By Mark Eatherton

Last month (CB Oct. 2007, pg. 102), I began telling you about an orphaned snowmelt system that I took over and how the boiler seemed to constantly and mysteriously lockout. The last thing I had tried was to completely change the venting arrangement.

About two weeks later at 4 a.m., the homeowner called me at home to let me know that the pump had started and the boiler was not lit. I quickly drove to her home, and as I entered the stairwell leading down to the boiler room, I again could smell the aldehydes associated with partially burnt natural gas.

As I walked down the long hallway to the remote mechanical room, I passed a plug-in type carbon monoxide monitor. I noticed the unit reading 150 ppm of CO. I unplugged the unit, and took it with me to the mechanical room.

As I entered the room, I observed that it was just about as cold in the unheated mechanical room as it was outside, which is a normal consideration for a remote mechanical room with code-required combustion air and no internal means of heat. I also noted that my eyes and lungs were stinging. I plugged in the CO detector and, after it booted up, it shot up to 250 ppm. I quickly left the room, closed the door, and went upstairs to talk with the concerned homeowner.

She was complaining of a headache that was borderline migraine and said she felt like she was getting the flu. She asked me if I wanted a cup of coffee, to which I obliged. She took a cup out of the cupboard, set it in front of me and poured it full of milk. I recognized this as the beginnings of carbon monoxide poisoning. Confusion, headache, aches and pains, and general disorientation. I told her we needed to air the house out, and we began opening windows. I found another plug-in CO detector on that floor, and after about a half hour it had settled back to 0 ppm.

I went back to the remote mechanical room in the basement and it, too, had cleared itself of the CO that was present before.

Then I did something that I would not recommend anyone do. I reset the boiler, climbed into the corner and held my head directly in the stream of the incoming fresh air and observed a killer at work.

As the boiler fired, there was a definite major rollout occurring at the combustion chamber/burner opening. It was so bad that many of the burners ignited the gas inside their respective burners.

A white vapor began spilling from every opening of the boiler into the room. I kept my eye on the CO detector. It started climbing quickly. I also kept an eye on the barometric damper. It was doing exactly as it was designed to do, staying completely closed, attempting to establish a negative draft within the 30-ft. tall stack filled with cold air. There was a considerable amount of spillage and white vapor coming from the lower combustion chamber/burner access opening.

The CO detector continued to climb until the digital readout read 999 — the maximum readout capability of the detector. Bear in mind, I had my head stuck right in the incoming fresh air stream to avoid being directly exposed to the deadly concentration of CO.

Suddenly, the main burner shut down and the lockout light on the ignition controller came on, indicating a lockout condition. I had finally witnessed what the boiler was doing in real time, and it was not what I had expected.

This potential killer was spilling products of combustion from the lowest part of the combustion process, an area that normally sees a significant flow of air into it. It filled the room with carbon dioxide, which helped to enhance the production of carbon monoxide. As we all know, you can’t sustain a flame in the presence of carbon dioxide. When the pilot flame signal was lost, the ignition controller immediately went into a lockout status, as it was designed to do.

The only problem was, the room was completely filled with the killer gas. These gases were leaking through the ceiling penetrations above, and into the spare bedroom that the owner of the home had been sleeping in. I took a deep breath of fresh air and ran upstairs and asked the homeowner to step outside and into my warm car to get away from the toxic fumes.

After having sat outside for a half an hour, I ventured back into the home and found the CO levels to be back down below 50 ppm. I turned on some kitchen exhaust fans to help clear the air and allowed the homeowner back into the house.

As we sat there, contemplating what had just happened, I told her that my only guaranteed solution was to replace this boiler with a sealed-combustion modulating condensing boiler. She didn’t ask me how much it would cost to do it, she just said, “Replace it.”

Of course, I thanked her for the order and explained that it was our policy to not perform any work prior to the owner knowing the cost of the recommended changes and that I would assemble a quote for her to consider that day.

After I left the jobsite, I kept going over the events that I had witnessed. I kept asking myself over and over why this boiler was not required to have some means of roll-out detection on the lower combustion chamber opening and the draft relief hood opening, just like what’s required on smaller residential boilers. I asked some boiler manufacturer representatives and got no response. I guess that’s one of those questions that nobody in the boiler manufacturing business wants to answer.

The other question that really kept bugging me is why the CO detector, a popular off-the-shelf brand, was sitting at 999 ppm and had not gone into an audible alarm. I understand that the UL approval process has a time-weighted value to it but, in my mind, if the CO rises above a certain deadly point, all timing elements should be thrown out and the device should go into an alarm state immediately.

In the end, we replaced her atmospheric boiler with a sealed-combustion modulating condensing boiler, which virtually eliminated the possibility of any carbon monoxide being released into the house. In retrospect, my little adventure into the boiler room to observe the killer at work could have backfired on me. As I stated, I would not recommend that anyone subject himself to this kind of dangerous situation, but in my mind’s eye, it was the only way I was going to be able to figure out exactly what was going on to cause this continual lockout situation.

In closing this article, I want to ask the industry two questions:

Why is it that these commercial grade boilers are not required to have any rollout safety switches anywhere on them like their residential cousins are required to have?

Secondly, why didn’t the CO detector go into an audible alarm status when the CO rose to a deadly level in such a short period of time?

If no one can give me direct answers to these questions, maybe somebody can at least give me some direction about how to get these requirements changed. I stand ready to help in whatever capacity I can offer.

Mark Eatherton is a Denver-based hydronics contractor. To reach Mark, e-mail markeatherton@ comcast.net or call 303/778-7772.