...It's paradox. They left us these technological marvels, yet with all their might and knowledge they failed to prevent their own doom...

Excerpt from a lecture by the High Historian of Berlin Falls

Welcome to a future. Mankind has brought doom upon themselves, their cities have been flattened by war and weather, and most of the northern hemisphere is radioactive badlands.

They've managed to avert revert global warming, in the process creating a global society of an unprecedented scale. Air-travel got reduced to the bare necessities, and the railways underwent a renaissance.

Closed stretches of track & stations all over Europe got reopened. Lines between bigger cities got extended to rail-arteries1. These arteries expanded to stretch all along the northern hemisphere, even connecting Berlin to Boston.

 - Central Europe, Russia, Kazakhstan, Beijing, Bering-Strait, Chicago, Boston
 - England, Central Europe, Spain, Gibraltar, Morocco
 - Kazakhstan, Afhganistan, India
 - Beijing, Hong-Kong, Thailand, Indonesia, Papua New Guinea, Sidney
 - Beijing, Korea, Japan
 - Chicago, Mexico, Colombia

Due to the lack of electrification of rails on the American, African and Australian continents scientists and engineers spent considerable efforts into advancing emission-free alternatives to diesel-electric engines.

A break-through was achieved in the field of SRG & RTG technology. Taking a hint from space-engineering, Radioisotope Heater Units (RHU) were expanded in size to serve as continuous heating elements in the boilers of steam-electric engines - these engines would then be used for trains and ships.

A typical train-engine would consist of a boiler in which 5 RHUs (~6 tons of Radium per unit2) superheat water or another conductor fluid in a primary fluid-cycle. Through conduction the heat in the primary cycle is used to superheat water to steam in a secondary water-cycle, which is then used to drive a turbine3 producing electricity.

┌─────────────────────┐     ┌───────────────┐
│ ┌─┐ ┌─┐ ┌─┐ ┌─┐ ┌─┐ ╞══╗|╔╡ STEAM TURBINE │
│ │R│ │R│ │R│ │R│ │R│ │  ║|║└──────────────╥┘
│ │A│ │A│ │A│ │A│ │A│ │  ║|║   ┌─────────┐ ║
│ └─┘ └─┘ └─┘ └─┘ └─┘ ╞⛒╝|╚⛒═╡CONDENSER╞═╝
└─────────────────────┘        └─────────┘

Assuming ~168W per kg of Radium4 and an optimistic efficiency of ~50% this gets us ~2500kW usable energy. Or about ~500kW per RHU.

Q: Is this concept for an engine workable or are there game-breakers that I missed out on?

A good answer:

  • addresses issues with the proposed design
  • proposes solutions to the addressed issues

1Lines with sections of up to 8 tracks next to each other in order to facilitate higher throughput. Sort of superhighways but for trains.
2Which results in cylinders of ~1m diameter and ~2m height (assuming we do not have to interleave the radium with too much other metal to get the heat out efficiently)
3Similarly to how a Nuclear Power Plant works, also known as Rankine Cycle.
4I've not too much knowledge in the area of nuclear physics, so I designed the described system based on the very helpful explanations I got from @kingledion on the chat.

  • 1
    $\begingroup$ But why would anyone do this rather than electrify all the existing & new rail lines, & generate the electricity with stationary nuclear plants near the rail lines? $\endgroup$
    – Jim Baerg
    Feb 18, 2020 at 1:09
  • $\begingroup$ Small reactors for ships crossing oceans would make sense & maybe the radioisotope power source would be good for relatively small boats. See templar.co.uk/downloads/Beyond_Fossil_Fuels.pdf for a discussion of such ideas. $\endgroup$
    – Jim Baerg
    Feb 18, 2020 at 1:11
  • $\begingroup$ @JimBaerg all the assumptions and absolutes for the setting are given in the question. Also as this is a set-up for a post-apoc setting the point is to have these desired engines available post-apoc, and not have a blown-out, unusable electrical setup $\endgroup$
    – dot_Sp0T
    Feb 18, 2020 at 8:27

2 Answers 2


A radiothermal train is a fun idea. Radium may be possible (given futuristic resources), though tricky, I'll need to give it more thought. Another fun possibility if you don't mind an actively controlled reactor is a natural uranium source, like in the CANDU reactors in Canada. For passive systems though, I'm going to make my case for Polonium-210.

One of the bigger issues of Radium and other seemingly suitable isotopes is the byproducts (or daughters), which are created in that isotope's decay chain. These byproducts would build up constantly in normal use and create technical problems or hazardous conditions. Keep an eye out for long-lived (= obnoxious) daughters, and any beta or gamma decays. To my knowledge, all of the common RTG isotopes have obnoxious byproducts, sadly, with one exception.

If you're willing to hand-wave the production of the isotope, 210Po is rather ideal. It decays directly to a stable isotope of lead-208 through a low-ish energy alpha decay which is quite easy to catch and generate heat from. The alpha particles mostly stop with only a few cm of air, or completely stop in a few mm of water. Polonium-210 has an extreme activity, allowing smaller weights of it to be viable, even on a large train. From my back-of-the-envelope calculations, it would take 'only' about 40-70kg per train (an amount which would fill a ~6in cube). As a note, producing this with modern technology is flatly impossible. If it's easier in the future though...

