Turning nuclear into a fuel dominated business
Under our current energy paradigm, nuclear power has the reputation of needing enormous up-front capital investments. Once those investments have been made and the plants are complete, the payoff is that they have low recurring fuel costs.
Just the opposite is said of simple cycle natural gas fired combustion turbines. They require a small capital investment that can be paid off even if the plant only operates a few hundred hours per year. They don’t have an optimized fuel efficiency and they burn a fuel that can be quite expensive during the hours when the “peakers” need to run.
Those peakers are responsive and are becoming more interesting to power producers with the continued growth In variable renewable energy sources like wind and solar.
Just thinking out loud here, but what if it was possible to build really simple, much lower cost nuclear plants on the condition that they have a safety case that is built around a fuel design that is several times more pricey than conventional commercial nuclear fuel?
For more than 50 years, scientists and engineers have been working on coated particle fuels where tiny particles of fissile material in various chemical forms is surrounded by tightly adherent and durable layers of material capable of withstanding very high temperatures without releasing fission products.
In the space program, incredibly powerful and energy dense reactors have been designed and tested using coated particle fuels, but for commercial power generation, the usual path is to create large low power density reactors that are considered to be “inherently safe” because they don’t need any active cooling systems to prevent the core temperatures from exceeding the much more generous limits allowed by high temperature fuel.
Unfortunately, the development of reactors using coated particle fuels has been held back by a couple of technical choices. One has been that the reactors have been seen as a modest improvement in conventional reactors that still need to have most of the expensive equipment of a steam plant power conversion system.
Using that heat engine choice, designers must include heat exchangers that are functionally equivalent to the high cost steam generators in pressurized water reactors. Since they need heat exchangers, they naturally look toward high gas pressures in the primary coolant loop in order to increase heat transfer rates.
That path leads to systems with capital costs that are not much different from conventional nuclear plants with the added burden of using fuel that is quite a bit more expensive, especially in the early years before the manufacturing lines become cost efficient.
General Atomics achieved initial marketing success with a line of GW class high temperature reactors (HTRs) in the late 1960s and early 1970s by emphasizing that their systems were somewhat simpler and used more conventional steam turbines than the lower temperature light water reactors. They inked about 10 contracts, but none of the plants were ever built.
X-energy, URENCO and HTR-PM all are pursuing updated versions of similar designs. They are not radically reconsidering the paradigm.
It’s possible to dig back into nuclear history and find that the HTRE (high temperature reactor experiment) used modified jet engines that sucked in atmospheric air, heated it in a modestly high temperature reactor (far lower temps than coated particle fuels enable today) and exhausted that air through a turbine and jet expansion system.
The capital cost of equipment for such an air breathing system today would be quite low, but there would likely be a problem with creating and emitting Ar-41 as well as the possibility that fuel manufacturing defects might allow some small quantity of fission products to be discharged. Regulatory hurdles prevent that path from being developed anytime soon.
With a modest increase in complexity and capital equipment, a mechanically identical system could use nitrogen gas separated from air. Because the gas isn’t air, it would need a closed system where a low pressure, moderate temperature heat exchanger performs the function of returning turbine exhaust back to atmospheric conditions for injection into the compressor.
This ultimately simple Brayton Cycle gas turbine would use fuel that might cost several times more per unit of heavy metal than conventional nuclear plants, but its initial investment should approach the cost of the combustion turbines that would be the heart of the system. Sure, there are costs associated with the piping systems, but those would likely be on a similar order of magnitude as the fossil fuel system pipes that would not be needed.
With dramatically lower capital costs and higher fuel costs, the total system cost allotment would be a complete departure from the conventional nuclear paradigm. No longer would equipment suppliers and financial providers be able to capture 50-75% of the total revenues, with personnel costs capturing 30-40% and the fuel supplier pulling up the rear with 5-20% of the revenue. Instead, financial costs could be far lower. Equipment costs would drop dramatically. Personnel costs per unit of output would fall.
The obvious result is that the fuel suppliers, the people who produce the fuel that is so capable that it is the safety case and safety boundary, would gain the lion’s share of revenue from product sales.
That situation has proven itself in the energy market. Customers and other stakeholders don’t necessarily like the fact that fuels people walk off with most of the money, but it has meant that fuel suppliers have adequate capital to both invest in new technologies and adequate incentives to promote the benefits of high energy use.
There is a massive amount of capital in the hydrocarbon fuel business. There is also a great deal of intellectual capital, some of which is scientific and technical, but some of which is business development and marketing.
