Tell me why I just spent $75.

September 21, 2007 7:30 AM

So the electrician said nothing was wrong. But something WAS wrong, only it fixed itself by the time he got there. Help me diagnose this maddening problem our tenant had.

So our tenant tripped the main breaker Wednesday night. I reset that and went in the apartment to check the breaker box, and nothing inside was tripped. This can happen, but isn't normally supposed to. In any case, all seemed well, the breaker didn't retrip, and I went to bed.

Thursday afternoon, I was called over because half her appliances were dead. The hot water heater, the HVAC, her electric range, and her fridge were all kaput. Her lights were on, as well as the TV, and most other things in the apartment. NO breakers were tripped. I reset them one by one by the book (off ... pause ... on) and still none of these came on. Even a fluorescent light in the closet that's on the same breaker as the hot water heater worked while the heater itself did not. The HVAC, on a wholly dedicated circuit, did not come on even after resetting. The stove's idiot light was even off, let alone any of the burners. The fridge was the odd man out -- the only appliance on a normal circuit, though it naturally had a dedicated outlet in the kitchen. I tested and reset every GFCI outlet in the apartment (one bathroom, two kitchen) to no effect.

Over the phone the service manager for the top-flight electrician in town (regional contractor with sterling reputation) explained that it could be the breaker was experiencing "loss of phase", i.e. one of the two 120V supply lines that make 240V service possible. This seems very plausible in that the big-ass appliances wouldn't come on, but not the fridge.

Just to check myself I went back to the apartment last night with a pigtail tester and verified that the situation had not changed. Most frustrating, I showed power to the fridge outlet but nothing happening inside.

Despite this sanity check, when the electrician showed up at 7am everything was working again ... except the fridge. He opened up the house panel and tested the points inside, finding nothing amiss. He opened up the main breaker outside and found nothing to concern him there, either. I asked about moisture (we've had some trips that could have been wet-related) and he showed me how there was no evidence of corrosion inside the main breaker box. But since he had verified the panel, the breaker, and even the fridge receptacle, there wasn't anything he could do that wasn't costing us money while stabbing in the dark. So I took it as a costly sanity check.

I'm assuming the fridge is, in fact, dead. It's not especially new and it could easily have kicked the bucket due to a power fluctuation. (But would that kill the light?!) But I can't rest easily knowing I have people living in a place where there's an undiagnosed power glitch.

To inoculate some questions:
* The fridge being plugged in was not a factor. It was plugged in last night and this morning.
* The hot water heater must have kicked in with everything else sometime in the dark of night. It was not all the way hot yet.
* I did NOT unfortunately pigtail test the stove receptacle (last night).
* The tenant added a satellite dish (w/o permission, and I'm pissed about it for other reasons) which is the only recent change to the apartment. That couldn't have done anything ... right?
* No other tenants reported any problems, so it isn't a problem with the building service.
* The interior and exterior electrical equipment is all Square D, and I agree with the electrician that this is a very solid brand. He said loss of phase was possible but very rare, unless you have another brand.

I've read that loss of phase can screw up motors, including compressors. Could that have led to the fridge failure many hours after the circuit was reset? But the fridge being plugged in should not have affected other appliances, unless they too were affected directly by the loss of phase. But why would it kick back in by itself in the middle of the night?!

Summarizing, I don't object to the call as it eased my mind about a couple of different things. But I wish I could have gotten things working last night and avoided it entirely. I did worry that we had to replace a hot water heater AND an HVAC, both relatively new (last 2 years), and that didn't happen, so that's good. The stove belongs to the tenant, and so she can't make a damage claim against us, so that's good. And we have to buy a new fridge, so that's bad. And we don't know why it happened, so maybe the new fridge will get the same treatment at some point?

