How could this happen?” That’s what they asked.

The place was pretty big and old enough to have steam heat. The mains ran from here to there and used to return to a boiler that was nearly as large as my imagination. It ran on gravity in those days. They kept the boiler pressure low so the condensate’s static weight as it stacked between the steam traps and the boiler’s waterline was enough to enter the boiler.

But then that old boiler earned its retirement. Its replacement was young, efficient and held much less water. That was a problem because this one would go off on low-water before the condensate had a chance to return.

So they added a boiler-feed pump to the mix, which meant they also had to add steam traps to the ends of the mains.

As they looked a bit deeper into the system they noticed that it was taking on a lot of fresh water, which probably had a lot to do with the old boiler’s demise. Cold feed water contains oxygen and that old boiler was made of iron.

Rust happens.

One of the guys suspected that long, buried return line might be leaking, so they looked into it, and sure enough, it was. Condensate is steam turned to water, but as it turns, it absorbs carbon dioxide, which is natural to the making of steam, and that creates carbonic acid, which eats steel pipes.

Rather than dig up and replace that buried return line, another of the guys suggested they abandon it in favor of a condensate-transfer pump. This would collect the condensate from the end-of-main trap and pump it overhead back to the boiler-feed pump. It’s a common solution to a problem that’s liable to reappear, and it usually goes off without a hitch, but not this time.

The condensate-transfer pump ran OK the first time, but the second time it came on it started with an enormous wallop that shook the overhead copper tubing. The guys shut it off and got in touch with me.

“It sounds just like water hammer,” one of them said. “But how could that be? There’s no steam in that line. It’s just condensate and it’s a wide-open line. It runs from the transfer pump straight back to the boiler-feed pump and that’s vented to atmosphere.”

So this was one of those moments when we all had to get back to basics.

“You can’t get water hammer in that line without steam,” I said.

“But there’s no steam,” he said.

“Are you sure?”


“Even though it goes against all that you’re seeing?”

“That’s why we’re calling you. We need help.”

“That’s a better place to start. Let’s get back to the basics and begin with what we know for sure.”


“You have water hammer in that line, right?”


“And that means that steam is shoving the water around, right?”

“I guess.”

“So we need to figure out where the steam is coming from.”

“Can you stop by?”


We looked at everything. You have to when you’re troubleshooting. I looked at how they had piped that 3/4” copper tubing from the transfer pump. It rose about 20 ft., and from there, it went horizontal for 200 ft. or so, and then it dropped into the boiler-feed pump’s vented receiver.

As the transfer pump started, we heard the water hammer in the horizontal copper tubing and we saw the tubing shake. It happened every time the pump started. Clearly, there was steam in that horizontal run, but how could it have gotten in there?

We went back to basics. We also left the room where the transfer pump was and walked to the other end of that line. The water was trying to fall down the 20 ft. and into the feed-pump’s receiver, but the top of that line wasn’t vented, so as the water fell, it created a vacuum in the horizontal portion of the run. The condensate, being very hot, flashed to steam and just stayed there in the horizontal tubing. The next time the transfer pump started, its pressure caused the steam to quickly condense and the water that had been pushed back rushed in to fill the void. And that’s what we call water hammer.

I had them turn the end of the copper tubing up about 18 in., and then go horizontal again before it dropped to the feed pump. I also had them install a vacuum breaker at the top of the drop. Now the condensate could drop because air could get in when the vacuum breaker opened, and the horizontal line would stay filled because we trapped it with that 18-in. rise before the drop.


Keep an open mind

Problem solved.

There are always undeniable truths when you’re troubleshooting, but sometimes we miss them because we arrive at the job with our minds made up already.

Let’s say we’re on a problem job together. This one has a steam boiler that goes to 10-psi pressure when it’s supposed to never go over 2 psi because that’s the presssuretrol’s setting. The guy before us had changed the pressuretrol, the gauge and the pigtails for each. The new gauge still showed 10 psi. So what’s up?

Let’s begin by stating undeniable truths, and these are:

  1. The only way pressure can build is if we fill the system with steam or air and then add more steam or air. Can we agree on that? Good.

  2. The only way a boiler can make steam is by firing the burner. Yes?

  3. Only two controls can start or stop that burner and these are the thermostat (or the central controller for the building) and the pressuretrol. Hard to disagree with that, right?

  4. Gauges sometimes don’t work, even right out of the box. New products are not always perfect products.

  5. If the gauge is indeed good, and the pressure is actually going that high and it’s not an illusion, then the culprit has to be either the thermostat, the pressuretrol, the burner or all of the above.

  6. So that’s where we’ll look for the answer.

But while you’re doing that, keep in mind that some people who have a problem would rather have the problem than the solution. They want you to fail so they can tell others that even you couldn’t solve it. This has to do with ego, not engineering.

These folks have no idea what’s causing the problem, but when you give them a list of what might be the cause, they immediately say, “No, it can’t be that.” And that always makes me smile. They don’t have a clue, so how can they discount any possible cause without first checking it out?

Faced with folks like this, all you can really do is smile, wish them well, maybe shake their hands and then walk away.

You can’t help them.


This article was originally titled “Running down a mystery” in the September 2017 print edition of PM Engineer.