24

The cabin electric bill came and I gave it what I thought would be a cursory glance and it looked unusually high. Not just for this cabin, but in my experience with our houses. Our main house is a 1965 crapshack in Northern Virginia, very poorly insulated, drafty, and in summer it's 80+ degrees there so the central AC is running all the time.

EDIT: Here's a snapshot of the past 2 bills at that hot-weather house, where the AC has been on for both months (newish Trane XR16 on XV 80 furnace/blower). It's ~900 kwH, so somehow less than the cabin. At this home, we have the same cameras, 2 ceiling fans running all the time, older fridge, plus a dehumidifier in the basement and 2x coway air filters, half dozen assorted wall warts.

enter image description here

Here's the past 30 days for the cabin. Again, No AC, cooler temps, but more juice used:

enter image description here

Somehow, our electrical use at this northern cabin is the same to slightly higher than at that house in the heat. I'm stunned, because here, there's no AC running at all. (It's broken, and turned off.) While we've been here (right side of chart), the heat hasn't been on either. (Heat is gas, in-floor, so only electrical needed is to run the circulator and exhaust blower.) That's even more surprising to me because this place is really well insulated. Our cooling this summer consists of opening the windows at night to gather up the cold air. Wake up, house is about 65 degrees. Close all the windows, and just keep doors shut, and on an 88 degree day, house is still <74 degrees by evening. (At the Northern Virginia house, if the AC is off b/c of power outage, the house zips to the low 80s in a few hour's time.)

So, I'm trying to figure out where we're wasting electricity.

electric use before and after occupancy at cabin

I compared the baseline of current use to before we arrived, and it looks like the average spring/summer use per hour when no one's here is 0.25 kwh/h. In the lower chart, occupied timespans are underlined in red.

Appliances running unoccupied:

  • Fridge
  • In-floor heat circulator and exhaust fan
  • Well Pump
  • water softener
  • Modem
  • 2 box fans
  • 4 USB power warts for security cameras
  • 2 LED floodlight security lights (~30 watts) that run dark-dawn
  • 2 60 watt front yard lights dusk/dawn.

Appliances running when occupied:

  • Fridge
  • In-floor heat circulator and exhaust fan
  • Well Pump
  • water softener
  • septic pump
  • bathroom fan 1-2 hours / day
  • Modem
  • Ceiling fan on low 24 hrs/day
  • 3 box fans
  • 4 USB power warts for security cameras
  • 2 LED floodlight security lights (~30 watts) that run dark-dawn
  • 130 watt TV 3-5 hours/ day
  • 3 laptops 3-6 hours / day
  • room lighting (almost all LED 4-9 watt bulbs)

The hot water heater is on an entirely different meter, so nothing to see there.

There's of course a washer and dryer and dish washer and stove, probably those bigger spikes in the chart.

What I'm trying to figure out is the lift in the baseline load from 0.25 kw/h unoccupied to about 1.25 kwh/h. I'm running out of ideas. (Again, there's no AC running, and no heat.)

I have been trying to figure out where the extra 1.0 kwh/h comes from. In my head, I see that as the equivalent of having 10x 100 watt lamps running 24/7, and I can't think of what would provide the equivalent draw, since we have LED bulbs everywhere, and we're not big light leaver-oners anyway. There's a ceiling fan running on low all the time, but I doubt that's using more than 1 kwh for the whole day. We've got a few more fans running in evenings, 3 of them at about 0.8 amps each, so even 10 hours of use there doesn't add much (and the nights don't jump all that much in the chart).

I thought maybe well pump? But a 1/2 HP pump at 10 gal / min for 3 people who use less water than average (I've checked the throughput in the softener) only comes to like 0.6 kwh / day.

Computers? We've got 3 laptops on 200w power bricks, but they only run a few hours a day (and that's less computer draw at that house back in Northern Va where the big machines are...) TV? A 130 watt flat panel that runs a few hours a day.

As far as I can tell, we run fewer and lower-power devices here than at the hot house in NoVa (because it's the cabin) but somehow we're using more electricity.

We're even drawing more electrical than when we were here last fall (see underline on left, lower chart range) when we had all the same stuff (fridge, washer/dryer, TV, stove, etc.) and at that point, the heat was running consistently, so we had the boiler exhaust fan and circulator pump running a lot. Moreover, the security lights would have been running many more hours per day then, because there was so much more dark time.

I feel like our meter is registering a summer AC surge for us, without realizing the AC isn't on. It feels crazy to me.

EDIT: Here's a month worth of winter, unoccupied "baseline," which is the circulator running the hydronic (house set at 63 deg), 2 window / box fans pushing that air into far corners, fridge running, security lights, etc. as above. There are electric baseboards set to very low temps in two of the colder rooms where there's plumbing to act as backup against frozen pipes, but AFAIK those never kicked on (no real green spikes). Caretaker said lowest temps he saw in house were ~58 deg. Load kicks up noticeably on the cold days. Orange spikes are water heater (separate meter) kicking on once in a while.

Thanks for all the comments, interest, and ideas. I'll probably start flipping breakers or using a kilowatt next to try to pin down the use. I'll update along the way.

enter image description here

18
  • 2
    My guess is the well-pump or the heating circulator/fan (or both). Water softener could also have more load than you would expect. Especially if it has and internal circulation pump that keeps going all the time. And what about those 2 "box-fans" (whatever those may be?). That is a whole list of always on stuff that may draw more power than you realize.
    – Tonny
    Jul 26 at 10:23
  • 11
    1kWh/h is just 1kW, isn't it?
    – fraxinus
    Jul 26 at 11:05
  • 7
    Do please keep us updated with your findings as you progress.
    – Criggie
    Jul 26 at 11:47
  • 4
    If something is wrong with your pressure switch or if your water line has an obstruction it's possible that the well pump is just running non-stop. You may also have a water line leak that is causing your pressure tank no not hold pressure and your well pump to run continuously or mostly continuously. Have you checked that?
    – J...
    Jul 26 at 13:57
  • 4
    If any answer was useful, please accept it and tell us your findings / what you've learned (you can edit original post with update) Jul 26 at 16:35
43

Well, the load is 100% consistent hour by hour, so it is not normal usage that would be very spikey and much lower at night.

So it should be observable by watching the meter spin (or marquee lights or whatever your smart meter provides for instantaneous draw indication).

Ramp down any loads you can, and stop people from throwing random loads on for awhile. Then turn off half the panel and see if it basically stops. Don't "hold out" certain breakers on the logic that they are important; turn them off!

If turning half the panel off doesn't slow it down, then turn those back on and turn off the other half.

Once you find which half makes the load go away, turn off half of those breakers. Divide and conquer thusly, until you find the 1 breaker that makes the spurious load go away.

If you can't find it that easily, then it may be a water leak causing the well pump to spin up day and night.


That reminds me: the load may be not continuous, but intermittent. So if the load is not immediately findable, then take the slow road. Re-do the sequence, but leave the breakers off a couple of hours -- long enough for the usage to appear on that "app" you are monitoring. Don't go too long on refrigerator or sump pump circuits, though.

8
  • 23
    Binary search on the breaker panel. If that doesn't narrow it down to a single breaker you have something directly wired after the meter and before the breakers. Which is a MAJOR problem that needs to be sorted. But the wiring for that can only be between meter and breaker-panel so should be relatively easy to trace.
    – Tonny
    Jul 26 at 10:25
  • 1
    Would a clamp/ring meter around the phase wire of each circuit in turm, provide a better picture than going by eye? Good excuse for new-tool-day.
    – Criggie
    Jul 26 at 11:47
  • 3
    @Criggie the main limitation is that a clamp meter can only look at one circuit at a time; which means it'll take longer on average to find the offending breaker than dividing by halves each time. It's a linear vs log time scale, how much worse it is depends on the total number of breakers. eg 4 vs 2 tests if you have 4 breakers; 16 vs 4 if you have 16. Assuming you only have 1 large mystery load; if for some reason you've got a half dozen smaller ones that are being missed you'll end up having to iterate over most of the binary tree anyway and a clamp meter might be easier then. Jul 26 at 13:40
  • 1
    We're looking for a kWh/h so If the problem is intermittent, with a duty cycle of 50% or less it would have to be a greater than a 2kW load. No guarantee but I would expect you could hear/feel/smell or somehow sense most large appliances running continuously like that. It's pure guesswork but I'm hoping for something continuous and closer to 1 kW load.
    – jay613
    Jul 26 at 18:41
  • 1
    @jay613 I noticed :-) And I know from past experience that many coders don’t realize that some programming concepts have applications outside computers. Figured a little hint couldn’t hurt.
    – Tonny
    Jul 26 at 21:43
12

That is not going to be an easy find. I would start with the well pump. When you are not using water turn off the power for a few hours and monitor the pressure or come back and read it. It should not change. If it is going down you may have a leak causing the pump to cycle. If it is a jet pump one of the tubes going down or the foot valve could have a small leak. If it primes easily just turn it off when not occupied. I have seen this before and it is extremely difficult to find. They make devices that you can plug in that monitor the current consumption of an appliance, you might want to borrow or purchase one and monitor each individual appliance etc. Start with the refrigerator. Replace your 60 watt front yard lights dusk/dawn with LED bulbs. By the time you get through this list you will know what your problem is. Let us know how it comes out.

10

The binary search option above would work, but would be time consuming because you'd have to wait until the next day (in my case) to see the power company's refreshed usage graphs. And you'd have had half the house off during that time. (And I have 3 breaker panels, meaning this would have taken forever...)

The simplest solution was what I stumbled into by accident: Have the power company send a tech (if they will) to open the breaker box and test the current being pulled by the various breakers.

Or, if you're qualified, do that yourself.

Here's the latest power graph showing that after replacing the septic pump, the use went way down. Somehow, that septic pump was drawing more power even than the new AC unit I had installed a week later.

enter image description here

EDIT:

The phantom power draw was definitely the septic pump circuit, a combination of improperly installed float originally (too low in tank), a cheap pump's wearing out, and horribly corroded electrical adding to the power draw, perhaps. The pump empties the tank only to a certain point, then can't push the water any further, so the float stays horizontal and the pump runs constantly.

I pulled the tank cover and took a look at the electrical. Yikes. This provided the power to the pump and float switch. (Piggy-back, pump plugged into the float plug.)

enter image description here Obviously, that needed immediate replacement, but it's still inside the riser for now, pending installation of an external pedestal, as should have been done originally.

It was drawing 10 amps, so at 110v, there's the extra kilowatt of constant energy use! With a new, clean outlet, it draws 8 amps when running.

Solution will be a new pump and float switch, new alarm float, new external power pedestal (and a new plastic riser too...)

Backstory to eventual final answer:

Why was it so hard to figure out? Well, the breaker labeled "septic pump" was wrong, for starters. Moreover, the pump was quiet enough (and deep enough in our cement tank) that when I did go listen for it, I heard nothing.

Since we have 5 separate breaker boxes here, I called the power coop and they sent a tech. He took a look and said "I have a breaker labeled "utility" that's drawing 10 amps. Well, there was nothing obvious in the utility room, then the tech noticed a thick black cable. Where's that go? I pulled up the floorboards and it went into a big steel box in the sump pit, and from there several wires emerged. Some of them were really thick -- probably the abandoned in-slab electric heat.

Only one was energized, a white nm cable heading down into the sand. We headed out, gave ourselves hernias pulling the cap off the cement tank, and the pump was running! I grabbed a rake and got down into the manhole and wiggled the float and it went off.

Since it started running the day we arrived, my working theory is that after months of no use, the float was gunked up somehow, and when we filled the tank and it floated up and switched on, it didn't return after the pump.

I'll keep an eye on it to see if the float continues to malfunction, and I wonder if running that pump 24-7 for 6 weeks hurt it, but at least we figured out the problem. I'll edit with the power use graph when that's updated.

EDIT:

Here's a chart showing (on the right) a few hours of the pump with the breaker entirely off, then back on, then off again, then on overnight. As mentioned, when I jiggled the float with a rake, it shut off, so I thought maybe the power draw problem would be over. Nope, when the breaker's on, it clearly draws all that power again.

enter image description here

Bonus problem, when the breaker's on and there's power to the pump, now I'm not sure it even runs at all, because we went out there, pulled the cover, heard nothing and I pulled the float up to try to get it to run, and it didn't. Tester showed it was getting power, so now I wonder if the pump isn't burned out and just drawing power and converting it to heat like the seized impeller in the hydronic system in the garage last year. (Yeah, the guy who sold me this place was a crook.)

1
  • 2
    Thanks so much for following up! I've been following this one (and I imagine some others are as well!) Aug 4 at 0:05
4
  1. If you haven't turned off the breaker for the HVAC system (or don't want to), go outside and pull the disconnect for the A/C outdoor unit. You said it's broken, but if the thermostat is calling for cooling and the compressor is seized, it could be continuously trying to start but popping the thermal protection every few minutes. There's also the compressor's crankcase heater that would be running if the HVAC is energized at all.

  2. Has your water softener run out of salt? If you have one of those newfangled "smart" water softeners with a sensor in the resin bed that only regenerates when the resin bed's softening capacity is depleted, it could be regenerating every night. If there's no salt, it's regenerating with plain water and the resin bed will stay depleted. Regeneration uses many gallons of water.

  3. If the well's foot valve or check valve is leaking, your pressure tank will send water back to the well, and your pump will run every few minutes. Observe the gauge on the pressure tank when no water is being used, and see if the pressure drops, then rises.

  4. Also, if the water line between the well and the house has a crack or split, the pump could be running continuously, with thousands of gallons of water returning to the water table from the underground split, yet no apparent problem above ground if your soil has high permeability. Turning off the well pump overnight would result in an immediate drop in power consumption.

3
  • There's no thermo calling for cooling as it's a mini split. On-demand operation only. And, I flipped off the circuit out there when I figured out it wasn't going to work. (Just in time for 90 degree days!) Jul 28 at 21:50
  • Softener's full, but good idea. (It's also LOUD on regen, so I'd know if it were going rogue.) Jul 28 at 21:51
  • Pressures good on the house water. Pump only runs when you'd expect, showers, washing, consecutive flushes, etc. At least from what I can hear. (I hear everything, as I sit right on top of the utility closet for work!) Jul 28 at 21:52
2

I once went down this road and found that the AC unit outside consumed ~100 watts when not running. Never figured out why... I think we asked a technician who had no answer. I've heard they have heaters on the compressor to keep the oil warm, but 100 watts seems like a lot for that.

You might try flipping the breaker to the AC unit and see what happens. I'd be astounded if that accounts for 1 kwh though.

3
  • 1
    I was told about the AC heater from my tech, and have gotten in the habit of turning off power to the outside compressor equipment out-of-season. He told me that mice like having the heat! I don't know the power consumption.
    – DaveM
    Jul 26 at 18:22
  • Your AC unit likely has a crankcase heater which will ensure there is no buildup of liquid refrigerant in the compressor. More modern/sophisticated systems will switch the heater off if the ambient temperature is high enough, but it seems like your system did not. Jul 27 at 8:33
  • Interesting idea. I flipped the cutoff for it a few days ago, so I should see if it made a difference. Jul 28 at 21:54
2

Is the meter on the outside of the cabin or on a pole nearby. Many years ago I worked for a power company and a customer on a farm that had their meter on a pole that fed the house and a tree branch add grown up into the wire between the pole and the house. It shorted causing a large power draw and running up a huge bill for owner of the house. We trimmed the tree and their bill went down dramatically.

4
  • 1
    The mystery load is (1) always-on, and (2) correlated with cabin occupancy. This eliminates the possibility that it's something like a tree branch shorting things out.
    – Mark
    Jul 26 at 22:32
  • 1
    @Mark - Where did you read that the mystery load is correlated with cabin occupancy? I re-read the text twice and the way I understand it it's the exact opposite - there is load when nobody is there.
    – Vilx-
    Jul 27 at 10:33
  • 1
    @Vilx- In case this is the confusion -- note that the lower chart is on a different (longer) time scale versus the main chart. The small shaded area of the lower chart corresponds to the entire main chart. The red line for the recent occupancy begins at the same time as the increase in usage around June 12.
    – nanoman
    Jul 27 at 12:28
  • 1
    @nanoman - Ahh, finally I understood. Thanks!
    – Vilx-
    Jul 27 at 13:38
1

One option not mentioned here is getting a device often called "kill-a-watt". Essentially it's a mini-power-meter that plugs between an appliance and the wall socket. Thus you can measure exactly how much energy a specific appliance is using. Even things like boilers or water pumps often plug in with just a simple plug so you can use it on them too. After you've eliminated all possible suspects, the rest can be done with the circuit-breaker method mentioned above.

Hunting down the culprit will take some time, but depending on the type of central power meter you have, this tool might make it faster. For example, my power meter only shows whole kWh and not a load, so if I wanted to use the circuit-breaker method, I'd have to turn off a circuit and wait for probably a few hours to see if it made a difference.

Also keep an eye out for the possibility that a neighbour has decided to take advantage of your absence and steal some power (that is, if there are any neighbours close enough). Inspect your connections carefully at the meter and other possible places where a cable could come in. Obviously, this is not a very likely possibility, so don't start blaming your neighbours right away, but do check.

And another idea from a different angle - if you're gone for most of the year, why not just shut everything down, right at the central breakers? Sure, it might later take a bit more time to bring it back on, but that's just a minor inconvenience.

Correction: I misread the question. I thought that you were asking about unexplained load while you are away, not while you're staying.

1

Here's another perspective. Based on the lower (long-term) chart, it seems not merely that usage during occupancy is anomalously high but that usage during the most recent part of the vacancy (April to early June) is anomalously low.

Think back to your move-in day, June 12, and consider what you did and whether you observed anything unusual. Based on the main chart, the first big spike happened in the afternoon, and then the sustained higher usage you're concerned about began that evening.

During the current occupancy, has the heat been turned on at all? It sounds like you have no need for it and the only issue is keeping cool. Does the heat circulator/fan run at all in this case? If for any reason it is drawing power now, but wasn't when the cabin was empty in the spring, that could be a reason.

During the vacancy, in addition to a general trend toward lower unoccupied usage in the spring than in the winter (presumably due to less running of the heat circulator/fan), it appears there might have been a sudden event in early April that brought the usage down to the very low baseline you observed. (You could check the hourly usage there and determine how sudden it was.) This could be something that failed or was automatically switched off at that time, and then was (inadvertently?) restarted when you moved back into the cabin. (Even if it was just the heat going off for good once the spring warmth hit.)

In other words, not as many things may have been running when the cabin was unoccupied in the spring as you thought. This could reduce the gap between the usage jump observed and what you can account for.

In addition, something may have failed when you moved in on June 12 in a way that caused increased usage, as converting to occupied mode put sudden stress on some part of the home.

2
  • The "unoccupied" load includes heating. The cabin is described as being in Northern Virginia, and mid-April is a reasonable time for temperatures to warm up to the point where heating is no longer needed to keep the pipes from freezing.
    – Mark
    Jul 27 at 20:17
  • @Mark Yes, I'm saying that the heating system may have effectively shut down in April when no heat was needed for a long time at a low thermostat setting, and then after move-in in June the circulator/fan may have become active if the heat was used even briefly and intermittently. I don't know the normal duty cycle or if there was a malfunction, but am trying to clarify if the circulator/fan could be running continuously now, since the seasonal history does indicate that it seems to draw a lot when it is operating.
    – nanoman
    Jul 27 at 22:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.