Implausible multiple failures in my gas central heating system

[Jonathan Scott]

How many things can go wrong simultaneously and apparently independently with a gas central heating system? Mine has been stretching credulity to the limits last weekend.

Saturday afternoon, we noticed house getting cold and spotted that pilot light was out. Couldn't get it to relight at all. Phoned our usual plumber / heating engineer but got voice mail saying closed due to illness.

Looked up a "local" emergency plumber (expensive) online, who arrived a couple of hours later. He couldn't get it to light either, but didn't seem to have any real idea why. I mentioned that on a previous occasion we had pilot light go out due to some dirt in the nozzle, so he borrowed a needle to try cleaning it out, and after that the pilot light flame would light. However, it would go out as soon as the igniter button was released. He then said it must be the thermocouple (which senses the flame and holds the gas valve open). I asked if he could do a test to confirm that it was the thermocouple (I assumed that he could check it with a meter and a flame), but he didn't seem to know how to do that, but instead he dismantled a lot of stuff and made a lot of mess (and threw away some bits of the "glass rope" which is supposed to seal the burner area). He also said it could be the thermostat and gas control, but as that doesn't get involved until the whole thing is running I couldn't see how, and he couldn't give a coherent reason. But his final conclusion was that the thermocouple needed replacing; it's a cheap part and easily available.

We then remembered we had the phone number for someone who had previously been a very efficient heating engineer assistant to our normal local plumber, and he said he could get the thermocouple and come and install it Monday morning. So for the rest of Saturday and Sunday we used a gas fire in our living room for heat (using fans to spread it around the house a bit) and used our electric immersion heater for hot water.

So on Monday morning, he came and replaced the thermocouple and it still didn't work, and it only took him moments to prove it must actually be the gas valve operated by the thermocouple that wasn't working.

So he got that part during the day and came back to fit it in the evening. So then the pilot light stayed on and the burner would light. But no water circulated - the boiler got very hot but the outflow pipe remained cold! He then found that the water pump was not working properly (again easily demonstrated), but he happened to have one in the van that he had left over from upgrading a working system in the last few days. But that one didn't work either, and eventually turned out to be in a worse state than our original one (although the system from which it was removed was very small, so it may well have worked there).

So he went out and got a new pump (expensive again) and installed it. And it still seemed to be not working! While he was investigating, he discovered that the backplate wiring for our central heating timer was incorrect, so the combination of functions he was testing (central heating without hot water) did not work.

So he decided to test with hot water instead (adjusting the thermostat to make sure it would come on). But it STILL didn't work. He and I were both getting quite stressed out by that point. However, he then realized that by changing the pump twice he had introduced quite a bit of air into the loop, which might be causing a problem, so he first temporarily connected the pump backwards, which finally seemed to make something flow, then he connected it forwards again, and after some gurgling noises it finally started working properly. Since then it has been working fine so far (somewhat better than before, probably because of the new pump).

However, we are still unable to explain how all these failures occurred simultaneously. It's just about possible that since the thermocouple had been holding the pilot light gas valve open continuously for years, the pilot light getting blocked could have triggered the first operation of the gas valve for a very long time, after which it failed. But how would that link to a pump failure? The heating engineer speculated that perhaps somehow a gradual pump failure caused the boiler to overheat, somehow triggering a problem with the pilot light, and that the gas valve was already about to fail. The bad second pump was just bad luck, as was the incorrect wiring (ignored for now, as we never use heating without also having hot water enabled) preventing the specific test from working.

Although it's all working now, I find it very unsettling to come across everyday things which seem to violate the laws of probability in this way!

 

[phinds]

I certainly agree w/ you that it is VERY unlikely that there were multiple independent and simultaneous failures. Much more likely it was, as you hypothesize, some sort of as yet unexplained cascading effect.

 

[russ_watters]

Im not sure i understand your system (pilot light+electronic ignition?) It seems to me that the pump failure could cause the other two failures by making the boiler turn on and off repeatedly until they failed.

 

[Tom.G]

For what it's worth.

• Heating Timer was miswired at initial installation and you never knew because you never used the heat-without-hot-water function
• The gas valve failed and was replaced (appearently the root cause)
• The servicer tested a function you had never used (heat only) and concluded a pump was bad
• System was opened twice for pump replacement, causing air bubble/vapor lock
• System purged by temporarily reversing pump and now system working normally

Looks an awful lot like one failure (gas valve) combined with a usually-valid assumption of a bad thermocouple (it should have been tested though) and an unknown original wiring error.

More like a Comedy of Errors than "multiple independent and simultaneous failures."

 

[Jonathan Scott]

That's a interesting theory, thanks, but I'm fairly sure the test which was impacted by the miswiring resulted in the pump and boiler not switching on at all, and it still doesn't explain the connection between the gas valve failure and the pilot light apparently getting blocked (until it was cleaned with a needle). That could however be explained by the thermocouple release triggering a pending gas valve failure.

 

[russ_watters]

...and it still doesn't explain the connection between the gas valve failure and the pilot light apparently getting blocked (until it was cleaned with a needle).

A couple of possibilities;
1. Cycling the valve knocked loose some corrosion which clogged the orifice.
2. The lack of flow through the orifice exposed it to oxygen and dust/dirt.

 

[Jonathan Scott]

Cycling the valve knocked loose some corrosion which clogged the orifice.

Thanks - that's an interesting theory, but it still raises the question as to what caused the valve to cycle if it wasn't the pilot light going out due to a blockage and hence causing the thermocouple to release the gas valve (although I must admit I don't know exactly how the mechanics work inside those valves).

I find it hard (although not impossible) to believe that the gas valve spontaneously failed when it had been open for years, so I'm more inclined to believe that being closed for the first time in ages is what triggered the failure. I guess that the initial failure could have been something else unusual, for example pilot light blowing out due to unusual wind direction (I did have the back door open for a while that afternoon, but it wasn't unusually windy), triggering the gas valve failure and releasing some dirt which then clogged the pilot light. But that still doesn't seem to have any causal connection either way with the pump being too weak!

 

[Bystander]

How old is the system? There are "safety" features required/recommended by code(s) in various areas of the country that will interrupt the thermocouple (stack switches and the like) that do not age gracefully, nor are they intuitively or analytically obvious to even qualified inspection/inspectors. The valve on my gas fireplace is an example; corrosion on the contacts for the stack switch resulted in several "failed" TC replacements before the problem was identified.

 

[Tom.G]

I have lived in some areas with enough contaniments in the natural gas to plug orifices over a period of years. This would even affect the kitchen stove.

I would go with:

• Pilot light clogged causing it to extinguish
• This caused the gas valve to close as a required safety measure.
• The valve, having been exposed to contanimated gas over the years, had a deposit build-up that prevented it from re-opening
• Continue with my post #4, above

p.s. A post mortem of the gas valve would be interesting.

p.p.s. I'm applying Occams Razor to this situation, based on similiar, but not identical, situations I have run into in the past. (The Razor could, of course, be dull!)

 

[Jehannum]

Pilot lights and thermocouples!? Just how old is your boiler? If you're in the UK it must be well over 20 years old because such things haven't been used in domestic boilers for many years.

Regarding gas contamination, the odourising components added to natural gas sometimes causes sulphidisation inside copper pipework which can block it completely. I've seen this happen with 22 mm copper - the copper sulphide came out as black discs with the same diameter as the pipe.