Yes, currently boilers are 50% efficient. So 50% of the fuel energy added makes it into the steam.
This would be comparable to a real world boiler that has some scale buildup from bad water being fired without the attention to detail necessary for good combustion efficiency. Such a boiler would likely be belching smoke, and would be a safety concern because a dirty boiler may experience local overheating that can cause a rupture and explosion.
Quoting the other part of my post in the other thread that didnt get copied here yet, regarding engine efficiency:
Engines on the other hand, Factorio's engines are laughably impossible.
Real life steam engines are really bad for efficiency.
Single expansion (Locomotive, small stationary engines): 5-10%. Most of the energy in the steam is lost to under-expansion, the steam is released while still at a significant pressure. This is why the blastpipe is so effective at inducing boiler draft, and why steam locomotives chug the way they do.
Compound expansion (Stationary and marine engines): 10-30%. Too complex and large to be reliable in a locomotive, this type was commonly found in factories and ships where the improved efficiency more than made up for the increased complexity and the added weight was of little concern. Even when run with a condenser so that the steam is in fact expanded all the way to a vaccum, there is still a significant energy loss in the latent energy needed to change the water to steam that cannot be recovered by any known engine.
Turbines: Up to 40% efficient. Like the compound engine, a turbine is able to expand the steam all the way to vaccum if designed to do so. But unlike piston engines, turbines are only efficient in a narrow envelope of inlet conditions and RPM. Stray too far from this envelope and your efficiency goes up the cooling tower that condenses the exhaust to be fed back to the boiler.
I would love to see a mod for factorio that actually implements realistic efficiency for steam and nuclear power. But given how disruptive to the game balance that such a configuration would be, it would probably have to be a mod and cannot be part of the vanilla game. Too much realism takes all the fun out of it for people who just want to play the game.
There were engines that could push the envelope- for instance a single expansion Corliss valve was significantly more efficient than the standard slide valve and piston valve arrangements, and compound corliss valve engines were well known for their fuel economy.
Factorio's engine layout suggests a triple compound design similar to what might have been found in a ship or large factory. Such an engine should be able to exceed 20% efficiency at part load, but without a condenser it will have underexpansion losses at full load. The devs talked about making a closed-loop steam system and including condensers and such, but ultimately decided against doing so for a variety of reasons. Shame too, a closed loop steam solution would make it possible to put the power station close to the fuel supply and bring make-up water to it by railcar instead of forcing power stations to be built on lakeshores.
So armed with some information of typical real world values, lets see how those numbers would look in Factorio.
1 coal = 8MJ.
Boiler efficiency, 50%
8MJ coal + water => 4MJ saturated steam
Engine efficiency 25%, make the math easy.
4MJ saturated steam => 1MJ electricity.
But in addition, changing the engines to 25% efficiency also means that each engine consumes 900KW in steam while only producing 225KW of electrical power.
So the 1:20:40 power station we were introduced to in 0.15, which currently produces 36MW theoretical power and I've tested to 35.3-35.8MW sustained, would now only produce 9MW of power.
End result is that players would need 4 times the power station capacity to produce the same power, and would consume 8 times the coal to produce that power.
If we were going to actually implement such a drastic change, I think it would be necessary to re-think the numbers across the entire system. For instance the following:
All fuel values x4. Coal would be 32MJ like so instead of 8MJ. Like so the consumption rate would only increase to 2x the present value.
Steam engines consume 4x the steam they currently do, but produce the same amount of electrical power. This would give an efficiency of 25%, but due to pipe flow limits power stations would be reduced to 9MW instead of 36MW.
Boiler steaming capacity stays the same.
The result would be a ratio of 1:20:10, which kind of echoes the pre-0.15 ratio if you think about it, and would have a 9MW theoretical output.
Similarly though, you could make the steam engines consume 8x the steam and produce 2x the power, and raise the boiler steaming capacity by 2 to make up the difference. The ratio would then become 1:10:5, still producing 9MW but taking up half the space. Savvy players would build these as 18MW pairs, in much the same way pre-0.15 power stations could be constructed by making pairs of 1:14:10 sets.
Either way, traditional steam power in stationary and marine applications had 2 or more boilers per engine. Large ships like Titanic had upwards of 30 boilers to power its 2 engines and turbine.
Factorio has been unrealistic for some time in the overall scale of boiler needs to engine size, in no small part due to the engines having 100% efficiency instead of 25%.
In my mind, Steam is the eternal king of the railway.