Did I miss any obvious steps here?

posted by dhartung to home & garden (17 answers total)

Does the fridge circuit take it's juice from the opposite power bar than the appliances/circuits that stayed on? If so then the phase loss would kill the 240V appliances, and the fridge, then when power came back on it spiked the fridge and killed it.
posted by zeoslap at 7:46 AM on September 21, 2007

You can get a repair person to check out the fridge. It may have an internal fuse that has blown. Or a wire that has melted. Their diagnosis might solve the mystery, since it is the only thing in the apartment with any evidence of the event.
posted by Eringatang at 8:31 AM on September 21, 2007

Refrigerator: I don't recall refrigerators having fuses (for some reason), so the light not coming on sounds like a wiring fault. Do you see a wiring diagram in/on the refrigerator?

Possibly the refrigerator caused the whole thing. Maybe a wire shorting out. This should have tripped a breaker, but...
Losing one of the hots to the 240V could certainly cause things like stove lights on with burners not working and HVAC/water heaters not working.

No idea on how it 'fixed itself'. Does the apartment have its own line from the street or does it come off a breaker in your house panel?
posted by MtDewd at 8:32 AM on September 21, 2007

We've had the problem of half the power being out. It didn't actually fix itself, but it was fixed without our intervention--the whole block was affected, and someone else called the power company to come out and fix it. No idea about why the fridge isn't working, though.
posted by MrMoonPie at 8:49 AM on September 21, 2007

Perhaps one of the poles in the main breaker is failing? It could be intermittent for along time before failing completely. I had a failed breaker, and you could hear the arcing inside it when trying to use that circuit.

When power has been out for a while, fridges and freezers have to operate for long hours to re-cool everything, and this can over-stress them (I get the feeling you've thought of this). I'm with you, I wouldn't think the light should be taken out by that situation, but it isn't beyond the realm of believability.
posted by Chuckles at 9:39 AM on September 21, 2007

I'd give the fridge a quick once over. Even if it doesn't have a fuse/breaker, the extended stress might have burned out an already poor connection. An easy fix, if the fridge seems to be worth fixing. Did the tenants notice any symptomatic smells?
posted by Chuckles at 9:42 AM on September 21, 2007

I think you have two problems.

First, your refrigerator has some bad problem that caused it to start drawing so much current it tripped your main breaker, and then later caused the voltage in one leg of your 240V supply to fall to a level where your 240V appliances could not function. By the time the electrician arrived, the problem had solved itself because the partial short circuit in the fridge had finally burned itself out and become an open circuit. I believe some failure modes of the compressor motor start circuit could have caused things to act like this.

Second, I think you either have a bad circuit breaker or an inadequate level of service to the apartment. The breaker the refrigerator is connected to should have tripped long before the main breaker did unless total normal demand from your tenant's apartment is running too close to the limits of the main breaker.
posted by jamjam at 10:07 AM on September 21, 2007

I guess nobody thinks the satellite dish may be improperly grounded? Thing is I'm not even sure how I'd verify that.

The fridge does seem to be an "appliance of interest" but I'm not sure we have evidence to press discharges at this point. I will look at it with a voltage tester after we get it moved (the new one comes tomorrow, sorry about her food, but I'm not to broken up about that for another reason I won't mention).

I am really concerned that the internal breakers aren't tripping first, but the electrician said that a blown (or incompletely reset) phase could behave that way (as opposed to a straight overloaded circuit), so maybe that's still the reason for that particular part of it. He still thought there would be visible evidence of that kind of failure, though, or all four apartments would have been affected a la MrMoonPie. There was definitely no unusual noise that I heard when I was resetting that breaker, or examining it yesterday during the weird partial outage.

As far as service levels, I think we're well within code specs. She's the only tenant with a big-screen TV and I have no idea what that thing draws. I think it's about the largest tube-based TV you can get. With a new (factory close-out) more efficient fridge, there will probably be less competition for the available power -- I hope.
posted by dhartung at 12:56 PM on September 21, 2007

What do you mean by the dish being improperly grounded? I've not inspected the equipment closely but isn't there just a coax cable from the dish to the box, possibly with a power supply? Have you looked at the dish setup to see if things look okay there?

In any case I'd separate your dislike of the tenant from the problem at hand, or at least keep it out of the post and comments.

