Feasibility of a Nuclear-Powered Locomotive

[lpetrich]

In another messageboard some time ago, that question came up: how feasible it is to power a locomotive with a nuclear reactor. Yes, that kind of railroad vehicle.

It would work much like a diesel-electric locomotive. That kind of locomotive works by a diesel engine turning a generator, which in turn powers motors which turn the wheels. In effect, an electric transmission. A nuclear-powered locomotive would have a nuclear reactor instead of the diesel engine.

A typical diesel-electric locomotive is EMD SD40-2 - Wikipedia
Length: 68 ft 10 in / 21.0 m
Width: 10 ft 3 in / 3.1 m
Height: 15 ft 7 in / 4.8 m
Weight: 184 US tons / 170 metric tons
Engine power: 3000 hp / 2240 kW
Engine weight: 18.2 US tons / 16.5 metric tons

Would it be possible to have a nuclear-powered locomotive that (1) fits in the above size and has less than double the above weight or (2) has tolerably low ionizing-radiation emissions, or (3) both?

 

[mathman]

I believe there would be a practical problem to make it small enough (including shielding). Also there are potential dangers of sabotage, accidents, etc.

 

[lpetrich]

Yes, the mean free paths of particles released by nuclear reactors will give minimum sizes for the reactor itself and for its shielding. But I've found it hard to find good numbers, at least unclassified numbers. I've found some discussions of small nuclear reactors, but I can't find any with power outputs less than about 25 megawatts, which is 10 times too much for a locomotive.

 

[etudiant]

Afaik, the US has experimented with small reactors for remote sites.
In particular, the McMurdo base in Antarctica was equipped with a PM3A 1.5 megawatt reactor built by the Martin Company in 1963. It was not a long term success and was removed in 1973. This was despite the well recognized hazards associated with the reliance on diesel fuel, including several serious fires.
I do not have any specifics of the design or its fuel load, but it shows there is some prior art.
As the reactor was shipped to the base, which did not have elaborate facilities, it was probably fairly lightweight.
My guess is that for a locomotive, the problem would be in the railyards, where the engines are cheek by jowl, with people active on the sides of the tracks. That does not leave much space for shielding, yet the reactor would have to be active to allow the locomotive to remain mobile.

 

[berkeman]

Wouldn't you need a pretty good "cold reservoir" of some sort to eliminate your waste heat? That's why nuclear power plants are located next to large bodies of water (or under them, as in nuclear powered submarines). I'm not sure an air-based radiator system would work very well, but I'm no expert in this kind of thing.

 

[lpetrich]

Thanx, etudiant, for your hint. I found Camp Century, about a base in Greenland that was powered by a PM-2A small nuclear reactor. It had a 2-megawatt output, and it used 93% enriched uranium. I think that that enrichment shows that it was on the small end of feasible reactors.

That page contains a picture of that reactor. Its core is about 0.5 m across, and its pressure vessel about 1 m across. So PM-2A could easily fit inside a locomotive, and its power rating is appropriate. However, PM-2A and its various additional machinery and plumbing and parts had weighed in at about 400 tons. Judging from its size, PM-2A itself had likely weighed in at somewhere around 10 tons, so much of the extra stuff was likely a turbogenerator and a heat exchanger.

Project Iceworm - Wikipedia, Army Nuclear Power Program - Wikipedia

 

[Astronuc]

In another messageboard some time ago, that question came up: how feasible it is to power a locomotive with a nuclear reactor. Yes, that kind of railroad vehicle.

It would work much like a diesel-electric locomotive. That kind of locomotive works by a diesel engine turning a generator, which in turn powers motors which turn the wheels. In effect, an electric transmission. A nuclear-powered locomotive would have a nuclear reactor instead of the diesel engine.

A typical diesel-electric locomotive is EMD SD40-2 - Wikipedia
Length: 68 ft 10 in / 21.0 m
Width: 10 ft 3 in / 3.1 m
Height: 15 ft 7 in / 4.8 m
Weight: 184 US tons / 170 metric tons
Engine power: 3000 hp / 2240 kW
Engine weight: 18.2 US tons / 16.5 metric tons

Would it be possible to have a nuclear-powered locomotive that (1) fits in the above size and has less than double the above weight or (2) has tolerably low ionizing-radiation emissions, or (3) both?

The SD40/40-2 (3000 hp/2.25 MW) was a standard in the late 70s and 80s. The typical units these days are the SD70 (4000 hp/3 MW) or Dash 9-44CW/ES44 (4400 hp/3.3 MW). With an efficiency of ~0.33, a nuclear system would require a 9-10 MWt reactor. The core would have to be compact, so highly enriched. Shielding could be a challenge. The hoods are about 6-7 feet wide with catwalks on either side for a total with of ~10 ft.