In the early days of nuclear energy, the hydrocarbon giants dipped their toes in the business. They couldn’t figure out how to make as much money in nuclear as they were used to making, so they quickly exited.
Perhaps this early Sunday morning essay will help stimulate them to reconsider their decision to abandon the business without figuring out how to make it a fuels business that could answer a lot of their future challenges.
Note: I have more details about the paradigm shift described above, but I think I’ll hold them closely for now.
Rod,
Very insightful and strange that I haven’t read more about this paradigm shift. I’m thinking upcoming events may bring this to light. Doesn’t hurt to be a step or two ahead of the others!
Thanks for the message.
Rod,
why use N2 as working fluid and not CO2?
The cross section for the 14N (n,p) 14C reaction is quite large across the whole spectrum, not only creatin lots of long-lives Carbon -14, but also eating precious neutrons.
The ratio of specific heats of CO2 is way different from air/N2, so you can’t buy off-the-shelf turbomachinery for it. Yet.
You can buy LOTS of OTS turbomachinery for steam, albeit not usually for saturated steam. I’m waiting for the first shipyard to announce a letter of intent with NuScale to produce an emissions-free container hauler. I’m betting it will happen after construction starts for UAMPS but before first criticality.
Are any binary noble gas mixtures suitable for off-the-shelf turbomachinery?
I am particularly intrigued by a mixture of He and Kr-85 (thus using a fission product).
The idea of binary noble gas mixtures has been considered, but (due to the cost of heavy noble gasses, I suppose) the consideration has focused on systems for space. See, e.g.,
https://www.researchgate.net/publication/268572981_Selection_of_Noble_Gas_Binary_Mixtures_for_Brayton_Space_Nuclear_Power_Systems
If the Kr-85 in SNF was recovered, though, the cost would be less (if, of course, the recovery of the Kr-85 was part and parcel to reprocessing that was already being conducted). While it would still not provide an avenue for a “shift” in nuclear (because of how few could be built), it could be the basis for some worthwhile reactors.
They’ll all have the same 1.667 Cp/Cv. The heavier ones would do wonders for your density, of course.
If you don’t mind almost 10% of your Kr plating out as rubidium each year, and a thermal neutron capture cross-section of something like 10 barns, go knock yourself out.
Thanks for the link, E-P — I have added it to my “nuke library.”
Small correction: Kr-85 has a half life of about 10 years.
If about 10% of it were to decay every year, its half life would be closer to 6 years.
More significantly: what characteristics of a gas (or a blend of gases) is required for compatibility with turbomachinery that is designed for operation using air?
I suppose that a match of both density and specific heat is required.
If so (and I am only looking for a very rough idea), what sort of diminution in efficiency is associated with changes in specific heat and/or density?
Mostly ratio of specific heats, IIUC. Density is a factor if it increases mechanical loads too much; it will change your absolute pressure but not your pressure ratio. What I’ve seen of performance curves suggests that absolute temperature is also a factor, as combustion turbine output falls off both above and below the design operating point.
Good question. Answering that would be utterly fascinating, taking a hypothetical design of a machine optimized for one set of characteristics and then checking its performance with a gas of different characteristics. But it’s way more work than I am willing to take on for a casual query even if I had the time. I am still stuck on a convoluted formula for calculating elapsed time for fractional elliptical orbits and then have to figure out how to work it into a spreadsheet.
But now that you made me think about it, you made me wonder if something like sulfur dioxide (MW = 64) plus neon (MW = 20) might not be worth a try. It would have a density considerably higher than air, and the cross-section of sulfur is around 1 barn. I suspect you’d have to use silicon carbide or other protective coatings on most of your hardware, though.
@RRMeyer
CO2 dissociates at temperatures of interest in conventional turbomachinery.
I think the problem is not so much that the CO2 disassociates, as that it reacts with the graphite/pyrolytic carbon cladding of triso fuel. Proposals have been made for sCO2 turbines running up to 700 C; suposedly supercritical carbon dioxide is less corrosive than supercritical water. Rather like the situation with light water reactors, which could run hotter if they didn’t use zirconium clad fuel.
@John
When CO2 dissociates, the resulting gas is even more reactive.
700 C isn’t an interesting injection temp for a simple cycle gas turbine.
CO2 is a well proven coolant for a steam system secondary, but it’s not very interesting in my quest for restructuring the nuclear cost pie chart.