As has been suggested the fridge seems to be the most likely culprit but you did state the "tenant tripped the main breaker." Do you think the tenant actually did something to cause the failure? Aside from the dish. If you ask don't be accusatory but you can see if she had a lot of appliances on at the time it tripped. Anyway, it sounds like you've already got a new fridge on the way but it wouldn't hurt to have someone look at it just to figure out what went wrong, if you can't diagnose it yourself. Is there anything else on the fridge's circuit? What's the rating of the breaker for it?
posted by 6550 at 3:12 PM on September 21, 2007

Did you actually check the voltage, or just for the presence of voltage? I have seen some strange things during brownouts. I have also seen lots of things die when the line is dirty (constant dips and spikes, caused by some heavy machinery in a warehouse below an office building, line finally checked with an oscilloscope).

Probably a shot in the dark, but could someone outside of the apartment have been using a lot of juice for something?
posted by bh at 3:54 PM on September 21, 2007

"I'm assuming the fridge is, in fact, dead. It's not especially new and it could easily have kicked the bucket due to a power fluctuation. (But would that kill the light?!)"

It's unlikely though there are few frost free models that use the defrost timer as a junction box. A blowout of the wiring in that location could take out everything. This is a $50 fix by the way if it was the case. It is also possible that when the compressor went it surged the voltage to the bulband burnt it out, wierder stuff happens.

It's really unusual to trip a main without tripping an individual circuit. A failing main breaker could cause this. To trip it normally would require a huge load, something that even a good direct short at the fridge outlet can't do because the wire size isn't large enough.

It's really unlikely the dish is the problem unless they managed to drive a screw into an electrical cable or attached it to the service mast. All grounding the dish does is provide some minimal protection to down stream equipment from lightning strikes.

It's unclear what the electrician means by phase. It is wildly unlikely you have anything but a single phase service. Losing a phase would mean no power at all. He could mean you lost one of the hot legs of your service. This would indeed present the symptoms you describe.

One thing you might want to do is use an actual meter if you are just using a light up tester to read the voltages across your two hot legs and across the legs to neutral and ground. You should see ~220-240 across your hots and ~120-120 between each hot and ground or neutral. You should have 0 volts between ground and neutral. Any significant potential between neutral and ground indicates a serious problem. Check this at several wall outlets and at your range and dryer outlet. If your A/C condensing unit plugs in you can check it there too.

Weird stuff can also happen if all the 110 volt loads by chance happen to be on the same leg of theservice. This causes an increased load on the neutral leg. No way of knowing if your tenant has this without mapping all the loads to the panel. Every second breaker in a SquareD box (SquareD is the best IMO too) is on the same hot leg.

This type of intermittent power weirdness can be seen with a bad service supply from the electrical company. A floating ground especially can cause all sorts of whack things to happen. There is a fair chance this is your providers problem; I'd give them a call just to check and make them aware you are having an issue. The electrical company here will sometimes lend you a voltage recorder if you suspect you have a problem. Many cheaper digital multi meters include a min/max recording feature. It won't give you a great picture but it will at least let you know if the voltage is surging up to 270V or dropping down to 180V.
posted by Mitheral at 6:34 PM on September 21, 2007

6550, it's simply that my IT troubleshooting background leads me to look at the last thing that changed, and that's the last thing that changed (which, repeat, was done without our foreknowledge or consultation, so I have a perfect right to be pissed if it ends up relating to this problem, which has cost us plenty). The prior tenant (still in the same building, switched to a bigger apt) reports that she never experienced anything similar.

And I've rarely seen a fridge just go "dead" like this. Generally the compressor just works harder and harder and we get a call around the time it can't keep the ice frozen.

In any case, Mitheral answered my question about the grounding sufficiently that I'm dropping that angle.

bh: I used a pigtail "light up" tester that has "110V", "220V", and some other idiot lights. It showed voltage. I had a better device, but it got fried at some point by being on the wrong setting when it touched hot; I suppose it's time to get another.

Mitheral: You're most likely right about the hot leg vs. phase question. He wasn't talking down to me and I don't feel he was deliberately (mis) using jargon since he wasn't trying to sell me any service thereby. I'm thinking the power company angle is a good idea.
posted by dhartung at 8:28 PM on September 21, 2007

It's really unusual to trip a main without tripping an individual circuit. A failing main breaker could cause this. To trip it normally would require a huge load, something that even a good direct short at the fridge outlet can't do because the wire size isn't large enough.

