Whither the pumps?
We have this grand tradition of heating here in the U.S. of A. that goes back to the days when heating was new and people didn't use pumps to move hot water because, well, they had not yet been invented. We used gravity instead, and gravity is wonderful because it is so dependable. Heavy things, such as cold water, sink, so when the hot water rose from the boiler, it would shove the cold water that was inside the radiators out of the way, and before long, we had a lovely circulation set up - hot water rising, cold water sinking, and continuous circulation to warm the people. Nice.
To make this happen, we needed to have pipes of a large size because friction is the enemy of flowing water, so those old systems had hefty pipes into which you could yodel, should the spirit so move you. The big pipes on the supply side matched the big pipes on the return side, and all of this made for a lot of work, a lot of money, and a lot of water.
The pumps finally arrived, and at about the same time as the Great Depression. People realized they could make the water move faster (thereby saving money on coal) if they used a pump, so onto the systems the pumps went, and the place they seemed to work best was on the return side of the system piping, down there where the water was coolest. Don't want to be burning up that new pump with red-hot boiler water. So down at the bottom they went, and down there they stayed. And these were big pumps, with packing glands that dribbled water into drains ,and manly motors that sucked up oil and electricity like beer at a ballgame.
The 1970s showed up, and with the new decade came small pumps that ran on high speed and needed no oil because the system water lubricated the bearings. We loved these little pumps because they got the job done well and they cost much less than their big, beefy, old uncles. They needed no drains and they were as quiet as can be, and when we changed them out, we placed them in the same spot as the old beasts - down at the bottom of the piping, even though water temperature was no longer a factor.
And that led to some interesting problems because of the greater pressure differential of the new pumps. Radiators that never had a problem before suddenly began to bind with air, and pipes that were once quiet were now gurgling. Air was causing all of this, but where was it coming from?
It took a while to figure out, but we finally got it through our thick skulls that it was the location of the pump that was giving us grief. We were all pumping toward the compression tank, and not away from it, and that was allowing the pump's now-greater differential pressure to play with the dissolved air in the water. Here's what it's all about.
Consider the closed-loop, hot-water heating system. We fill the pipes, boiler, and radiators with cold water, leaving no space at all, other than the space inside the compression tank, on the other side of the bladder. That bladder gives the water a place to expand when we heat it. We need the tank because we can't compress water; we can only compress air. Now when we fill any system, we use a definite amount of water, so let's take a case where we fill a certain system using precisely 40 quarts of water.
Okay, here's the question: When the pump first starts, does it shove any water into the compression-tank?
If you say yes, I'm going to have to ask you where it got the water that it's shoving. And if you tell me the pump got the water from inside the pipes, I'm going to have to ask you what took the place of the water that used to be in the pipe, but is now supposedly inside the tank. If you tell me that air has taken the place of the now-missing water inside the pipe, we're both going to have to wonder from whence the air came. Keep in mind that this is a closed system.
Truth is, it's impossible, within a closed system, for the pump to add water to the tank, and since it can't do that, it can't change the pressure inside the tank (because of Boyle’s Law). Nor can the pump remove any water from the tank because there's simply no place to put it. The pipes and the radiators are already completely full, and we can't compress water.
A pump in a closed system can neither add nor remove a drop of water from the compression tank, and since it can't do this, it can't affect the pressure inside the tank. Old Boyle's Law at work here. That's why we call the tank "the point of no pressure change." Whatever pressure you use to maintain the system in a filled and operating condition, that pressure will be a constant at the point where the tank is, and in the pipe that connects the take to the distribution system.
Now a centrifugal pump in a closed system is nothing more than a differential-pressure machine. It must produce a difference in pressure across itself, but it doesn't particularly care how it does this. It can raise the pressure on its discharge, or lower the pressure at its suction, or split the difference between the two; it really doesn't care. However, since the compression tank is the one point in the system where the pressure can never change, when we pump away from the tank, we will add the pump's differential pressure to the system's fill pressure. This is splendid because Henry's Law kicks in to help us get rid of the air that comes from the heated water (Henry’s Law states that gases dissolve in liquids in direct proportion to pressure). We use the pump's differential pressure to pulverize the air bubbles and move them out of the radiators and pipes and into the air vent on start-up. Whoosh! They're gone.
But pipe the pump on the return side, pumping directly toward the compression tank's point of no pressure change, and the opposite happens. The pump can't raise its discharge pressure because the tank is the point of no pressure change, so the pump creates its differential pressure by lowering the pressure at its suction side. Once again, Henry's Law kicks in, but not in a nice way this time. Henry lowers the pressure on the water, releasing all the dissolved gases, and turning tiny bubbles into balloons, and all in an instant. Suddenly the pipes are noisy and the radiators are bound with air.
And this is what we lived with for years in the U.S. of A. Nowadays, smart installers know to pipe pumps on the supply side of the boiler, pumping away from the compression tank. However, we do have our share of knuckleheads, and these proud people persist in piping the pumps on the return side so that they pump right at the compression tank, in spite of the air problems this causes. And do you know why they do this? It's because this is the way their grandfathers did it. And they will try anything new - as long as he tried it first.