I would imagine that it would have to use a Brayton cycle rather than a Rankine cycle, but that would require some basic analysis to verify.

The safety case would have to address impact loads related to collisions and derailments.

 

[lpetrich]

Wouldn't you need a pretty good "cold reservoir" of some sort to eliminate your waste heat? That's why nuclear power plants are located next to large bodies of water (or under them, as in nuclear powered submarines). I'm not sure an air-based radiator system would work very well, but I'm no expert in this kind of thing.

Steam locomotives had this problem also, and the usual solution with them was not to try, to release the used steam. That made them very thirsty.

Some steam locomotives have been able to recover their water, however: Condensing steam locomotive - Wikipedia Some of their tender cars had big radiators in them. So a closed-cycle steam system may be feasible.

While on the subject, it's worth noting that most steam locomotives have had direct-drive piston engines. There were a few built that had reduction gearing and/or steam turbines, however. Direct-drive piston engines may have had some engineering reasons in their favor, but they would require a lot of re-learning for a present-day locomotive maker.

 

[lpetrich]

I would imagine that it would have to use a Brayton cycle rather than a Rankine cycle, but that would require some basic analysis to verify.

I looked it up:

Rankine cycle - closed-cycle steam engine: piston or turbine
Brayton cycle - air compressor turbine - heater / combustor - expander turbine

 

[Pkruse]

I’m not a nuclear engineer and have no inside information in the field. But I have observed some things that make it seem possible to power a locomotive with nuclear power. Perhaps some of you who are more knowledgeable concerning nuclear power can help straighten out my thinking.

The basic design of all the nuclear plants we have built in the past is a very old and obsolete. By now we have figured out that they will eventually melt down, which is a major bad thing. I was told by a nuclear engineer that the reason for selecting that design concept back then was that it was highly desired at the time to produce bomb material as a byproduct. They had other options, but they did not pursue them. We could be recycling the spent fuel, but our Peanut Farmer signed a law making that illegal because a byproduct of recycling the fuel is bomb grade Plutonium. So now we argue endlessly about how and where to store it for many generations.

Now General Atomics is actively marketing two new products that they have developed. One is a new design nuclear power plant that has a zero probability of a meltdown. The controls are all passive and depend on the laws of physics to remain safe with no human intervention in the event something goes wrong. It uses Helium as a working fluid for a gas turbine and has no water in the system to flash off. Their second product is a plant that uses our spent fuel currently in storage to get an extra 35 years of power with equal safety, and when that is used up the spent fuel is safe and does not have to be stored like the old system. Looking at their design on their web site, I believe it could be fit into a locomotive. Perhaps some of you can tell me what I’m overlooking.

Then there are the RTG’s that NASA has always used on their deep space probes. They weigh about 400 pounds and are about the size of an oil drum. They produce large amounts of electrical power for many years. I don’t know how they work, but they seem to me to be very suitable for a locomotive. Again, if anyone can shed some light on this for me so I can see what I’m missing, then I’d appreciate that.

Harris and GE have developed new distributed power systems for rail use. They can’t put more than three locomotives in front of the train because the traction force by more would damage the tracks. This has always been the limit on the length of a train. They can’t pull more cars than three locomotives can handle. But these new systems permit an unlimited number of locomotives, spaced about every 50 cars. For the amount of power I see talked about in this thread, we need to talk about powering several locomotives in a distributed power system.

I see lots of questions like this about various subjects in these forums, where someone wants to know if something new is possible. Then everyone starts telling them why current old technology can’t do that. How much better it would be to then identify what new products we would need to make it possible, and then brainstorm means of developing it. But in this case, perhaps all the new technology is already here.

 

[jim hardy]

an old friend of mine worked at that Antarctic power plant.
It grew out of a military project (of course) that put a small nuclear powered electric generator on a flatbed truck trailer.

So your locomotive would be technically quite feasible. The weight of shielding would help give it traction.

But practically , I'm sure it would be unpalatable to many people to have operating reactors hurtling down railroad tracks. And look how it would complicate maintenance...
I think it would be a tough sell to the railroads.


There was an atomic aircraft engine built in the 1950's. It was test run suspended from a tower and was loaded into bombay of a large bomber (B47 i think) but never propelled an airplane.