Is one of the touted advantages of the new Lightbridge zirconium alloy fuel, that because the rods run hundreds of degrees cooler internally, that they can be rated-up to generate higher temperature steam? Can they work at superheated steam temperatures, or are there moderation issues?
@turnages
AFAIK, Lightbridge is focused on being a replacement fuel in conventional light water reactors. There are numerous components in that system that limit core coolant outlet temperatures, so I have not seen any discussion about higher outlet temperatures or attempting to generate superheated steam.
They do talk a good deal about the improved flow made possible by eliminating fuel pin spacer grids. Since power out is a function of mass flow rate and change in temperature from hot to cold leg, the increased mass flow with the same pumping power results in increased steam production and increased turbine power.
‘One key to a turbine’s fuel-to-power efficiency is the temperature at which it operates. Higher temperatures generally mean higher efficiencies, which in turn, can lead to more economical operation. Gas flowing through a typical power plant turbine can be as hot as 2300 degrees F, but some of the critical metals in the turbine can withstand temperatures only as hot as 1500 to 1700 degrees F ( up to 900 C ) Therefore, air from the compressor might be used for cooling key turbine components, reducing ultimate thermal efficiency.’ ( From the DOE, Office of Fossil Energy ).
That’s in a mix of air and combustion products, but the Allam cycle is designed to work at turbine inlet temperatures of ~1150 C, with 97.5% CO2, 2.5% water. At those temperatures, you will get monatomic oxygen and carbon monoxide in the mix, so resistant nickel-based alloys have to be used, but combustion turbines already are built to withstand nitrogen oxides, soot particles, and acids. Pure CO2 from a reactor would be comparatively benign – but not to the pyrolytic carbon in the fuel compacts. S CO2 acts as a solvent to flake off the graphite, and CO2 can donate an oxygen to the graphite to make two carbon monoxide molecules.
The French are working on nitrogen as a turbine gas in their next generation fast reactor, for similar reasons – CO2 might react with sodium if the heat exchanger leaked. ‘Despite the capabilities of the supercritical CO2 PCS to reach a high efficiency greater than 42%, the Nitrogen PCS option has been preferred due to its higher maturity. Nevertheless, SCO2 PCS remains an interesting option for commercial reactors, if Na/SCO2 reaction characterization will be done, N2 PCS being a first step towards operation feedback of a Brayton cycle.’
‘Progress in the Astrid Gas Power Conversion System Development’-https://media.superevent.com/documents/20170620/44a4aaa7d2807f0bfd13a547a5cacb97/fr17-285.pdf
‘
Thanks. What about the completely wacky idea of using carbon monoxide? As a reducing agent it may well be quite friendly to the graphite and metal components. Also has cp/cv=1.4, just like N2, adressing E.-P.’s concern.
Regarding carbon monoxide, you can expect major headaches from coking. Look up “Boudouard equilibrium”.
I agree. Terrible idea. Steel is a catalyst for 2 CO –> CO2 + C with the C embrittling the steel Probably also a major problem with CO2 at high temp, as it partiall dissociates to CO with the help of C from the Graphite. Leaves only Helium, as N2 is an absolute no starter due to ite appetite for neutrons.
No wonder that the Chinese looked at all the options and chose helium again for ther htr-pm.
You are right, carbon monoxide is an extremely bad idea as iron and nickel catalyse the 2CO –> CO2 + C reaction, and the C gets deposited in the steel, embrittling it.
Yes it is ironic (or tragic) that nuclear power’s greatest asset — fuel energy density orders of magnitude greater than its alternatives — is also in some ways its curse. No “uranium lobby” comparable to the coal, oil, and gas lobbys.
Breaking the paradigm to build “cheap” plants that use expensive fuel is a potentially great idea. Getting it through the regulators unscathed will be a hurdle.
The concept being promoted by Hybrid Power Technologies is a paradigm shift.
http://www.hybridpowertechnologies.com/
Way over-complicated and retains fossil-fuel dependency. I believe Cal Abel did it much better with a solar-salt storage tank between a high-temperature reactor of some kind and a stock steam turbine.
Over-complicated? How so? It is basically a nuclear powered air compressor connected to a combined cycle power plant (though there are variations that are nuclear-only, utilize coal in various fashions, and can work with solar thermal plants). The combined cycle side is basically stock, except for some modifications to the compression side of the combustion turbine.
Retains fossil-fuel dependency: mostly true (there is a nuclear-only option, but that option is not the best financial option) but so what. What is the point of going nuclear-only if it drives up costs to the point that they are not built? If you can tremendously reduce the fossil fuel usage and the associated emissions, is that not better than doing nothing but make paper reactors that are always just on the horizon?
