Hi Martin,
As Colin B says, efficiency of marine steam engines was measured in how much coal was burnt- but this was equally affected by the efficiency the boiler too.
To answer your question directly;
The efficiency of a steam engine was measured using a 'steam indicator card'- a few members on here will have used one in anger I'm sure. The device measured the pressures in the cylinder (steam works expansively, so) the pressure should not stay constant through the revolution, it should drop on a slow curve, drop off to nothing as the exhaust opens then rise rapidly, just a tad before TDC or TDC as this acts as a cushion and stops the piston hitting the cylinder covers before it starts its power stroke.
A chord was attached to the crosshead and also showed on the card the relevant position of the piston for any given pressure.
An efficient engine was the pride of any chief engineer, who was always the person who oversaw the setting of the valve lap and lead (the amount of steam admitted before power stroke, and the amount of steam admitted during the stroke before exhaust)
This job became harder the more cylinders you had, and yes, as you say Triples were very efficient, but Quadruple Expansion engines were very common in smaller marine plants too, owing to the fact that on big marine plants the last cylinder would have been too large to manufacture- though some large 5 cylinder quadruple expansion engines were manufactured which split the last expansion of the steam between two cylinders that shared the volume of the steam equally- though you sacrificed weight and length to do so.
The next, if not as important point was thermal efficiency of the pipework and cylinders- early steam engines had solid cast cylinders, as most [all] of the models on here do. However around the 1880’s experiments were made to cast the cylinder walls hollow and supply steam to this ‘jacket’, which became known as the ‘steam jacket’, and was then used in every steam engine design right upto the last days of steam. This ‘jacket’ counteracted the thermal inefficiencies inherent in an external combustion engine- ie. The constant pressure, and therefore temperature, differentials as the cylinder went through it’s expansive working. It also massively reduced condensation in the cylinders and valve chests- condensation is (not to be patronising) condensed steam- in other words a little like taking all the ‘Octane’ out of your preferred IC fuel- still the same stuff, but would be useless and would just serve to stop the engine working properly.
All marine steam plants also utilised a means of changing the valve movements while the engine is moving- the ‘valve gear’.
Predominantly ‘Stephenson’s link’, although ‘Hackworth’, ‘Marshall’ and ‘Bremme’ were all experimented with.
These allow the steam to work all the way from ‘full steam ahead’ to ‘full steam astern’.
The term ‘full steam’ meaning that the cylinder was working with minimal expansion- giving lot’s of power, lots of rev’s and quick responsiveness- just what the Helm needs when manouvering.
Everything in-between just improved efficiency as the steam in a hot engine could be taken to admitting just the slightest ‘whiff’ of the stuff, which sacrificed all the things the Helm needed when in close confines, but at sea or in the middle of a lake, but gave huge fuel savings. This was as long as the valve chest was provided with full pressure steam. This process was termed ‘linking up’ or ‘linking down’, a term from the use of Stephenson’s link when you could physically see the link taking a smaller arc as less steam was admitted- as you can on the models that use this gear too.
That’s efficiency in a nutshell, though there are other factors- dry steam, superheat, piston valves or slide valves, well packed glands, no leaks etc etc.
Greg