I was wondering what it'd take for a fluid to be more efficient than water for steam engines.

My current idea is a fictional fluid with three properties: 1) It boils at 80C, thus requiring less fuel/energy to heat to a boil. 2) It has twice the density of water. Since vapour is always the same size regardless of fluid, this should mean that you can have a smaller boiler with a lot more pressure? The smaller boiler also means it can be lighter. 3) It's highly available within the setting.

There is one problem, though. Water is the (real) fluid with the highest Latent Heat of Evaporation. I'm not sure how important that is for a steam engine.

Would the fluid need to have a higher Latent Heat of Evaporation in order to be a more effective fluid than water for generating steam power?

  • 2
    $\begingroup$ Quote: The heat of vaporization of water is the highest known. Sorry, no dice to fit in real-life physics. Yet, I think that for steam engines it's not the heat of vaporization, but specific heat of steam is what matters, as steam does not condense while expending internal energy. $\endgroup$
    – Vesper
    Commented Jan 11, 2023 at 9:19
  • 5
    $\begingroup$ One of the important questions has to be: how to dispose of the used liquid. Water in steam is easily vented to the air and is not toxic. Water is also easily obtained. CO2 vented to the air will cause serious problems. If you have to keep the liquid contained, then you have a heat distribution problem. $\endgroup$
    – David R
    Commented Jan 11, 2023 at 15:35
  • 2
    $\begingroup$ What if... theoretically, you were able to heat a liquid so much that it would essentially blow up. Like combust? A liquid like oil... which you could combust to drive an engine? Like a combustion engine? $\endgroup$
    – Shadowzee
    Commented Jan 12, 2023 at 3:48
  • 2
    $\begingroup$ You probably mean liquid. “Fluid” means both liquid and gas in contexts where their flow is important. $\endgroup$ Commented Jan 12, 2023 at 11:57
  • 5
    $\begingroup$ One fluid which has been used (frighteningly enough) is mercury. It's theoretically better than water for the Rankine cycle, commercialised up to the 40MW scale. Was never super popular because ... well, guess! But there was even a big Mercury power plant in New Jersey and that state seems just fine. $\endgroup$
    – Dan
    Commented Jan 12, 2023 at 18:38

5 Answers 5


This might possibly be a better question asked of physics.SE, but the problem, uh, boils down to efficiency of heating.

Firstly, it is generally easier to heat a liquid than a gas. Loosely speaking, this is because the thermal conductivity of gasses tends to be lower as a result of their lower density. Liquid water, for example, has more than 10 times the thermal conductivity of steam. You can get the same mass of liquid and a gas to the same temperature, of course, but it can take longer to heat up that gas.

Secondly, the efficiency of a heat engine is proportional to the temperature difference between the hot end (the boiler) and the cold end (the condenser). You want that hot end to be as hot as is practical.

Finally, your engine delivers power by basically moving energy from the hot end to the cold end. If it takes too long to heat up the working fluid, the flow rate will be low. If the heat capacity of the working fluid is too low, the heat energy that can be moved in a given time will be low.

Water hits a sweet spot of heat capacity and boiling point under pressure, and one which is hard to beat. There are lower temperature working fluids which get used for things like geothermal power plants using binary cycles or things using an organic Rankine cycle, because the underlying heat source isn't hot enough to generate enough steam to run a turbine.

This is why modern power generation still frequently use steam as the working fluid, centuries after the first practical steam engines. Look on the bright side though: hydroflurocarbonpunk just doesn't trip off the tongue. It'll never catch on.

(Note that I use a lot of weasel words above, because these things are always more complex than they initially seem, see also supercritical CO2 which can be an efficient working fluid at lower temperatures than steam, though it does require much higher pressures. That sort of thing isn't very steampunk though, so it doesn't fit your specific requirements so well)

edit: forgot to actually respond to your original question, oops

How could a fluid be better than water for steam power?