Mitheral makes excellent points here, particularly the one about the wire size. You mention one other thing I think you could take a look at, though I hate to risk increasing your animosity toward your tenant, and that is her stove. If her stove pulls a huge amount of current in some modes of operation, whether by malfunction or design, it could in combination with the water heater and the HVAC, say, cause the voltage in one leg of your supply to fall quite low.

This in turn could have caused your refrigerator problems. Refrigerator compressor motors typically have start windings which produce a lot of torque and draw a lot of current when they first turn on. When the motor reaches a certain speed, a centrifugal switch is tripped, which causes the start windings to turn off and the run windings to operate. The run windings draw much less current. If the voltage the refrigerator was seeing when it turned on was too low, it might not ever reach the speed necessary to trip that centrifugal switch. But the start windings are not designed for sustained operation, and energizing them for any extended period would cause them to overheat and then ultimately burn out. I would be surprised if even an older fridge was not thermally protected, however. Some have automatic resets when they cool off after they overheat, but I've seen some with manual resets as well. The manual resets I've seen can be accessed by pulling off the ventilation grill under the door. Look for a rod with perhaps a red or orange end sticking out. Push that rod in to the point it clicks and try turning the fridge on then.

If that's the situation you find, and the refrigerator is actually OK (plug it in and try it anyway, the automatic resets can take quite a while to kick in sometimes) I would go back and look into the stove. I would start by running the hot water enough to turn the water heater on, turning the HVAC on, and then turn the stove on maximally and see what happens.
posted by jamjam at 11:32 AM on September 22, 2007

Well, the DOA fridge is now in my garage. I'm gonna put it on a GFCI extension cord and start messing around with it that way. The new one is showing up any minute now.

Her stove is practically brand new, her pride and joy (for her ethnic cooking). I don't mean to appear so passive-aggressive but I have good reason to nastygram her already, despite her having been a good tenant when she lived here previously. The last thing I want is conflict with a tenant but some things are impossible to ignore, have already cost us time and money, and that's probably all I should say.

The building service (and unit service) seems to me to be above average. The building was gutted and rebuilt in 1996 and everything but the heating is electric, so it's not a matter of old wiring of any kind and I'm reasonably sure it still meets code. I'll have to take some figures and do some back-of-the-envelope calculations, but I will still probably get the measurement equipment from the power company.
posted by dhartung at 12:58 PM on September 22, 2007

A kill-a-watt is never a bad instrument to have in your arsenal. It's a good way to see the power factor of an appliance, which is the phase relationship between the current waveform and the voltage waveform.

Circuit breakers operate either thermally or inductively, and a very low power factor can confuse them. If the branch circuit breakers sense one way, and the main senses the other way, a power factor problem could produce the symptoms you describe. I think. (AC was never my thing.)

My main thought, until I read the latest update, was "try the fridge on an extension cord from another known-working outlet". I guess your garage will work! I'm gonna check this thread for a few days and see how it pans out. Keep us posted!
posted by Myself at 3:28 AM on September 24, 2007

I'm in the middle of a flea-bombing mess for the next day or so .... but I will get back to that fridge (if only so I can get it out of the damned garage).
posted by dhartung at 12:07 AM on September 25, 2007

OK, I should have known.

The fridge is working fine -- the old fridge, the one that was moved to our garage, the one that wouldn't turn on when it was plugged in to the kitchen receptacle that the electrician said was pulling 120V.

So, apparently, is her new fridge, plugged into that same receptacle.

But the water heater now doesn't heat up a full tank of water. (Usually this means the top one of two heating elements has gone bad. A bottom element heats water that then circulates by convection, and can still heat the full tank, just more slowly.) This is a WH that is less than one year old -- and it's a warranty replacement of the previous one, which began leaking out an element gasket when it was less than a year old. I suppose I should consider the possibility of a manufacturing defect, although I can't see how a bad element translates into a blown breaker other than that (like a failing fridge) it might be on all the time.
posted by dhartung at 10:01 PM on September 25, 2007