The small pump is actually a boiler feed pump so that just tops up the water level in the boiler as the steam is drawn off.

The larger pump is an air pump which blows or draws air through the condenser this acts much like the fan on your car engine that blows water over the core of the radiator. This makes the tubes that run through the condenser cooler and encourages the steam to condense which creates a vacuum to draw the last bit of usable steam out of the engine.

In the marine applications I'm familiar with, the 'large' pump, the vacuum pump or wet air pump, pulls a vacuum in the condenser shell which 1) lowers the pressure which causes greater pressure across the LP cylinder, 2) by lowering the pressure causes the steam to cool and condense faster, and 3) pumps the condensate out to the hot well. The hot well is where the oil and crud is separated from the condensate which is then fed to the boiler by the 'small' pump. Oil in a boiler will ruin it very quickly as it burns on to the surfaces and seriously interferes with steaming.

I have an excellent document (30 pages of .pdf) on the subject. If you are interested I will email it to you. I don't know how to 'link' it here…..

Pete

Bazyle and I were typing at the same time… The document I mention is specifically about the Edwards air pump which was designed for the wet pump function and is the best for the purpose.

Edited By Peter Krogh on 30/11/2016 13:42:00

So looks like I was a bit wrong.

Is a third pump now needed to move the water through the tubes of the condenser? I see some models have 3 pumps – air, feed and circulating.

Yes Bazyle, the air just bubbles out of the waste water pipe, after it had been pumped from the condensor by the"air pump". Air pumps on horizontal engines have a vent on the top of the pump, and just vent into the engine room. Marine engines have a vertical air vent pipe adjacent to the air pump. Barometric condensers require no pump to remove the water, as the exhaust is arranged as a very large barometer. Once the column of water reaches about 28 feet, the atmospheric pressure cannot suport any more, and a vacuum forms above the water. Any additional water entering will just run out of the bottom. all the air pump has to do is just that; remove any air that may be present. I have a picture of such a condenser serving several engines. It just looks like an oversize soil pipe network on the outside wall of the building. The Newcomen engine does not figure here, as it condenses inside the cylinder.

Maurice

With the Edwardes pump, it is important that the condenser be above the pump so that the water can flow downhill all the way to the lowest point. The pump does actually pump both air and water, and this design is quite clever in that it uses the water to help give good seals on the piston and on the delivery valve. There is no foot valve, which is also quite a good feature, since these can be troublesome and are hard to get at. As noted above, the conical end of the pump plunger dispaces water up through the ports around the cylinder, so as the ports close on the upstroke there should be water lying on the piston, helping to give a good seal there. Then on the delivery side, the pipe to the hotwell should be uphill, and should deliver into the hot well over the top of the water level, so water will tend to lie on the delivery valve and help it to seal.

Full size plant usually does/did have a circulating pump. It is preferable if this can be driven separately from the main plant, but not vital. The worst effect is that it can take a little longer to establish vacuum if the condenser has got a bit hot.

I've seen it claimed before that the pump is not creating the vacuum. I'd like to point out that on starting, the condenser and exhaust spaces are liable to be full of air, and unless that is removed, a decent vacuum will not be obtained. Then once things are running, the pump has to remove the condensate, otherwise the whole system will become waterlogged. So while the reduction in volume by condensation obviously helps, there will never be a vacuum unless there is a pump there, or some equivalent like the above mentioned standpipes or a vacuum ejector.

Some beam engines used a jet condenser. This uses the vacuum to suck in a jet of water, sufficient to condense the steam, and a pump to remove the air and condensate. From experience watching a beam engine like this being started at MOTAT in Auckland, the vacuum tends to take a while to establish and then come in with a bit of a rush. While there is no vacuum, there is nothing to suck in the water, and the pump is trying to deal with a wet fog of steam and air. Then once a bit of water starts to get in, the steam condenses, and the pump suddenly has a much easier job.

John