 

[etudiant]

an old friend of mine worked at that Antarctic power plant.
It grew out of a military project (of course) that put a small nuclear powered electric generator on a flatbed truck trailer.

So your locomotive would be technically quite feasible. The weight of shielding would help give it traction.

But practically , I'm sure it would be unpalatable to many people to have operating reactors hurtling down railroad tracks. And look how it would complicate maintenance...
I think it would be a tough sell to the railroads.


There was an atomic aircraft engine built in the 1950's. It was test run suspended from a tower and was loaded into bombay of a large bomber (B47 i think) but never propelled an airplane.

If memory serves, the portable nuclear generator was not designed to be used while in motion. Rather, it was built to be easily transported to the location where it would be put into service. So shielding could be achieved simply by setting it up a few hundred feet from the people involved. The locomotive engine however would brush by train workers and travellers a few feet away when in use, so it might be rather more challenging to shield.
I do know the Nuclear Powered Aircraft carried a shielded crew capsule at a substantial distance from the engines and that maintenance and service access near the engines, for changing the tires or loading the bomb bay, was an unsolved issue.

 

[Pkruse]

Why talk about what we did a long time ago that did not work out?

Let’s talk about what has worked and is currently powering a number of spacecraft that travel away from the sun such that too little solar power is available.

These are small light weight devices that put out a great deal of electricity for a very long time, and which can be handled safely in a normal shop environment without radiographic hazards to the people. Yes, the ones installed in spacecraft put out only a tiny fraction of the amount of power required for a locomotive, but does anyone know how effectively they can be scaled up to locomotive size?

These would be the Radioisotope Thermoelectric Generator (RTG) or the Stirling Radioisotope Generator (SRG). The first uses thermocouples to derive electrical power from a heat source, while the second uses a Stirling engine for the same purpose. I would propose even a third scheme like that currently marketed on the General Atomics website that uses a similar nuclear heat source to replace the combustor in a gas turbine. They are talking about scales up to 65 MW, which would have to be scaled down for a locomotive. All of these are free from the normal melt down concerns of the reactors currently working around the world.

Does anyone on this forum have enough knowledge on these devices to help educate the rest of us?

 

[lpetrich]

I found this document on RTG's

Their figures for a GPHS RTG:
Diameter: 42.7 cm
Length: 113.1 cm
Total mass: 56.0 kg
PuO2 mass: 11.3 kg
Heat output: 4.5 kW

It uses Pu-238

Scaling up to 2.25 MW yields 28 metric tons and 103 m^3, using rectangular bounding boxes. So RTG's could fit.

According to GPHS-RTG - Wikipedia, they produce only 300 watts of electricity, implying a thermal efficiency of 6.7%. That's because they use thermocouples, a very low-maintenance way of generating electricity. Diesel engines can achieve much higher thermal efficiency, as much as 50%.

Scaling up a further factor of 10 would yield excessively high size and weight for a locomotive.

 

[aquitaine]

Perhaps a good reference would be the SAFE reactor or the proposed Mars Surface Reactor.

 

[mfb]

Now General Atomics is actively marketing two new products that they have developed. One is a new design nuclear power plant that has a zero probability of a meltdown. The controls are all passive and depend on the laws of physics to remain safe with no human intervention in the event something goes wrong.

Where does this claim come from? The company which wants to sell these, or an independent analysis?

Their second product is a plant that uses our spent fuel currently in storage to get an extra 35 years of power with equal safety, and when that is used up the spent fuel is safe and does not have to be stored like the old system.

Same question here. http://www.ga.com/energy/em2/? The presentation just shows a collection of things they would like to achieve. Nice, but not really interesting without any prototype. And it would still have waste. As it does not use any external neutron source, I would expect that it cannot fission a significant amount of transuranium elements - the most problematic part of nuclear waste. Reducing the mass of nuclear waste is easy and can be done with chemical methods. However, this does not reduce the activity. It increases the activity per mass in the remaining waste.

 

[jim hardy]

One is a new design nuclear power plant that has a zero probability of a meltdown. The controls are all passive and depend on the laws of physics to remain safe with no human intervention in the event something goes wrong.

This might be the hydrided fuel design that Freeman Dyson described in his book "Disturbing the Universe".
Dyson is just about the last living Manhattan Project scientist and still sharp as a tack. Watch for his articles and occasional TV appearances, he's just great.

Anyhow GA built reactors with the hydrogen moderator right in the fuel molecules. They can't run away because as the fuel heats it shuts itself down without having to wait for heat transfer out of fuel pellet into surrounding moderator as in LWR's.. They are used for research and training i think TRIGA is their name.
One demonstrator power plant was built, Ft St Vrain, about a mile North of Denver (near Lyons) Colorado.
It was a failure because of bureaucratic and business reasons not scientific ones.