Replace the air compressor with a generator, and everything to the right of it goes away. You’ve got a pile of extra machinery which mostly reduces the conversion efficiency between the helium turbine and the grid, unless you burn FF to supplement.
Cal Abel’s scheme is much better. Nuclear heat is really cheap, so just run the reactor at 100% feeding heat into a solar-salt tank at something like 500°C. Pull hot salt from the tank as required to make steam for a steam turbine; the steam generators have a considerably higher thermal power maximum than the reactor. This lets you (a) decouple the reactor from the instantaneous electric output so it can likely be controlled by the system operator without violating NRC regs, and (b) follow the entire daily load curve without burning fossil fuel. That’s something that Hybrid Power manifestly cannot do.
If you need more than the ST’s rated power or run out of hot salt, you fire up some open-cycle gas turbines and use heat-recovery systems on the GT exhausts to help top up the salt tank.
I *think* I’ve seen that someone was working on using TRISO fuel with the hear removed by a molten salt. If the salt then heated air to go through an open cycle gas turbine, that would likely work out very cheap with no problems of radioactive material going through the turbine & out into the atmosphere.
@Jim Baerg
Kairos Power is using a salt cooled pebble bed reactor, a heat exchanger and an open cycle gas turbine.
Their concept is to use natural gas to boost the turbine inlet temperature when they want to go To higher power operations.
Salt and the heat exchanger limit the temperature at which nuclear-only operations are possible.
I think it maxed out at something less than 800 C.
Might be viable and competitive, but it’s not my preference.
Actually, Kairos has decided to abandon this concept — at least for the first reactors that they intend to build. They probably will want to revisit it later, if they get the chance, but their effort to bring a reactor to market will skip this hybrid power system as too much risk and go with something more straightforward.
What Kairos might want to do is come up with a power block that is designed for air cooling. Ideally, passive circulation air cooling. Something you can put anywhere, not just where you’ve got a ready source of water.
Some time ago I read about a fellow who proposed using ammonia-water mixtures instead of pure steam to improve the thermodynamics of a vapor turbine cycle. From what I recall, the experts were initially skeptical but were eventually convinced by the numbers. One of the features of an ammonia-water mixture is that the vapor pressure is much higher over all temperatures. This would allow a dry cooling tower to have the turbine outlet vapor plumbed directly to its heat exchangers and not have to worry about collapsing the tubes under vacuum. One less heat exchanger means lower cost.
Something I suspect is that the higher vapor pressure means higher density especially at the last stages of the turbine, so a much more compact turbine; that could be a major cost saver. I’d have to dig pretty deep to verify this, though.
Replace molten salt with helium and you get the Hybrid Power Technologies concept.
Energy storage: improving fast reactor economics.
Cal Abel and Bojan Petrovic,
American Nuclear Society
Annual Meeting, June 2013
Yup, that’s the paper (or a summary thereof) I was talking about.
Everyone who isn’t familiar with the concept, follow that link. You’ll be glad you did.
Moltex have also picked up the thermal-energy-storage principle, using what they call their “GridReserve” technology, basically tanks of solar salt operating over a 565 – 270 degC range. See http://www.moltexenergy.com/learnmore/An_Introduction_Moltex_Energy_Technology_Portfolio.pdf .
RRMeyer — You might care to ask Sandia National Laboratories if they would care to research your idea of using carbon monoxide. With the right materials choices this could well be a feasible alternative.
Counter intuitive…. Make the fuel expensive.
Brilliant!!
@scaryjello
I trust you are not being sarcastic by using “brilliant” to describe my post.
But if you disagree with the premise, please elaborate.
You could be right about nuclear following a fuel model, but I think hydro is a better one for nuclear. The US has the world’s largest nuclear industry, but it’s been stagnant for a long time. In most other countries, nuclear has been pushed through by governments, as has hydro. The short term rewards for the developer may not be as attractive as from fossil fuels, but the long term benefits for society as a whole should be paramount. The government-led hydro programmes in Canada, the US, Norway, Australia, New Zealand, and many other countries have left solid assets that will last for generations. Reactors could be just the same.
It’s a shame that it tends to be authoritarian governments pushing nuclear, but that wasn’t always true. Hopefully the efforts of educators like yourself can help swing public opinion enough, to get more politicians with long term vision elected.