One convenience of Polonium is that it's fairly noble as a metal, in other words it doesn't rust or dissolve well in water unless there's a bunch of chloride in the solution. For this reason, it could be used in small chunks or even possibly as a thin (mm) plating of solid metal in the boiler or on the walls of the heat exchanger. If you wanted to use it in water solution instead, it's soluble in 3%HCl, or more simply in EDTA (like in shampoo). Without either HCl or EDTA (or similar) it automatically tries to plate itself out of water solutions as metallic Polonium. A downside of HCl is it attacks Copper and Iron, common heat exchanger materials.

A radiothermal train would be producing heat constantly, so it would need to be constantly boiling water, even when not moving. This could be very obnoxious, because water tends to leave deposits and scales on boiler / heatEx equipment and without an obvious opportunity to clean them they could become ineffective or (frighteningly in this case) leaky.

Because Polonium-210 has an extraordinary activity and decent bioabsorption, it's a nefarious toxin (worse than cyanide). The good news is it has a very short halflife (138 days), so spills would lose toxicity within decade. The immediate effects would be devastating though, and would travel freely with ground water if using the solution form. Also because of Polonium's short halflife, the trains would need to be refilled regularly, perhaps making the 'chunk' version more appealing.

Lastly, the alpha particles would cause wild amounts of embrittlement in any metals within striking distance. The Polonium would need to be held well separate from anything critical.

I think I could keep going for a while about little quibbling engineering problems a radiothermal train would face, but it's wonderful concept that doesn't break any major laws of physics while having just enough troubles to make it interesting for the people stuck aboard.

Math for amount of fuel required on a radiothermal steam engine:
2ton coal /hr reference
19.48 *10^6 BTU / ton coal reference
1055 J/BTU , 3600 s/hr
~6MW heat
141W/gm Po-210 reference
42kg Polonium-210 per train.
Polonium's quite dense at ~9g/cm^3 resulting in something like a 17cm cube of polonium. A cube would explode though, so let's assume small chunks or a plating.

Best of luck with your world! Also, I really like your ascii art.

Edit: I noticed that a train to last through a nuclear winter would be handy, so I put a bit more work into this. It seems that Sr-90 would also be appropriate, with the advantage that it has a much longer (29yr) half life. A train would require about 1500kg of it, and beneficially it's available in large amounts in nuclear waste. Sr-90 undergoes beta decay (medium-nasty radiation) to make a daughter which also quickly beta decays to something nice and stable. Beta radiation is best caught by acrylic plastic, though a few cm of water would work as well.


Problem: You will not get 50% efficiency. Nuke plants run at about 30% for safety reasons, the same rules would apply to your trains. Thus you need to increase your powerplant by 50%.

Clarifying this: Your efficiency is limited to the Carnot limit, which is a function of temperature. Since nuclear power doesn't have an inherent limiting factor like combustion-driven power you need to keep the temperature farther away from the point your system breaks--for fission plants that ends up being 30%, I would figure the same factors would be at work here and thus the same limit.

Problem: You also have to dissipate 5,000kw of waste heat. Major woe if anything goes wrong with your cooling system as there's no off switch possible. Expect any serious train accident to turn into a nuclear accident as your system bakes itself.

  • $\begingroup$ I am not sure I follow your reasoning: Do I not get 50% due to technical constraints or due to ideological constraints? $\endgroup$
    – dot_Sp0T
    Jan 3, 2018 at 8:58
  • 1
    $\begingroup$ Actually you get 30% because no thermal system can operate beyond the Carnot limit. Some systems "seem" to get around it because they are essentially ganging up two different Carnot cycle engines, for example a gas turbine generator who's exhaust is used to power the boiler for a steam turbine can show a 60% efficiency overall. Non Carnot systems like fuel cells or MHD generators can exceed these limits, but we are not talking about this here. $\endgroup$
    – Thucydides
    Jan 3, 2018 at 21:35
  • 1
    $\begingroup$ You are assessing a nuclear plant's efficiency, not a radioisotope generator's efficiency. Depending on design characteristics, there is no reason that a radioisotope steam generator will have the same constraints. A supercritical water steam cycle (say, 1000 K + at the hot reservoir) will give you near 50% efficiency, even without any regenerative heating components. @Thucydides the Carnot limit for such an engine would be about 70%. 50% is reasonable. $\endgroup$
    – kingledion
    Jan 4, 2018 at 1:30
  • $\begingroup$ @kingledion I'm figuring they won't want to run the radium generator hotter than we choose to run our fission generators. I don't know what temperature they picked, just that fission plants are held back by this--we get better efficiency from fossil fuel plants because we run them hotter. $\endgroup$ Jan 4, 2018 at 5:24
  • $\begingroup$ The Carnot cycle is not related to the heat source, using wood or antimatter to power a Carnot engine (i.e raising steam to run a generator) makes no difference: infogalactic.com/info/Carnot_cycle $\endgroup$
    – Thucydides
    Jan 4, 2018 at 5:29

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .