Following on from this question which specifically asks about the consequences of aeolipile development, how can my ancient Roman civilization develop effective steam power for use in their ships?

I assume this is just about impossible in our word, but what changes might make this possible? Or at the very least make it fairly easy to suspend disbelief and avoid compromising the whole ancient Rome setting?

This is an earth-like world with basic Roman technology (+steam ships!) There may be other steam powered engines or not, the steam power may be of any type and may be relatively primitive but should be practical (so no aeolipiles). I would rather not use magic as this makes the question trivial.
edit further references added
This article in reddit is very relevant: https://www.reddit.com/r/AskHistorians/comments/1k30h1/could_the_romans_have_build_a_steam_engine_the/

Here is a reference to the transited work by Hero showing fairly extensive knowledge https://www.reddit.com/r/AskHistorians/comments/1k30h1/could_the_romans_have_build_a_steam_engine_the/ http://web.archive.org/web/20120211202905/http://www.history.rochester.edu/steam/hero/section74.html


8 Answers 8


TL;DR - Romans would be able to build some riverboats with atmospheric-pressure engines, the main obstacle is understanding of physics, not some technology.

  1. Building a simplest steam engine is not as hard as people tend to think. Boiler may use atmospheric pressure and be made from copper and lead(upper part) - that's exactly how first boilers were made. Main increase in efficiency of steam engine in 18th century came not from increasing pressure, but from using separate condenser - making it requires understanding of physics, not some high technology.

    Cylinders were made by hand and leaking of steam was prevented by winding rope around piston.

    Wood fuel became quite expensive at late Middle Ages, but in Roman times most of Europe was covered in forest so at northern borders of the Empire fuel was dirt cheap and even in Rome it was mostly cost of transportation that formed price of fuel.

    Also first engines valves were hand-operated, though automating that proved to be easy enough(funny, but it was less fuel efficient that human operator).

  2. Romans had quite an industry dedicated to mine draining so they would pay for draining steam engines. That's how first engines are likely to be used.

  3. So it is technologically possible and economically feasible. The problem is to understand physics enough to make them. For example understanding workings of pressure is not as simple as it may seems - otherwise Magdeburg hemispheres or Pascal's experiment with barrel would not become so famous.

    The notion of latent heat is quite hard to get as well, it's understanding was crucial for Watt's invention of separate condenser.

    The required theory leap would be tremendous, frankly I don't see Ancient Romans making it. But if we assume that they made it somehow or were helped by time-traveler then they can build the engine.

  4. But there is big difference between some Fulton's "North River Steamboat"(100 tons displacement and 20 h.p.) and ocean going steamship(Great Western, 2300 ton, 750 HP).

    Making river-going steamboat is withing Roman technology level. Ocean going steamship requires technological revolution.

P.S. Aeolipile is completely useless because of extremely low efficiency.

  • $\begingroup$ Very good answer. As a related aside I wonder if there is any natural event (or any man made event that could be contrived happen by accident) that would clearly demonstrate either latent heat or the power of a vacuum? $\endgroup$
    – Slarty
    Commented Nov 17, 2017 at 16:42
  • 1
    $\begingroup$ It is easy to demonstrate latent heat - boil water and measure time it takes to boil and time it takes to boil out. This was known long ago, and anybody who knows his way around kitchen knows it intuitively - but he knows it just as empiric rule that it take water so much to boil, so much to boil out, not some abstract knowledge that is easily transferable to thinking about steam engine. To understand it in physics terms and to think about heat as some entity that can be spent by spending fuel is surprisingly hard. $\endgroup$
    – Vashu
    Commented Nov 17, 2017 at 19:58
  • 1
    $\begingroup$ Demonstrating vacuum is easy too - put steam in some vessel, close it, watch it break. But the leap it takes to think about it in abstract terms of force and pressure, to think about air as another form of matter is a little hard to get after you get it. The main problem is that you need several ideas to meet to produce the invention, but those ideas are pretty much useless on its own. $\endgroup$
    – Vashu
    Commented Nov 17, 2017 at 20:05

The truth is they already have it; the Greeks developed the Hero steam engine (a simple sphere on a spit so to speak, that you could put some water in and place over a fire, with 2 exhaust ports opposite each other that allowed the steam to escape and the ball to spin) - they just never developed it further.

Roman ships were sturdy, and they were great traders. There are 3 reasons they didn't develop steam power for their ships;

1) Human labour - why invest in coal, expensive machinery et al when you can get humans to row? (All you need is food, but you already need that for the rest of your crew - a single resource to load up)

2) Sails - wind power was reliable enough in the Mediterranean.

3) Military focus - ships were for trading, not really for fighting. The Romans were more focused on infantry and land fighting

Ultimately, if you want your 'Romans' to get into steam power for their ships, make rowers more expensive and make sea power more important. Having unreliable winds at sea would also help.

Slaves would still be needed for coal mining and the like to support this technology. The real issue with steam power though is converting the energy to propulsion. In Mark Twain's time, this was done with paddle wheels. It wasn't until the end of the 19th century that the 'underwater screw', or propeller was invented.

That said, using steam in military applications would have eventually led to an earlier development of this I'm quite sure.

The other point coming up is the increased cost of metal in ships with steam power. This is a good point and also why Roman mechanisation would have more likely relied on spring technology. Their fifth legion (effectively their engineering corps) used to build large weapons that launched massive arrows that were launched using cords made of bovine tendons that had been dried in place. That kind of energy storage would have been more likely for their ships as well.

  • 3
    $\begingroup$ Or have someone show that mining coal is a more effective way to use labor. If there is an abundance of coal for a large number of purposes… $\endgroup$ Commented Nov 16, 2017 at 20:30
  • $\begingroup$ That's actually a very good point; use the 'free' labour in resource harvesting instead of logistics... $\endgroup$
    – Tim B II
    Commented Nov 16, 2017 at 21:29
  • 2
    $\begingroup$ It is a common misconception that slaves rowed ancient galleys, in actuality you needed a skilled and trained team of oarsmen to carry out complex manoeuvres. The practice of using slaves and convicts to row galleys occurred in the 1500's, when Spanish silver caused inflation and priced oarsmen out of the market, and a brace of cannon and arquebuses aboard substituted for ramming. $\endgroup$
    – Thucydides
    Commented Nov 17, 2017 at 3:51
  • $\begingroup$ the aeolipile is a toy, not an engine. there is no mechanism to take power from its rotation to apply to something that would use it. And it produced so little power that hooking it up to anything that offered resistance would bring it to a halt. $\endgroup$
    – vulcan_
    Commented May 18, 2022 at 21:51
  • $\begingroup$ another correction of Tim B II, the romans were a very strong sea power. They were building it up from about 400 years BC so that after the end of the First Punic war they were able rack up a string of naval battle wins against the Carthaginians, which continued after the 2nd Punic war right up to about 190 BC. After that they kinda owned the Mediterranean so their victories where mostly on land. $\endgroup$
    – vulcan_
    Commented May 18, 2022 at 21:57

This is a very interesting question, because it shows how the world is all interconnected. Could the Romans develop steam engines? Let's look first at the technological requirements and their implications.

Technological requirements

The Romans knew perfectly well that there's force in steam. There were even some crude demonstrative devices, such as the aeolipile mentioned in the question. So what more was needed, from a purely technological point of view?

  • First, they needed fuel. The classical Greco-Roman civilization used relatively little fuel, and that fuel was wood and charcoal. They used little fueld because they lived around the Mediterranean, where winters are mild, and they lived during a warm period. The Mediterranean region has few extensive forests, and it has essentially no mineral coal.

    Since there is no coal around the Mediterranean, the Romans had a very vague idea of what coal was, or how to use it. They knew that in Britain one could find some black rocks which burned, but that's about all they knew about coal. But let's say that after somebody building the first engines, some Roman engineer returning from Britain would have experimented with coal, and the Romans would have pushed northwards to tap the extensive coal reserves in Britain and Gaul.

    So fuel is not such a big deal; it's perfectly possible to imagine Romans mining coal in Britain.

  • Second, and here we come to the really difficult technological problems, they would need metal and metalworking techniques to make the engine. This is a big problem. The only suitable metal available in those times was bronze; bronze is definitely not cheap. Why not cheap iron? Well, the problem is that in Roman times they knew how to cast bronze, but nobody had ever seen molten iron, much less attempted to cast it. Bronze remained the only metal suitable for casting for a thousand years after the fall of the Western empire; that's why cannon were made of bronze up to the dawn of the Modern age.

    Remember that in the classical world metal was used on a much smaller scale than in the Renaissance and the Modern Age. Metal was precious.

    And metalworking techniques were also problematic. The Romans did not have lathes able to machine metal; all they had was hand-powered lathes suitable for turning wood. The first lathes able to machine metal appeared (again) one thousand years after the fall of the Western empire. No lathes means no (reproducible) screws and no way to make accurate cylinders or pistons. Making a steam engine without accurate cylinders and pistons is non-trivial; yes, a crafty artisan could have made one cylinder and one piston, at a great cost, but this does not a navy make.

  • But the killer problem is metal pipes. They just did not have the technology to make high(-ish) pressure metal pipes. The only metal pipes which they knew how to make were very low pressure lead pipes; don't try to pipe live steam through one of those. Why are pipes important? Without high(-ish) pressure pipes one can make only an atmospheric engine, like Newcomen's original (1712). This kind of engine is suitable for pumping water out of mine shafts and little else.

Implications of the technological requirements

One can of course imagine the Romans solving the technological problems. (Or more likely the Greeks, but then by the 2nd century the Greeks were fully integrated in the empire, which they no longer perceived as an alien domination but rather as the natural culmination of the history of Greece.) The question is then, what kind of world would it have been? Would it have preserved the essence of the classical world?

I would say no. The kind of technological advancements we are speaking of imply a world with a highly qualified workforce, with an intense market economy, with systematic technological training, with widespread use of metal parts, with precise measurements... This is not the classical world. It just isn't.

Could they have done it while preserving something of the classical world? Yes, of course! After all, we did it -- we still have senators, who (in the U.S.A.) meet on the Capitol hill, we still have tribunals, and (in many European countries) we still have prefects and quaestors, and our languages are full of Latin words. If you are content with a world which preserves some (maybe most) of the forms of the Roman world but little of its essence, then yes, this is imaginable.

So why didn't they?

But the Romans did not develop the technologies necessary to make steam engines. They did not develop the kind of intense market economy where steam engines would have been useful. Why? Because they couldn't. The Romans did invent the state in the modern conception of the word, they did invent the concept of uniform laws, and codes, and bureaucracy, and public service, and public infrastructure and many other aspects of civilization which in the fullness of time made the modern world possible. But one thing they did not invent -- a functional and stable mechanism for the transfer of power.

Peaceful transfer of power was always a goal: most emperors were decent people, who really strived to preserve and develop the empire. Most of them actually succeeded in transferring power to their successor peacefully; for example the chain Nerva–Trajan–Hadrian–Antoninus Pius–Marcus Aurelius (the adoptive Nerva-Antonine dynasty, the "five good emperors") provided stabilty and good governance for an 84 years, almost a century, for 96 to 180 CE. But once in a while the transfer of power failed, or the successor was poorly chosen, and havoc ensued.

So the empire had periods of peace and development -- under the Julio-Claudians, under the Flavians, under the Nerva–Antonine dynasty -- and periods of anarchy and civil strife. And in the 3rd century one of those periods of anarchy -- the Crisis of the Third Century -- tilted the balance too much to the side of darkness. The empire was fatally wounded; it would never recover.

It could have been avoided, especially in fiction. In fiction, one can always make Commodus a good emperor; one could imagine a rule establishing that imperial succession must go to an adopted heir; or at least one could imagine the Romans instituting legallythe office of emperor and making it hereditary. (One of the big issues with the Roman empire was that before Diocletian it was not legally and empire and did not legally have an emperor; the state was officially a republic, and the men whom we call emperors ruled by cumulating key positions in the state; it's complicated, but basically they cumulated the position of speaker of the Senate, commander in chief of the army, high priest, and most importantly tribune of the people.)

Key points

  • You must make the economy of the Roman empire much more intense; steam engines are no use in a sleepy economy.

  • You must somehow overcome the Greco-Roman disdain for artisans; maybe have Alexandria develop into a hotbed of industry, with Rome forced to compete, I don't know.

  • You must avoid the crisis of the 3rd century at all costs, even if that means instituting the dominate early.

  • You must imagine a chain of unlikely but not impossible inventions...

  • And as a result you will get a world which resembles the classical Greco-Roman world a bit, but it's quite different, and you will have changed everything to follow. There will be no Middle Ages, no Islam, no British Empire...

  • $\begingroup$ // Greco-Roman civilization used relatively little fuel - It produced 80,000 tons of iron per year. It DID NOT use little fuel, no matter what you mean by relatively. // bronze is definitely not cheap - Nevertheless cilinders of first steam engines were made from bronze. // Romans did not have lathes able to machine metal - Hand-powered lathes was perfectly well used to machine metal. And you do not need lathe for steam engine making. $\endgroup$
    – Vashu
    Commented Nov 17, 2017 at 1:28
  • $\begingroup$ // No lathes means no (reproducible) screws - Just google when first reproducible screw was made and when first steam engine appeared. // did not have the technology to make high(-ish) pressure metal pipes - Typical Watt engine used 0.5 atm. That's not high pressure. And it was used on steamboats. $\endgroup$
    – Vashu
    Commented Nov 17, 2017 at 1:28
  • 2
    $\begingroup$ @Vashu: 80,000 tons of steel per year is one kilogram per inhabitant per year; for comparison, in 1750 (Newcomen's time) Britain alone produced 28,000 tons of iron per year, 5 kg per inhabitant, raising to a whopping 250,000 tons per year in 1800 (35 kg per inhabitant). With no lathes you cannot make bearings; since the engine must eventually turn a shaft, this is rather important. $\endgroup$
    – AlexP
    Commented Nov 17, 2017 at 7:43
  • $\begingroup$ Given that iron swords had replaced bronze swords all over Europe in the 4th Century BCE, and that their stereotypical armour was made of iron and mild steel, I find the assertion that Romans had never seen molten iron to be, shall we say, a tad suspect. $\endgroup$ Commented Nov 17, 2017 at 17:02
  • 3
    $\begingroup$ @KeithMorrison Iron working was not done by melting as it takes a considerably higher temperature than smelting ore into a spongy mass of semi-soft iron which can then be hammered into shape. Though China may have been able to produce cast iron during the time of the Roman Empire, there is no evidence for such blast furnaces in Europe until considerably later. $\endgroup$ Commented Nov 17, 2017 at 18:15

Steam engines require an industrial economy. Even Mark Twain's riverboats had a huge amount of iron in them from a Roman point of view. (Foot diameter pistons, multi foot stroke...) You need a way to produce and cast iron in multiton lots.

That means you have to have ways to heat it. That means coal and coke. England was able to bootstrap an industrial revolution with large forests that made good charcoal, and some deposits of coal that were near or at the surface. The steam engines drained the mines to get deeper coal, allowing more engines and railroads to be built. Rail allowed moving coal and iron ore from further apart locations.

Even so, steam engines were in widespread use on land before they were applied to ships. Ship use required higher pressures to allow both more efficient use of fuel, and to get the engine to a reasonable size.

One historian commented, "You can't railroad until it's time to railroad"

That said: The Romans were in Brittania for several hundred years. Put the whole industrial revolution at AD 200


Walking beam pressure pumps (which used cylinders and pistons) date to the 3rd Century B.C. (they used bronze, of course, but it would have worked perfectly fine for steam engines).

The crank (sun and planet gearing was used in early steam engines to get around Pickard’s patent on the crank; British patent law of the time allowed someone to patent things that were already in use) dates to at least the 1st Century (the Nemi Ships used them in their bilge pumps)

Flywheels predate civilization, and valve technology sophisticated enough for a water organ could have come up with valves for a steam engine.

Putting the pieces together in the appropriate combination was beyond them, for reasons others on this thread mentioned, but all the necessary components for making small- to medium-sized steam engines were available to the Romans.

  • $\begingroup$ Welcome to WorldBuilding John! If you have a moment please take the tour and visit the help center to learn more about the site. Have fun! $\endgroup$
    – Secespitus
    Commented Nov 20, 2017 at 13:43
  • $\begingroup$ All necessary components were there for making a one off as another answer mentions, either on this question or another one of the OP's. Full-scale production is going to need a little more revolution. $\endgroup$
    – Mazura
    Commented Nov 20, 2017 at 20:05

They didn't have the metallurgy, the machining skills, or the knowledge base to develop a practical, economical steam engine.

The first practical steam engine was the Newcomen engine, which depended upon an invention that the Romans did not have: the cylinder and piston, to translate the partial vacuum from steam condensing in an enclosed container, into power. The piston and cylinder was an idea that was first developed by Otto Guerickie as the vacuum pump, and he got the idea from an unknown inventor who had created a household fire extinguisher using a cylinder and piston as a water pump. Nor did the Romans have a planetary gear system to translate the push and pull into rotary motion, invented by William Murdoch.

True that the Greeks had developed a simple steam engine with a spinning ball, but that design did not scale up easily (no bearings that could allow a large ball to spin efficiently, no metals that could withstand the higher pressure and temperature), nor did they have a way to pump water into the engine while it was running under pressure (also requiring a cylinder and piston), so it wouldn't have had much endurance.

And finally, there is the knowledge base to consider. In Roman times, metalworking was confined to hand beating bronze, copper, and iron into shapes. You can make a good sword and shield, but you can't make a practical steam engine like that. Consequently, even the idea of using steam or anything other than human, animal, water wheel, or wind for power, would have been beyond even the brightest minds back then.

Theoretically, Romans could have created early firearms: they had the raw materials for both black powder and bronze cannon barrels... what they lacked was the knowledge base that such a device might be possible.

  • $\begingroup$ "The piston and cyliner was an idea that was first developed by Otto Guerickie..." the Romans had knowledge of pistons and cylinders in our time line. But if the Romans had been given a working example of a steam engine? $\endgroup$
    – Slarty
    Commented Nov 22, 2017 at 0:26
  • $\begingroup$ The Romans didn't have the metal machining techniques with which to build a practical, functioning steam engine. Their metalworking was limited to hand hammering swords and armor, not nearly precise enough. Going from hand hammering to machining metal is a huge leap, and didn't happen for at least another 1000 years. $\endgroup$
    – tj1000
    Commented Nov 22, 2017 at 16:34

Essentially, they would have to get rid of slavery - something they never did.

All the "economy" of the Empire was built upon such institution, and if anything, their navy was even more predicated on it than anything else (save perhaps mining). Their boats were moved around by oars, sails being a mere supplement (their sails only allowed navigation leeward).

This posed extreme limitations to their technological progress. Since slaves don't earn wages, there was no pressure to lowering wages by improving productivity, so there was very little need for technological creativity, and even less to put such creativity into "economic" enterprises. Also slaves typically cannot be trained to improve their abilities - there certainly were skilled slaves in all trades, but this depended on the capture and enslaving of already skilled artisans, not on catching unskilled workers and then teaching them a trade (plus there will always be some underuse of slave skills - think 12 Years a Slave, where a skilled musician is enslaved to perform utterly manual and unskilled labour). Such problem, if anything, grew increasingly as the original sources of skilled slaves - Greece and other already "civilised" regions - where integrated into the Empire, leaving it with only "barbarian" unskilled enslavable reserves of manpower.

Imagine trying to convince the owner of a Roman trading ship that he should replace his already paid for slaves with an equally or even more expensive machinery that then would need a much improved set of workers, that he would not easily, if at all, be able to procure at the slave market, resulting probably in needing to pay wages. Why would he pay twice for his power (machinery and then wages), if he had an already paid for crew?

In such an environment, not only technological innovation was stimied, but even more the integration of any eventual advances into the productive system. The intellectual powerhouse of Classic Antiquity - the Museum at Alexandria - certainly provided a few inventions, but those were at the service of military engineering, not of improvement of labour conditions or of labour saving in agriculture, artisanship, or commerce.

So, a social revolution - the abolition of slavery - was a precondition. Without that, even if they had the technological ability to make a steamship, they would build it, at best, as a "wonder of the world" - something impractical, made to astonish people at the power and might of their empire, not as something to put into daily practical use.


The Printing Press

You might have to have your Romans invent the printing press first.

There's a ton of ideas that had to come into place before the steam engine could be viable, and the best way to trigger an accumulation of ideas is to have a way to put those ideas down and to widely distribute them.

But then the question would be why the Romans would want a lot of books around. Maybe a religious movement that centered around books. (But that's been done.) Maybe books could be used for instructions for setting up colonies. Or for field manuals for the troops. Or the printing pressed could be used for creating massive numbers of fliers for propaganda.


What if Archimedes was captured by the Romans instead of killed, and he invented the printing press? That could be fun.


You must log in to answer this question.

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