq/g (and you'll always manage to detect something) in it (the upper limit being 100 Bq/g where it's considered as low activity waste).

In turn it mean that a shit-ton of valuable materials that could be recycled goes into the nuclear waste circuit with all the afferent cost and paperwork.

Not even mentioning the fact that it makes the nuclear industry look like it produce a lot of radioactive waste and send the message to the population that it's actually dangerous from the first becquerel!

[LostCosmonaut]

Vogtle 3 is getting closer to actually going online; https://www.world-nuclear-news.org/Articles/Georgia-Power-orders-fuel-for-Vogtle-3

 

With this and Nu-Scale working on getting an SMR test plant going on INL, things are actually looking not too terrible for civilian nuclear power.