I recommend that book to anyone interested in thoughtful science from a philosophical old scientist.

old jim

 

[Pkruse]

Old Jim;

GA has an updated concept that they are marketing on their website if you would care to look.

 

[jim hardy]

Thanks PK i honestly didnt know they were still around ! Will check 'em out.

old jim

 

[zbret]

I think the idea of a Radioisotope Thermoelectric Generator (RTG) locomotive is interesting, but I wanted to comment on its current feasibility.

The Pu238 used for NASA's RTG is actually so rare now that NASA is basically out of the material. We were buying it from the Russian stockpile but they ran out as well. So, even if we could build this (and were allowed to), the Pu238 is not available. Since it is also desperately needed for any space mission much further than mars, it will also likely be earmarked for that over any terrestrial transportation options in the near future. Still, an interesting idea...

One question I have - is could a LFTR (thorium based) reactor be made small enough? I've heard rumors that one might be built to fit on a truck bed, so why not a locomotive? Thought experiment - what would the smallest LFTR look like? Given its safety features, I wonder if it might just be a reasonable option for a locomotive?

One side note, a LFTR would actually generate a reasonable amount of Pu238 which as I mentioned we disparately need for space exploration, as part of its "waste" products.

 

[rmattila]

The way I see it, railroads are an optimal means of transport for electrification by means of a fixed power line running over the track. In Europe, most trains run by the electricity from the national grid. Taking into account the challenges of radiation protection an the possible consequences of a collision/de-railing, it is not very easy to think how a portable nuclear power plant on a locomotive could be competitive with simply laying the power line over the track and building the power plant at a suitable location without having to consider the safety challenges caused by miniaturization and train crash loads.

 

[lpetrich]

Several European nations have a lot of electrified trackage: railways through europe — network maps and interoperabilty -- they even have electric-locomotive freight service.

http://www.parovoz.com/maps/supermap/index-e.html -- also heavily electrified

North America, by comparison, is pathetic. The only electrified intercity lines are Boston-Washington and Philadelphia-Harrisburg. New York City, northern New Jersey, Philadelphia, and Chicago have some electric suburban lines, but apart from purely urban lines, that's about it.

 

[zbret]

I suspect this idea comes from a few sources - first much of the US track is not electrified. Since electricity and long distances is not a good mix, and much of the US power generation is from coal, the US uses diesel-electric locomotives. They put out about 4 MW each since rail travel is quite efficient with energy, its an OK system. In Europe, the fuel is more expensive and the distances are shorter, so electric trains and tracks make far more sense. I'm not saying its not illogical to do in the US, but other factors need to be considered.

But - what if the locomotive could be replaced with another power source? 4 MW (about the power of a locomotive) seems like something that would work well for the nuclear range. In the case of thorium, it would only require a few grams or so per trip. There are plenty of reasons why it might be suitable, but I thought it was an interesting thought experiment, which I suspect is how this thread began.

 

[wizwom]

From: http://www.alternatewars.com/BBOW/Nuclear/US_Army_Reactors.htm

Mobile Low Power Plant 1 (ML-1)

Reactor Type: Gas Cooled
Cooling Type: Nitrogen to Air
Designer: Aerojet General Corporation
Power Output (Thermal): 3.4 megawatts
Power Output (Electrical): 330 kilowatts at 4,160/2,400 volts; 3 phase, 60 Hz.
Plant Volume: 2,978 ft3
Plant Weight: 38.5 tons
Number of Packages: 3
Plant Cost: $5.5 million (FY63)
Erection Time: 12 Hours
Criticality: 30 March 1961
Shutdown: 1965

Notes: To test a reactor package that was transportable by semitrailers, railroad flatcars, and barges.

330 kilowatts is just 442 HP, so this is a not terribly powerful, but then again, at less than 40 tons, an engine could have two, easily.

 

[zbret]

Interesting, and a LFTR should be lighter more powerful and more efficient.

I went to the link provided and noticed the units seem to go to a fairly high heat output, but don't scale electrically. Just so you know, a modern diesel locomotive is about 3MW (4000hp+):
wiki/List_of_GE_locomotives#Evolution_Series_.28introduced_2005.29 (fill in the link yourself or search locomotive on Wikipedia)

I'm still curious what the feasibility would be based on Gen IV nuclear tech. Thanks for sharing that info though - quite interesting.