You working fluid should ideally be:

  • non-toxic
  • non-flammable
  • minimally chemically reactive, even at high temperatures and pressures
  • liquid at ambient temperature (so your engine doesn't congeal when it gets frosty outside)

To be better than water, it should probably have higher thermal conductivity in both the liquid and gas phase, but a lower heat capacity in the gas phase (so adding heat to the gas causes a greater pressure increase), though I'm not sure if there are any practical real-world materials that would fit all these requirements.

I don't think that a higher latent heat of evaporation is necessary or even desirable. It is useful for purely moving heat, but I think it makes it harder to develop pressure and for a steam engine that moves stuff you want plenty of pressure. If the thermal conductivity of the gas was higher, superheaters can work better which might let you lower the latent heat of evaportion and form more gas and high pressures for the same energy, but I'm speculating wildly at this point.

  • 5
    $\begingroup$ @Salli hydrogen is a nightmare because it is flammable or explosive, extremely difficult to seal in anything and causes stuff like hydrogen embrittlement, not to mention how hard it is to get it to condense. Helium makes a good choice for a purely gas-phase working fluid, eg. for a Stirling engine. $\endgroup$ Commented Jan 11, 2023 at 10:09
  • 8
    $\begingroup$ I’d add “sourceable” to your attribute list. Water is ubiquitous. It took huge effort to make gasoline available everywhere for car network. Getting your fluid out into the world to make engines able to keep running may actually drive a lot of your story. $\endgroup$
    – SRM
    Commented Jan 11, 2023 at 16:57
  • 2
    $\begingroup$ I think heat conductivity is totally irrelevant. Internal energy of the working fluid is all what matters in power plants. All transport in power plants is purely advective (with some heavy pumps aiding the natural draft in the circulation), while diffusive phenomena like heat conduction play only a negligeable role. Since liquids can carry more internal energy per unit volume than gases there's still a point in preferring a liquid over a gas on the hot section of the pumping cycle. $\endgroup$
    – 5th decile
    Commented Jan 11, 2023 at 18:54
  • 2
    $\begingroup$ No, the fluid in contact with the steel vessel becomes lighter and rises away through draft, giving room for colder fluid to make contact with the vessel wall. The vessel is made in steel and this the only place where a good heat conductivity is relevant. $\endgroup$
    – 5th decile
    Commented Jan 11, 2023 at 19:04
  • 5
    $\begingroup$ @Vergilius I imagine the water molecules pick up that heat out of a sense of duty. $\endgroup$ Commented Jan 11, 2023 at 19:06

Good idea. Someone had to have had it. I backed into it.

  1. What fluid is denser than water and boils colder than water.

How about bromine?

Density: 3.119 g/dl compared to water at 1 Boiling point: 58C

OK! And here is the patent for the engine from 1982


Bromine-vapour gas engine Abstract The medium of bromine is heated, compressed and expanded after leaving the turbine nozzle. Thereafter, renewed liquefaction of the medium of bromine is performed in a special condensing region. The condensate runs back under the influence of gravity to the bottom trough and is once again heated and evaporated there. This is performed by a permanent circuit similar to that of water. The energy required for heating is supplied to the bottom trough with the aid of heat exchangers. After working dissipation to the turbine, this energy is once again completely destroyed after the expansion and condensation. The essential components of this drive system are 1. spec. weight = 3.14 and 2. boiling point at approx. 59 DEG C. A pressure of 9300 bar can be generated given a design height of 30 m. 1. Energy production without primary energy. No stress on environment. No costs for transport and delivery of oil, coal, gas, uranium or the like. The drive is performed using solar energy, collector greenhouse with large energy roofs - 25 km<2> - waste heat from coal-fired and atomic power stations or, e.g., the heat from refuse incineration plants. Hot springs and geothermal energy are also possible. 2. The plant is suitable in particular for installation in updraught power stations. 3. A further decisive "plus" is the possibility of installation at the site of very expensive large cooling towers. Instead of the destruction of energy in cooling towers, in bromine gas turbines the exhaust heat is forced to do further work and thus converted in an environmentally kind and profitable fashion.

The inventor proposes to use the engine to capture waste heat. The only problem is bromine is poisonous. At least it is not explosive.