@John ONeill
You may be correct, but history tells me that the government dominated models of hydro and nuclear are not well suited to today’s US economy. It might be regretted that we no longer seem to be able to pull together in support of multigenerational projects that require long term support and leadership, but that is where we are.
The fuel dominated model has proven that it works reasonably well to attract the necessary investments from institutions, banks and individuals. Those entities develop sufficient interest and equity in expanding their market to also invest in the public marketing and political activities required to allow them to continue doing business with a certain amount of freedom.
Rod,
I love your work, but I think pressurised nitrogen as a reactor coolant is an extremely bad idea.
22 years ago you wrote as an aside:
Modern British gas cooled reactors have systems that inject nitrogen into the coolant system as a means of reducing core activity in case the normal shutdown mechanisms do not function.
How big is that reduction of core activity? Massive:
According to this book, page 389:
Nitrogen is a good absorber of neutrons, for example, it has been estimated that filling the Hinkley Point B reactor with nitrogen at a pressure approaching normal operating pressure [24 bar] would have a reactivity worth in excess of -20 Niles, twice the worth of the control rods.
That is a lot of reactivity. -20Niles is about -32$. This means in an Adam’s engine, where the enrichment is increased to make it critical while flooded with nitrogen at, say, 15 bar pressure, the loss of just 0.75 bar of that pressure would make the reactor prompt critical and lead to a Chernobyl-style runaway power excursion.
There is also a very large positive coolant temperature coefficient. A temperature increase by 5% would again be enough for prompt criticality.
The best accident-tolerant fuel is worthless in a reactor that allows a runaway power excursion.
@RRMeyer
I believe you are mixing up temperature coefficient of reactivity with pressure coefficient of reactivity.
I’ve had simulations and calculations done that indicate a much lower pressure affect on reactivity in a graphite moderated pebble bed reactor. I’m not a specialist in reactor design, but I trust the people who have performed the calculations for me.
On an “arrow analysis” level of detail, there is a strong negative temperature coefficient of reactivity that works in the opposite direction of the pressure coefficient that you describe. Both pressure and temperature changes happen relatively slowly (especially in a low power density pebble bed reactor with lots of material inertia). There is sufficient negative feedback to prevent any kind of prompt reaction.
I was talking only about the coolant temperature. Neutron absorbtion is proportional to coolant density, which, at constant pressure, is inversely proportional to temperature.
Coolant heat capacity is small and coolant temperature would rise rapidly if flow is reduced.
Pressure can drop very quickly as a result of a failure in the primary circuit, e.g. a failure of a turbine disc. When that happened on the No 3 engine of that Quantas A380, pressure would have dropped very rapidly indeed.
@RRMeyer
Though there are some similarities between nuclear gas turbines and jet engines, the heat source part of the system is several orders of magnitude larger per unit power. Gas pressure in the large, low power density reactor will not change very rapidly, no matter what happens in the turbine section of the system.
When flow is reduced, the entire core heats up, not just the coolant/working fluid.
This comment regarding the reactivity effect of the nitrogen being worth anywhere near -$1, let alone the quoted -$32, seems to be terribly incorrect and should not stand without a sanity check because it throws doubt on AAE.
Assuming that 10% of a HTGCR core is coolant channel by volume (i.e. void) I’m finding that the SIG(A) for graphite is 10X that of nitrogen at @ 25 bar, 1000C if I homogenize everything and use thermal sig(A).
I’m not seeing a lot of worth in the coolant, which is consistent with Rod’s long time assertion.
People ought to be more careful (RRMeyer).
After further analysis I am estimating about -$0.06/bar N2 with an Isothermal Temperature Coefficient (ITC) of -$0.005/deg-C for a cold, clean, new, isothermal condition. So, a loss of coolant from 10 bar would insert about +$0.60 and a loss of coolant from 20 bar would insert about +$1.20. So, to some first-order approximation, the LOCA from 20 bar will quickly increase average temperature by 240 deg-C. Maybe it would overshoot, or maybe not; that would depend on the rate of pressure loss.
The model analyzed was a 3m diameter right cylinder filled with homogenized graphite, uranium carbide, and nitrogen reflected by 6″ of stainless steel all-around. Enrichment was 19.5% and core void fraction (pebble gaps filled with N2) was 35%. Operating point was 927 deg-C and 10 bar N2.
Follow on post LOCA temp rise will bring RX to zero (shut down).
Obviously there is a lot more gas inventory in a reactor than a jet engine, but it shows that a sudden big leak in the turbine is a plausible scenario.
