Heating Boston's Old North Church

In this episode, Dan Holohan shares a story about replacing the boiler on the unique heating system in Boston's Old North Church, as well as some Dead Men tips for a false-water-line bottle.

Episode Transcript

Huge, cast-iron, column-type, steam radiators warm Boston's Old North Church, the place made famous by Paul Revere’s lanterns. They are of the two-pipe-steam variety, but instead of having a supply valve on the inlet and a thermostatic radiator trap at the outlet, they have two angle valves – one at each end. The supply valves are mostly 1-1/2", the returns, 1-1/4". If you're at all familiar with two-pipe steam you'll immediately realize those return lines are much larger than normal. And to make things even more mysterious, each radiator also has a one-pipe-steam air vent!

What we have here is a "two-pipe, air-vent" system. This peculiar blending of one-pipe steam with two-pipe steam developed out of necessity more than a hundred years ago. You see, one-pipe steam doesn't lend itself well to churches of this size. The relatively high heat loss created by vaulted ceilings and single-pane, stained glass windows calls for enormous radiators. To satisfy the demand of such huge radiators without creating water hammer or spitting air vents, a fitter would have to install very large supply mains and risers, making sure each was perfectly pitched and properly drained. This simply wasn't practical in most cases, so during the 1890s, heating engineers came up with the idea of installing a second pipe on each one-pipe steam radiator. The second pipe's job would be to drain condensate back to the boiler so that it wouldn't have to flow against the in-rushing steam. No steam would travel through this return line, only water.

The turn-of-the-century engineer would oversize each radiator’s drain line to make sure it quickly removed the large volume of condensate that formed on start-up and returned it all to the boiler. Since very little condensate flowed against the steam, the engineer solved most of his water hammer and spitting-air vent problems before they occurred.

Each radiator drains into a wet return. This piping arrangement seals one radiator from the next, ensuring steam will not travel from the supply of one radiator into the return of another.

In practice, each radiator works as a one-pipe radiator with steam pushing air out of the system through the radiator's air vent. The key difference is that the condensate leaves through its own pipe instead of flowing back to the boiler through the supply risers.

It's a beautifully simple system with few moving parts. But as you can imagine, installing twice as many pipes as you'd need to heat the building with a one-pipe system added considerably to the installation costs. This is why you don't often run into this type of system. Few could afford it.

With the invention of the thermostatic radiator steam trap in 1903, two-pipe steam with small, "dry" (above the boiler water line) returns became possible. Radiator traps immediately made the two-pipe, air-vent system obsolete. The few that remain are there solely to keep us on our toes.

A contractor was hired to replace the steam boiler in the Old North Church a few years ago. He sized the replacement boiler by measuring the connected radiation load, and determined that the replacement boiler would need a boiler-feed pump because of the new boiler’s relatively low water content.

The challenge of using a boiler-feed pump, however, was that he had to leave the feed-pump's receiver open to the atmosphere. On this job, he knew that when his crew installed the boiler-feed pump and opened the return side of the system to the atmosphere, the system's wet returns would drain. With no water leg to stop it, the steam in the supply risers would suddenly have access to the radiator return risers through the formerly wet, horizontal return mains in the basement. This would have caused a tremendous water hammer as the steam criss crossed into places it had no place being.

The contractor realized that even if he placed one large float & thermostatic steam trap at the inlet to the boiler-feed pump he still wouldn't be able to keep the steam out of the return risers. At best, that single trap would only protect the boiler-feed pump from damage. But as pressure built, he knew the steam would push back to the single F & T trap and create water hammer, spitting air vents and uneven distribution problems. He reasoned that for this job to work properly as a two-pipe system, he would have to install thermostatic radiator traps at the outlet of each radiator, and float-and thermostatic traps at the ends of all the steam mains. He would also have to deal with the asbestos that covered all the pipes.

The trouble, however, was that the steam traps, with their related system repiping and necessary asbestos removal, would add tremendously to the price of the job. But of even greater concern was the disruption a project of this size work would cause to the operation of this National Historic Landmark during the height of the tourist season.

In reading my book The Lost Art of Steam Heating, the contractor noticed a neat trick called a “false-water-line bottle” that I had learned from an old-timer. He immediately recognized it as the way to keep the system intact and still be able to use the boiler-feed pump with its open receiver.

He decided to build the “bottle” from a short length of 12-inch diameter Schedule 40 pipe. He then joined all the wet returns in the boiler room and piped them into the bottom of the “bottle.” Next, he equalized the top of the bottle with a steam line from the boiler header. This sealed all the system piping – both supply and return lines – from the boiler-feed pump's open receiver.

The next step was to determine where the false water line should be in relation to the system's supply and return piping. His goal was to keep what had always been a wet return wet, and what had always been dry return dry. Once he'd figured out what went where, he knew how high his false water line would have to be. The top of the false water line would mimic the water line of the old boiler. It would be the height to which he'd allow condensate to rise before it flowed through a series of inverted bucket traps that he would weld onto the “bottle” and into his new boiler-feed pump. By doing this, he kept all the wet returns free of steam and immune to water hammer.

He used a series of three, 3/4" bucket traps to move the condensate but not the steam from the “bottle” into the boiler-feed pump. He welded the traps into the “bottle” at slightly different levels. The lowest bucket trap established the height of the false water line. Once the returning condensate rose to that height, it would open the trap and spill into the boiler-feed pump’s receiver. If the level of returning condensate was lower than the lowest bucket trap, the trap would stay closed and neither steam nor condensate would leave the “bottle.”

The reason he used three, 3/4" bucket traps instead of one large trap was because the start-up load will be greater on mild days than it would be on colder days. If one 3/4" trap couldn’t handle that heavy start-up load, the other two traps (piped at progressively higher levels in the “bottle”) would jump in to help.

On a cold day, the steaming cycles will lengthen, lessening the number of times the traps will have to deal with start-up condensate. Had he used one large trap instead of the three smaller ones, the large trap would probably have been damaged during light-load periods by wire-drawing.

The purpose of the 12"-wide “bottle” is to lessen the surging effect of the returning condensate and to protect the steam traps from water hammer. In The Lost Art of Steam Heating, I called for a bottle made of 10" pipe, but I gave this as a minimum size. The contractor had access to the larger-size pipe so he decided to use it instead.

Bucket traps work better than float & thermostatic traps in this application because there will always be some water hammer inside the “bottle.” Bucket traps are more resistant to water hammer than F & T traps, however, a bucket trap generally doesn't handle the removal of air as well as an F & T. On this job, air removal wasn't a concern (each radiator had its own vent), so the decision was easy.

As further protection against water hammer, I recommend in my book that he pipe two short nipples and an elbow inside the “bottle” to force the condensate into a 90-degree turn before it enters the trap. This piping trick lessens the effects of water hammer and greatly extends trap life.

Once he had everything piped, he started the system and the returning condensate spilled from the traps into the vented boiler-feed pump's receiver. From there, the condensate returned to the boiler whenever the pump controller called for water.

The system piping remained unchanged. He touched nothing except the near-boiler piping. All in all, it was a beautifully simple and cost-effective solution to a challenging problem. The tourists never even knew the work was going on. The church saved about $70,000 and the new boiler ran efficiently without creating any water hammer whatsoever.

Put the “bottle” in your bag of tricks and save it for the next time you have to replace a steam boiler and add a boiler-feed pump to some old gravity return system.

I hope you enjoyed that Dead Men Tale. And if you did, please share it. And if you haven’t already, please subscribe to this podcast. I have many more Dead Men Tales to tell you, and I appreciate your taking the time to listen. Thanks.