  • 7
    $\begingroup$ Bromine is also very reactive, so material selection for the vessel containing the bromine may be especially important. (But maybe it's as simple as a fluoropolymer coating.) The linked patent didn't say much about materials of construction. $\endgroup$
    – Theodore
    Commented Jan 11, 2023 at 20:23
  • $\begingroup$ It's also a lot more expensive than water, and this particular method can't be used on a moving conveyance due to the extra space/weight of the condensation equipment. $\endgroup$ Commented Jan 11, 2023 at 22:52

First some feedback:

  1. It boils at 80C, thus requiring less fuel/energy to heat to a boil.

I don't think there's a reason to want this (apart from danger of burning your hands etc.): your engine will have a low efficiency if you work at or below 80°C. Likewise there's no reason per se to keep the amount of energy to get to the high temperature as small as possible. See further along for more info.

Would the fluid need to have a higher Latent Heat of Evaporation in order to be a more effective fluid than water for generating steam power?

The difference in the amount of internal energy in your working fluid between the hot and cold part of the operation cycle is what's important I think (there's a contribution to this from phase transitions but also just from the ordinary heating in between phase transitions). The greater this difference, the more you can miniaturize your reactor for a given amount of desired power.

Although I'm not a specialist (I'll update this answer, i.e. consider it a work in progress and please edit it to improve it if you happen to be an expert on the subject), I think transcritical CO$_2$ in the real world is already considered 'better' than water (in a kind of theoretical sense, with many unresolved practical problems standing in the way of real applications).

What is clear is that we would a priori prefer a working fluid that we can easily heat to and handle at a very high temperature $T_h$ so that we can get a very good baseline (Carnot) efficiency $1-\frac{T_c}{T_h}$. Moreover, the difference in internal energy per unit volume $\Delta e=e(T_h)-e(T_c)$ for this working fluid is preferrably as large as possible so that we don't need to make our tube diameters and/or flow rates all too big. At first, water seems very good from this perspective because $\Delta e$ gets a big contribution from the latent heat that is requires to boil it from liquid to gas. Unfortunately, what seems to bring it back down is that the vapour phase appears to have a low density (at some given pressure that is, which by the way is also constrained to be not too high in order to not blow the reactor. In practice $150$ bar seems the best we can handle today) compared to some other candidate-fluids like supercritical CO$_2$, so while its "$\Delta e$ per unit of mass" is very good, the "$\Delta e$ per unit of volume" is suddenly much more meh. Another thing seems to be that super-hot vapour seems to be quite reactive and corrosive and stainless steel has to be used to address that problem (but even then...).

Crunching the numbers apparently reveals that CO$_2$ is twice as dense in the relevant high-T, high-P conditions (This statement could not be true if both gases behaved like ideal gases in these conditions, for then the number density and energy density would be equal for both under similar $P$ and $T$, I think). Transcritical CO$_2$ is a bit in a sweet spot in between fluid and gas and seems to remain so far beyond its critical temperature of $30°C$ and this seems to be ideal for the purpose.

Besides potential application in future CSP projects, the Japanese are recently suggesting to use gas-cooled high-temperature nuclear reactors for much of the same reasons that I mentioned.

Have a look at https://en.wikipedia.org/wiki/Rankine_cycle to learn more

  • 3
    $\begingroup$ I think high pressure is overall more of an advantage. Sure, it requires sturdier pipes, but that's made up by generating more torque on the pistons/turbines. With lower pressure, you would need higher volumetric flow rates to generate the same power, requiring basically everything to be bigger. $\endgroup$ Commented Jan 12, 2023 at 13:58
  • 1
    $\begingroup$ I just updated my answer (major overhaul TBH). My framing of using higher pressure to prevent boiling was not really correct, e.g. in actual power plants the boiling happens in the heating vessel, not immediately prior to the turbine room, as one might have inferred from the previous version of my answer. @leftaroundabout: you're absolutely right. $\endgroup$
    – 5th decile
    Commented Jan 12, 2023 at 14:09

The first improvement to steam engine is to split the expansion into smaller stages using a compound steam engine.

