Gravity Hot Water Heating Insights
“You sure? I asked.”
“Yep. I looked everywhere.”
“Did you go in the attic?”
“That’s the first place I went,” he said. “I know that’s where they installed those open tanks. They vented them out onto the roof.”
“Maybe it was a closed gravity system,” I said. “The Dead Men built those too, you know. They installed closed steel compression tanks either in the attic or in the basement. Those jobs didn’t have circulators, but they all had tanks. I’ve never seen one without a tank.”
“Would you like to see one?” he asked.
So we got together, and I did what he had done, that being traipse up and down the stairs a few dozen times. No tank. The job was heating just fine, though; it’s just that there wasn’t any compression tank. Anywhere.
Now, you’d expect the relief valve to be popping, wouldn’t you? Right. But it didn’t. The water heated and cooled between cycles, but the pressure gauge hardly moved. I tapped it a few times, figuring it was busted (but who can tell by tapping?).
After a while I noticed that every cast-iron radiator on this job was piped with a bottom-to-bottom connection. That got me thinking because those radiators were huge, and I figured, given their size, they might be a great place to store air.
Each radiator had a manual vent so I opened one and a lot of air came out. This was pretty weird because the radiator was heating well. Where the heck was all this air coming from?
My buddy and I bled the rest of the radiators throughout the house, and we got air from all of them. A lot of air. We wondered why this particular Dead Man had never finished the job of bleeding his radiators. The radiators were heating just fine, though, and that added to the mystery.
Once we got the air out we walked down to the boiler room because we couldn’t think of anything better to do. To our delight, the relief valve was popping and splashing hot water across the floor. We looked at each other and smiled. We had found the “compression tank.” Actually, we had found a lot of compression tanks. The Dead Man had used the top portion of each cast-iron radiator as a little compression tank. He saved himself some money. Pretty clever, eh?
So I started digging in my old book collection (I should have done that sooner), and I learned why the Dead Men liked to connect their freestanding, cast-iron radiators across the bottom, even when they planned to use an actual expansion or compression tank. When you pipe bottom-to-bottom, a cast-iron radiator will never air-bind. Had they used a top connection for either the supply or return line, trapped air could stop all circulation to the radiator and create problems for them. With a bottom-to-bottom connection, the radiator always heated. It might not heat all the way to the top, but it would heat.
The Dead Men didn’t have circulators; they ran all of these old systems on gravity. The hot water would rise out of the boiler because it was lighter than the cold water up in the radiators. The cold water would fall from the radiators and shove more hot water up. Before long, they had this neat Ferris wheel of convective currents going, and the whole house was beautifully warm. Hot water will rise inside the freestanding radiator the same way it does inside the system piping. The colder water falls from the top of the radiator and leaves through that bottom connection. It’s beautifully simple.
But when you add a circulator to one of these old gravity systems, wacky things can happen. If you use the wrong circulator - one that’s too powerful - the water will zip right through the radiator’s lower nipples, significantly slowing the natural gravity circulation within the radiator. Get it? The cold water that’s trying to fall out of the top of the radiator will be stopped by the hot water that’s zipping through the bottom outlet connection. You wind up with a radiator that heats only across the lower nipples; the rest of the radiator will eventually heat but it takes a long time.
You can solve this problem by using the supply valve at each radiator to throttle the flow through the radiator (assuming those supply valves are still working). By slowing the flow as it enters the radiator, you’ll promote gravity circulation within the radiator and increase its output. That may not make sense at first, but it works.
And something else might happen when you add a circulator to one of these old gravity jobs. As soon as it starts, the circulator will look for the compression tank (hydraulically speaking) because the circulator needs to know what to do with its differential pressure. The tank tells lets it know by acting like a hydronic bookmark, a point of no pressure change. The circulator responds to this point by either raising its discharge pressure (if you pump away from the tank) or lowering its suction pressure (if you pump toward the tank).
When you lower the system pressure, air comes out of solution, makes noises in the pipes, and often gets trapped in the radiators. When you have a lot of air trapped in radiators, you wind up with a lot of points of no pressure change. This confuses the circulator because it doesn’t know what to do with its pressure differential. It sometimes lowers the pressure at places where it shouldn’t, and that can leave you with a lot of annoying gurgling.
Another friend ran into this problem once. The Dead Man had connected his freestanding, cast-iron radiators in a down-feed system, with the inlet at the top, and the outlet at the bottom of each radiator. Both connections were on the same side of the radiator. Now, my friend’s a very bright engineer, and he figured out what was happening when he installed pressure gauges on either side of the circulator. He was pumping away from the compression tank, but he was still getting a wicked drop in pressure at the circulator’s suction flange. This told him the circulator was seeing more than one “compression tank” out there in the system. He knew the air in the radiators was the problem, but since the Dead Man had gravity circulation in mind, and not pumped flow, my friend was having a very hard time getting the air out of those radiators.
He finally solved the problem by installing a good air separator at the boiler, and then raising the system pressure. When you raise the pressure, the water absorbs more air and brings it around to the air separator. It took him a few days, but he finally got rid of the air. He could see the result of his efforts on the gauges, which now showed a positive pressure at the circulator’s discharge.
It’s nice that we get to work with both the new stuff and the old stuff, isn’t it? America is a heating museum. Enjoy it!