Why Do Steam Coils Freeze?

This e-mail showed up in my inbox one day.

"Our problem is that five out of hundreds of univentilator coils freeze up on a regular basis. They are part of two pipe steam systems with pneumatic control valves and condensate pumps. The F&T traps have been replaced and the coils are grading down to the returns. The traps on some are two feet or so below the coil outlet. The boiler pressure on all buildings runs between 4 psi and 8 psi and the boiler shuts down at night until the coldest room in the building requires heat. If the outside temperature falls below -15 C the boiler stays on pressure. Do you have any ideas on how to solve this problem?"

I figured he was a Canadian (the Metric system gave him away), and if you're going to freeze steam coils, what better place to do it than Canada? I wrote back and told him that he should look and see if there's a vacuum breaker between the control valve and univentilator. Vacuum breakers do exactly what their name implies, and vacuum is what you get when steam condenses, unless air can get back in to break it. And unless you're dealing with a mechanical vacuum system, vacuum is not your friend.

What's going on in that coil is exactly what goes on when you hold your finger over the top of a straw that's filled with water. Think of your finger as the vacuum breaker. Lift your finger and gravity does the rest.

But when it comes to steam coils, you can't depend entirely on gravity to get rid of the condensate. The guy in Canada said that his traps are two feet below his coils, right? Okay, think like condensate. You're building up inside the coil. The control valve is closed because the space is warm enough. The vacuum breaker just opened. You now have just two feet of vertical pipe available for you to stack and build static pressure. How strong are you feeling right about now? You're producing less than 1-psi pressure down there at the bottom of your stack, aren't you? There's a float & thermostatic steam trap down there and you're pressing against it with a measly 1-psi pressure.

Feeling wimpy?

Me too.

And all of this assumes that the vacuum breaker is in place and working, of course. If it's not, there will be a vacuum tugging you back into the coil and there will be little or no pressure on the inlet side of that F&T trap, which is why vacuum breakers are so important.

Now let's take this a step further. There's a valve pin and a valve seat inside that F&T trap and the size of the hole in that seat has a lot to do with how much condensate is going to flow through the trap. A low-pressure trap has a relatively big hole in its seat because there's not much pressure available on the inlet side of that trap (15-psi maximum) to shove the condensate through. High-pressure traps have much smaller holes in their seats because there's more pressure available to do the shoving.

The challenge here (and I'll bet this is a big part of our Canadian friend's problem) is that a lot of people select these F&T traps for a differential pressure that's going to occur when the control valve is fully opened. They're figuring that steam at the maximum available pressure (4-8 psi in this case) is actually there. And a lot of people will assume that the pressure at the outlet side of the trap is atmospheric because the trap probably drains toward a vented condensate receiver.

So let's say you pick a trap to move a certain load of condensate (which would be the unit ventilator's rating) at a pressure differential of 4 psi (that being the seemingly worse case when the control valve is wide open). As long as the control valve is open, the condensate will drain by gravity (assuming there's no pressure in the return line). But what happens when the control valve closes? Now all you have going for you is static pressure, and that's less than a 1 psi. Assuming there's no pressure in the return line beyond the trap, the condensate may decide to hang around a while inside the coil because there's just not enough differential pressure to make the trap work. I mean I'd hang around if I were the condensate. Wouldn't you?

And let's make our day a little bit more miserable. What if the pressure in the return line isn't atmospheric? All it takes is one nearby trap to fail to put pressure in that return. Or one high-pressure trap that's dumping flash steam into that return line (a common mistake). Or a vertical lift to an overhead return (which creates static backpressure). Or someone plugging the vent line on the condensate receiver because he's tired of seeing the plumes of steam spewing from it (a nasty habit that some building superintendents acquire). Any of these things will change the pressure differential across the F&T trap, and cause the condensate to back into the coil where it will meet Jack Frost.

Sadly, the truth is that an F&T trap is basically a dumb machine. It will not do the thinking for us and it doesn't care if we are happy or miserable from day to day. It just sits there like a meatloaf and does nothing until we bring the proper combination of load and differential pressure to it. And you have to figure those two things at their worst-case conditions. Play the "What if?" game.

But let's get back to our Canadian's misery. To make things more interesting, the frozen-coil problem probably doesn't show up every day. In fact, he'll most likely be joyous during the dead of winter when the control valves are open most of the time. It's the early spring and the late fall that will drive him crazy because these are the seasons when the control valves will throttle. Throttling reduces the steam pressure at the inlet to a coil and changes the differential across the F&T trap. So the condensate backs up and freezes inside the coil on the days when it's least expected.

Can you see it in your mind's eye? When control valves are involved, you have to size the F&T trap for the absolute minimum differential if you want it to work. Either that, or place it very far beneath the coil, which may not be possible. And don't forget the vacuum breaker.

If you're not sure how to size steam traps, call a manufacturer's rep. Those guys will fall over themselves to help you. I used to be one of those guys, so I know.

A final thought for you: If it's a heating systems and the F&T trap is the same size as the pipe that it's on then it is the wrong size for the application. Guaranteed.

I once stood with an engineer at the base of a two-inch steam riser that reached up into an apartment building in Westchester County, NY. That riser was feeding radiators and the condensate was draining down a separate return. The F&T trap at the base of that riser had to drain the condensate that formed when that riser went from ambient temperature to steam temperature. That's all. I'll bet that on the average day it didn't see more than a bucket of water. It was a two-inch trap on a two-inch pipe. A three-quarter-inch trap would have been too large for this application. I was there because the trap had failed. It failed because it was too big.

"Why didn't you use a three-quarter-inch trap?" I asked the engineer.

"It's a two-inch line," he said. "Three-quarter would have looked stupid."

I suppose he was right about that, but even though the trap looked nice, he looked stupid.

If you liked this article, you'll love Dan Holohan's The Lost Art of Steam Heating.