Low-water cutoff turning off and on

saikosis

Two days ago, I cleaned out the sight glass on my boiler. The glass was dirty, but it wasn't clogged. While I had the glass out, I opened the bottom gauge valve and let the water run out until it was clear. I also opened a drain valve and let it run clear. It didn't take much, less than a gallon total. When I had everything back together and turned the boiler on, the low-water cutoff kicked in pretty quickly and triggered the automatic feeder. The number on the meter didn't change, so I assume less than a gallon went back in.

Yesterday, I turned on the heat for the first time this season as a test. The boiler worked fine. After the house reached temperature, I lowered the thermostat to keep the heat off for the rest of the day. Later that day, I noticed that my thermostat was offline. When I checked the boiler, I saw that the light on the cutoff was turning on and off. It was making a click with each transition. Every time the light came on, the thermostat went offline.

While the cutoff was cycling, the water in the glass was calm and about even with the probe. It cycled through a few of these on/off transitions while I was watching, but the automatic feeder never kicked in. I eventually used the manual bypass to raise the level above the probe. (I only learned just now that the feeder has a FEED button that I could have used instead.) After raising the water level, the cutoff light turned off and stayed off.

I've got a Weil-McLain PEG50. The cutoff is a Safgard 400. The feeder is a VTX 24 with the switches set to LWCO and 2 minutes. The thermostat is a Nest Thermostat E (with all the smarts turned off). I added my own C wire, connected directly to the ground screw.

I figure there are a few possibilities.

1. This is totally normal and happens all the time, but I just never noticed before. Maybe the cutoff waits for a certain number of low readings before it calls for more water. If this was the case, then it probably would have recovered on its own if I was more patient.
   
2. The probe is dirty and was getting a false reading. I've never cleaned the probe myself. I don't know if it's done as part of the annual maintenance.
   
3. The probe is bad. The boiler was put in 10 years ago, and, as far as I know, this is the original cutoff.
   
4. I got really unlucky and the water level was exactly at the probe level, causing the reading to bounce. If this is true, I'm not sure it would have ever gotten out of this state on its own, which means that filling it manually was the right call.
   

What do you think?

I suppose a good test would be to call for heat, slowly drain the water, and make sure the cutoff kicks in. Anything else I should check? Any good videos showing how to clean the probe?

I understand that the cutoff should kill the boiler. Should it also kill my thermostat? It seems overly dramatic, but I guess it doesn't matter as long as the low-water condition is rare.

Thanks.

neilc

I thought I saw the word Nest,
yup, I did,

is the Nest a new addition, or even a year or 2 old?
folks have had issues as the Nests go looking for recharge power,
I wonder if it's pulling power from the LWCO and relays,
any chance you try an old thermostat there just to rule that out?
(put the Nest back if I'm wrong)

it is suspicious that the waterline is even with the probe height,
maybe the cleaning and flushing stirred up a top scum layer that's trippin the probe,
and or landed on the probe,

if you didn't break your sightglass then you ought be able to fanagle the probe out for a look,
wipe it clean with paper towel,
do not use tape putting it back in, only paste,
tape can isolate the base from completing its circuit,
paste only

post a picture of that sightglass,
if it was dirty you may have carryover, oiley water washing in from the top,
what's your skim port situation ?
picture?

Jamie Hall

Depending on how the system is wired, it wouldn't be unusual for the low water cutoff to kill power to the thermostat as well. In fact, it would be rather common.

Now. The way your VXT is set up, if the water level is right at the probe (sounds like it is) and there is any bounce at all or even uncertainty in the electronics -- not necessarily the water level -- the VXT won't feed. Why? Because it will only feed when the LWCO is calling, and if that is only momentary, with a two minute delay on the feed, the poor thing gets confused. Try resetting the VXT to feed a gallon or so after the delay and see if that helps

EBEBRATT-Ed

I would clean or have the probe cleaned. I would have the Nest wired so it does not lose power from your limit controls or from someone shutting the boiler on and off. You may need a relay and an additional transformer to accomplish this.

saikosis

My theory was that I just got unlucky with LWCO. When I cleaned the glass and drained some water, the boiler hadn't made steam in several months, so there was no condensate coming back. The VXT was set to trigger on LWCO, so it stopped as soon as LWCO was satisfied. That could have left the water level right at the probe. If the boiler had been steaming, condensate coming back would have raised the level to somewhere above the LWCO.

I'm interested in the oily water theory though. My glass was squeaky clean after I cleaned it. A week or so later, it's dirty again at the top. The boiler has fired for the indirect water heater during that time, but it hasn't fired for steam since the test I mentioned in my original post.

You can see some dirty residue in this picture:

Here's the front of the boiler:

Is the skim port the one between the glass and the outlet?

Here's the inside of the LWCO:

How do I get the probe out? I couldn't find a video, but it looks like a hex nut might be hiding behind the box. What should I use for paste? I've got Rectorseal #5. Is that the right stuff or should I get something else?

EBEBRATT-Ed

Disconnect the low water cutoff from the probe. Shut off the power first. You have to disconnect the probe (usually a small wing nut). Then loosen the two screws near the probe and rotate the control so the screws fit through the key holes. Drain the boiler below the probe level and remove the probe. Rectorseal #5 is fine. Any pipe dope is ok (no Teflon tape).

Reverse the procedure to re install. Fill the boiler before turning on the power.