In the eighties, it was suggested that you could use some of the waste heat from a steam engine to drive a second engine that used ammonia instead of water. This was proposed for power stations, as the engine would be large and need the economies of scale.

  • 9
    $\begingroup$ Having a second heat engine that runs at a lower temperature is referred to as a bottoming cycle. No giggling at the back, now. $\endgroup$ Commented Jan 11, 2023 at 13:40
  • $\begingroup$ There are also "economizers" in fuel-fired boilers that preheat water for the main boiler in the exhaust stack(s), cooling the exhaust gas and reducing the amount of heat needed to boil the working fluid in the boiler. $\endgroup$
    – Zeiss Ikon
    Commented Jan 11, 2023 at 14:28
  • $\begingroup$ Bottoming cycles are, of course, paired with topping cycles. A term which (now) no longer only refers to the line of motorcycles parked outside the local leather bar every Saturday night, in my head. $\endgroup$
    – FeRD
    Commented Jan 13, 2023 at 9:56

In general you'd want these properties in an ideal working fluid:

  • A boiling point at your cycle's low pressure that is just above the temperature of your heat sink (i.e. cooling water). Condensing steam at 100°C with cooling water of e.g. 15°C loses you 85° which could be used to extract more work from the steam if water had a lower boiling point. A fluid's boiling point depends on the pressure, but too low pressures aren't practical because then you need very big pipes and condensers to move enough mass around.
  • Chemical stability over the entire temperature range. There are organic rankine cycle turbines to extract work from low temperature waste heat which use hydrocarbons as working fluid. Unfortunately the right hydrocarbons are not stable at the high temperatures steam turbines use. The higher the temperature, the more efficient your engine is.
  • I think you'd ideally actually want a very low latent heat of evaporation. In a steam turbine you don't want the steam to condense in the turbine because droplets can damage the blades and condensation lowers the pressure. The condensation happens after the turbines in the condenser, where the latent heat of evaporation/condensation is just wasted and not converted into work.
  • A high expansion ratio when heated. The higher the volume ratio between the working fluid at the peak temperature vs the lowest temperature, the more work you can extract from it per unit mass. This is why we usually use water/steam instead of a gas-only cycle.
  • A high volumetric heat capacity. The higher the heat capacity, the more energy a given volume of fluid can transport per unit time. If we had a magic substance of which 1 liter had all the properties of 10 liters of water, we could scale down a steam engine by a factor 10 (volume-wise) and still get the exact same efficiency and power output of the larger engine using water. Volumetric heat capacity is the most important difference, I think, when using a 10xWater substance.
  • Cheap and non-toxic.

The efficiency of a steam turbine cycle at scale is for a large part determined by the difference between the input temperature (the highest steam temperature) and the condenser temperature set by your cooling water. The peak steam temperature in current steam turbine cycles is limited by the creep temperature of steel alloys, which is around 700-750°C depending on the alloy. Going higher is possible, but that requires non-ferrous high temperature alloys for the boiler and pipes and the first stages of the turbine. Those alloys are much more expensive, so the cost of your engine will be a few times higher. (Gas turbines use this approach, but there you need much less of the expensive alloys because you only need them for the turbine and you don't have a boiler and lots of piping.)

The question doesn't specify what constitutes 'better', so I am assuming you want some combination of efficiency and size/weight of the engine.

As you are not changing the main limiting factors for the peak and lowest temperatures in the cycle, using a different working fluid is not going to change the overall efficiency of your engine much, assuming you prioritize efficiency over size and weight.

The choice of working fluid is mainly going to influence how big and heavy an engine will be, or how efficient it will be if you are constrained in size and/or weight. An analog of water with twice the density would also have twice the volumetric heat capacity so the engine could be smaller. But if the molecules of your fictional working fluid were twice as heavy as water molecules but the same size, their interactions would also be stronger which would translate into a higher boiling point. If you fluid had smaller molecules that had the same or a slightly lower weight as water molecules that would get quite close to your fictional working fluid, and as the number of molecules is higher the gas volume and thus the expansion ratio would also be better.


You must log in to answer this question.

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