Let’s put some numbers on the speed of inventory loss. The HTR-PM has 78 m^3 core volume, 47 m^3 of which is taken up by the 420000 6cm spheres. Leaves 31m^3 gas. If that was N2 at an average temp of 500 deg C and 15 bar, that would be about 190 kg N2. The 250 MWth heats that by 500 K in just 0.4 seconds, so the inventory has to be recirculated 2.5 times per second.
If something happens to the turbine whereby half the usual flow is lost instead of recirculated, it would take just 0.04 seconds to lose 5% of the inventory and inject a $ of reactivity. Followed by another, and another, for a total of 20$ reactivity insertion.
The most challenging exercise calculated for the HTR-PM is a total control rod ejection, where all control rods are completely ejected in 0.1s starting from the normal full power position (which is already close to the top). This inserts about 1.3% reactivity, or 2$.
If using N2 as a coolant, a loss of coolant at half the normal full power flow rate would insert 10 times as much reactivity at a faster rate. This is too big to be reined in by the negative fuel temperature feedback, and would lead to a massive power excursion.
A key safety feature of pebble bed reactors is the very low reserve reactivity (slide 15). This is made possible by the continuous re-fuelling.
Nitrogen coolant would completely destroy the safety case by introducing a large reactivity reserve that can be easily and quickly released.
‘Semiheavy water, HDO, exists whenever there is water with light hydrogen (protium, 1H) and deuterium (D or 2
H) in the mix. This is because hydrogen atoms (hydrogen-1 and deuterium) are rapidly exchanged between water molecules. Water containing 50% H and 50% D in its hydrogen actually contains about 50% HDO and 25% each of H2O and D2O, in dynamic equilibrium.’ (Wiki)
‘Heavy Nitrogen’ – N15 – has a neutron absorption cross-section of only 0.000024 barns, versus the natural mix of nitrogen isotopes at 1.9 barns. Since the two atoms in N2 are tightly bound with a triple bond, rather than swapping around promiscuously like the protons in water, enrichment would have to work on ammonia, NH3. Heavy Nitrogen is about twenty times as common as heavy Hydrogen in nature – 0.37 % versus 0.015% – but the mass difference being selected for is only about 5% versus 11% for heavy water. Deuterium in the ammonia would skew the results a little, but not enough to affect the neutronics. So at a guess, if scaled up, the cost of 2N15 should not be too dissimilar to that of heavy water. But a Candu reactor needs five hundred tons of liquid to fill its calandria and operating circuit, whereas you estimate a pebble bed reactor using nitrogen would only need 190 kg of gas. So an initial cost to set up the enrichment system, and for each reactor’s inventory, but not a deal breaker.
I estimate about -$0.60 at 10 bar and -$1.20 at 20 bar N2. Big, but not so strong as to override strong temperature feedback in LOCA.
Thanks for the response, E-P. It is of typical “E-P Quality” (which is to say “both thoughtful and useful”).
That’s high praise in this crowd, Rich. Thank you.
I can’t hold a candle to Feynman but I definitely qualify as a curious character. I wondered if I could figure out what the properties of a Ne:CO2 or Ne:SO2 mixture of Cp/Cv of 1.400 would be.
As I suspected, the Cp/Cv of the triatomic gases CO2 and SO2 are so close as to be indistinguishable at 1.289 and 1.29. Neon is of course 1.667.
Something few people know is that Cp-Cv = R, the gas constant. Knowing the ratio of specific heats also gives you the values of both of them. How neat is that? So what we probably want is a gas mixture with a Cv of about 2.5 R.
Noble gases are about 1.5 R. CO2 and SO2 are about 3.46 R. It seems to me that a mixture of slightly more than 50% of either tri-atomic gas with the balance neon would have the desired Cp/Cv to replace a diatomic gas in OTS turbomachinery.
50% SO2 in neon: MW 42.
50% CO2 in neon: MW 32. This is VERY close to air; almost eerily so.
@E-P
On a practical note, can you estimate the cost of the quantity of neon needed for a 50% mixture in a system with gas inventory requirement approximating 200 m^2 at STP?
Then help me figure out how much it will cost to supply machines produced at a commercially useful rate in the range of between 10 and 100 each year. If some of these machines are at sea or in other remote areas, how much will it cost to maintain an additional inventory to make up for losses from leakage or to enable maintenance that might require machinery access and loss of gas.
Note: I’m not actually asking for calculation assistance, just attempting some humor in explaining why I remain focused on using N2 (or air) rather than some kind of more complex mixture of gases.
Well, Rod, I just gave it a look-see. The one price I could find on-line appears to be 33 rupees per liter. At the current exchange rate that’s about 45 cents per liter, so 100 scm would cost you $45,000. I bet this is a drop in the bucket compared to the cost of fuel and other hardware.
If you need to do maintenance, you can pump the gas into tanks for storage. You can flush with the cheaper component to remove air before going back into service. Neon will separate easily from triatomic molecules, particularly those with high boiling/freezing points. Note that this is not so different from what’s already done today with common refrigerants; recovery and reuse is the norm.
A serious subject deserves to be treated seriously. What are the costs of dealing with C-14 contamination and the effects on neutron economy?
Look at it this way: if you find that nitrogen is too much of a headache, you’ve now got a pretty solid backup position.
Alibaba has price of USD 0.2-0.4 / litre in bulk https://www.alibaba.com/showroom/neon-gas.html
Thanks @John ONeill. Pure N15 would certainly be great for neutronics, but no longer simple.
Also the question is: For which benefit? It is not like air or N2 is a particularly good working fluid. Water/steam is much better. The reason for the price advantage of gas turbines is internal combustion, which removes the need for a heat exchanger between the heat source and the working fluid.
Just compare the huge htr-pm pressure vessels (at 70 bar to get sufficient heat transfer at acceptable flow rate) with the tiny combustion chamber of a gas turbine.
Nobody would use air as working medium where the heat is not generated in the air itself by combustion.
CO2 is much better yet on some specifications; it is especially good for gas turbines as the compression can be done near the critical point which seriously reduces the temperature rise and thus the compressor back-work.
The problem with that is that there is no OTS (off the shelf) turbomachinery for CO2 systems (yet), and establishing a supply chain for it is a billion-dollar investment. It may be that Toshiba (?) is doing this as a partner with NetPower. If so, such turbomachinery will be available for other users. Time will tell.
@E-P
Once a particular compressor and turbine set has been designed, manufactured and proven using CO2, it might be worthwhile to use it as the basis for a new power plant design.
The interesting thing – to me – about choosing to work on nuclear heated systems that can fit within the limitations of proven combustion turbine design parameters is that there is already a massive supply chain in existence for machines in a wide range of shapes, sizes and specialty applications. There are tens of thousands of qualified mechanics, spare parts bins, etc. As the title of this piece attempts to indicate, the idea is to copy a successful business model and make it rather boring for machinery designers while producing acceptable profits and massive quantities of clean power.
Critical pressure of CO2 is 78 bar, so at the 7-10 bar Rod is aiming for, you would ger none of these benefits. Critical temperature is as low as 31 deg C, so the lower pressure of proposed SCO2 systems is 78 bar, the higher something like 200 bar.
Using that as direct reactor coolant would be ambitious because of the conflicting aims of making the vessel small enough to be buildable and making the power density small enough to allow decay hear removal in case of pressure loss.
Critical pressure of CO2 is 78 bar, so at the 7-10 bar Rod is aiming for, you would ger none of these benefits. Critical temperature is as low as 31 deg C, so the lower pressure of proposed SCO2 systems is 78 bar, the higher something like 200 bar.
Using that as direct reactor coolant would be ambitious because of the conflicting aims of making the vessel small enough to be buildable and making the power density small enough to allow decay hear removal in case of pressure loss.
Good point re: pressure but note that those pressures are still only a fraction of what’s in every-day use for vehicular hydrogen storage in FCEVs. Long story short, it comes down to money and some comparison of values at the bottom of a spreadsheet is going to determine the pics.
@RRMeyer
Many of our most useful inventions were the result of people who actively chose to become “nobody” in the sense of making decisions that others have proclaimed that “nobody” would make.
How much hands-on experience do you have in designing, building, operating, maintaining and paying for steam plants?
I agree that helium at 70 bar presents some design and cost challenges, but the considerations are different for nitrogen at 7-10 bar in the heat exchanger (reactor) portion of the Brayton and roughly 1 bar in the heat rejection portion. Obviously pressures in the compressor and turbine sections of the system transition from the low to the high and back down from the high to the low.
Flow cross sections have to be much larger in the ducting and heat exchangers than they are in a combustion turbine, but they are similar in the turbine and compressor as similar output combustion machines.
The reactor is large but the ratio of graphite to heavy metal is about 20:1. The reactor walls are more comparable in size and cost to walls in a steam plant furnace than to the pressure vessels used in light water reactors. The heat rejection portion of the system is closer in size and cost to air handlers in industrial air conditioners than to condensers in a steam plant. There isn’t much cost associated with obtaining or storing the ultimate heat sink or piping it to the plant. It’s just the surrounding air.
The large mass of graphite in the core performs more than just moderation duties. It is configured to contain the heavy metal and fission products and it provides a substantial amount of thermal inertia that allows the system to rapidly respond to changes in power demand.
I freely acknowledge that there are many alternative ways to convert heat into power and that the system I’m describing has numerous disadvantages and trade-offs compared to other proposed designs. It’s not the be all and end all. But it might point out a way to address the biggest current hurdles that slow nuclear development. It might not.
Thanks for your detailed answer. I have ni experience in power engineering, but I ca see with great distreas that coal plants are stull being built all over the world, so it would appear that it is not the steam cycle that is holding back nuclear.
Why would N2 at 7-10 bar be as good at removing heat as He is at 70 bar? The Chinese did not go for such a high pressure for the fun of it, and given the size of yh vessel, it must be a real headache.
In addition to the information John offered, consider fact that heat transfer is also a function of temperature differences.
The rate of heat transfer from core materials to gas improves when the entering gas is much cooler than the core. The macroscopic heat transfer from the core to the flowing gas is also greater when there is a bigger temperature rise across the core.
Simple cycle, non recuperated Brayton cycle machines have a bigger heat rise than recuperated cycles. Brayton cycle machines have a bigger temp drop across the turbine than a two loop system across a heat exchanger.
Rod, I don’t see your argument for un-recuperated machines here.
1. You still have to do something with the heat between the turbine and the compressor. In a closed-cycle machine you’ve got to move it through a heat exchanger whether you recuperate or dump it.
2. The recuperated machine has higher thermal efficiency, so supports higher output power for the same size reactor.
3. If fuel costs dominate, losing thermal efficiency means higher operating cost.
Re: Look at it this way: if you find that nitrogen is too much of a headache, you’ve now got a pretty solid backup position.
A suitable gas blend could also serve as an introductory application of the Adams Engine. After a few binary gas cooled reactors are operational, the proven technology could inspire confidence for an investment in an industrial scale N-15 isotope separation plant.
With all of its ammonia production, Trinidad and Tobago could see N-15 as a lucrative value-added product.
‘Why would N2 at 7-10 bar be as good at removing heat as He is at 70 bar?’
Specific heat of helium is 3.12 kJ/kg/degree Kelvin, for nitrogen it’s only 0.743. Multiply that by the atomic, or molecular, weight, though, and for helium it’s 6.24 , for nitrogen 15 it’s 22.29, three and a half times higher.
From my understanding, at the same temperature and pressure, with different gases, the same average number of molecules will be impacting the walls of the chamber, but the lighter molecules make up by hitting with higher average momentum. So if you substitute 2N15 at 10 bar for He2 at 70 bar, the ratio of average molecule impact masses per area would go from 140 to 300. The hot fuel would kick each incoming molecule away again with increased energy, to contribute most of the heat flow – infra red is much less important.
I am slow, but I finally realized your idea could be extended to replaceable cartridge molten-salt reactors like Terrestrial or ThorCon. Instead of just fuel as the consumable “razor blade” type sale, make the fuel, the reactor vessel, the primary heat exchanger, and the first air-tight barrier the consumable. Now the initial cost of the plant does not include any of those items (great way to reduce initial capital cost).
If the company could swing the cartridge finance, utilities could be charged monthly for electricity/heat produced. The company would have skin-in-the-game incentivizing cartridge improvements. The fuel would be owned by the company. Recycling/disposal is the company’s responsibility. Multiple countries could be serviced by the company.
Recurring income is the best defense against recessions or really bad PR. Disclaimer: I lost my job during all but one of the recessions in my lifetime.
Martin – I like the way you think. Yes, your model is a reasonable analog to the fuel based model I proposed.
One difference is that the consumable section is less flexible and less amenable to true mass production on the scale of marbles or M & Ms.
I believe you are looking for a business model that is not only sustainable but likely to improve the technology plus support advertising comparable to “natural” gas advertising. The business model must have a very large recurring income. The measure of “very large” is not the number of items produced but the percent of the industry recurring income going to this business.
Leasing a cartridge perhaps based on a few cents per kh while providing transportation, installation, cartridge use, removal, and refurb/recycle/disposal seems like a wonderful deal to both the utility and the business and the industry.
I do not see how to create this business but I certainly will cheer if someone can. I won’t even complain if the business creates a few billionaires.