Welcome to the new Home Performance Forum.

This site is read-only for archival purposes. Please visit the Home Energy Pros Forum to continue the conversations.

Power factor and motor loads

I did an inspection today for a homeowner with high electric bills. He is gung-ho to start some monitoring, and had already researched Kill-A-Watt, TED, and similar meters. He has some motor loads, such as furnace blower, septic and well pump, water recirc pumps, etc. He asked about how accurate his results would be.

What I told him--and I hope it's true--is that if he used typical electricity monitors, he would get the same numbers for his motor loads that the utility is getting using their service meter, but that they would not be entirely accurate due to power factor. His cost figures should match theirs, but he wouldn't be getting actual watts consumed.

My understanding is that reactive loads can't be measured entirely accurately with typical inductance tools like the current clamps used with multimeters, Kill-A-Watt / TED type monitors, etc., and that an actual watt meter would be needed if one really wants to know exactly. A commercial or industrial customer with a lot of motors would care about this, but a typical homeowner would not.

How close am I?


  • Even though residential customers aren't billed for low power factor, the energy monitor must account for PF (e.g., it must measure watts) or else it's not going to match what the customer pays for.

    Unlike a multimeter, TED and Kill-A-Watt measure power, not VA. In fact, both report the power factor of the load being monitored. I'm not sure about the accuracy of Kill-A-Watt but since it's just for plug loads, it should be close enough. TED's accuracy (or any system that uses current transducers) should be within about 5%. If you have a reference, I believe there's a calibration setting in the Footprints software that can be used to adjust the output.

    I installed a highly accurate watt-hour meter adjacent to my main breaker panel. I have it set up so it can be connected in series with any circuit. I use it to calibrate my web-based energy monitor (Brultech ECM-1240), which uses CT's. One thing I've discovered is that the unit's voltage reference (required for measuring true power) is off by about 5%, thus I had to adjust it's calibration settings. Otherwise, the CT's are typically within 1%.

    Another interesting thing I noticed is that utility electro-mechanical watt-hour meters like the one I use as reference aren't accurate at very low loads. For example, the 35 watt crankcase heater on my heat pump isn't enough to make the wheel spin (lack of inertia). However, this is only an issue when measuring small isolated loads since in a whole-house scenario, the baseload will almost certainly remain much higher than that.
  • BTW, I had a client with high bills that discovered his septic pump was running 24/7, even though it had a timer and probably only needed to run an hour or so a day. HW recirc pumps typically don't consume much energy but they sure increase the DHW load! Even using a timer, say 2 or 3 hours a day, can dramatically increase heat loss from the pipes, even if they're insulated. My home came with a timer-based recirc loop. I installed one of those 5 minute wind-up timers. Problem solved.
  • Good thing I'm a carpenter and not an electrical engineer.

    I had assumed that a TED, since it uses current clamps, is a volt-amp meter like a multi-meter with a current clamp, but it sounds like you're saying this isn't the case, and I'm guessing that it's because the TED knows the voltage from inputs you give it via software.

    It also sounds like you're saying that utility meters (they are digital here) are measuring power factor so that the customer is billed correctly.

    Just looked at one of my Kill-A-Watts, and yes, it does indicate power factor.

    If the bottom line is that the utility meter and the various electrical monitors are measuring the same thing, that's good.

    I read in another discussion about issues with measuring the electrical usage of ductless heat pumps with various electrical monitors, but wasn't able to find that thread (it was on another forum).
  • A monitor can't know true power simply from user-input voltage. It must measure voltage in real time, simultaneously with current. The TED MTU's connect to the power bus. This enables them to measure the phase angle. TED also uses the power line as its communication medium between the MTU's and the gateway device (one of TED's serious flaws IMHO). The Brultech ECM-1240 uses a Zigbee radio IIRC.

    Analog watt-hour meters measure power directly, so they don't need to explicitly measure power factor. I'm less familiar with the inner workings of digital utility meters but many do report the power factor over the distribution side com link.

    I recall seeing a couple of low end energy monitors that just measure amps, and require the user to enter the voltage. This yields imputed VA, which is not very useful. Bottom line, you have to know what you're buying.

    I'd be interested if you can find the thread about issues monitoring ductless mini-splits, but I suspect it's a case of someone who simply didn't know what they were doing.
  • I echo Dave's comments. I've had several TED systems for several years. TED does monitor voltage and current so can calculate power factor. I have found TED to be quite accurate, or at least in tight agreement with utility meter, sometimes to within 0.5%. I have a newer electronic utility meter presumably not subject to zero drop out Dave describes.

    TED's PLC (Power Line Communication) is a PITA. Any switching power supply, one of which is aboard pretty much every plug-in electronic device sold today, will stop PLC dead. So will CFL and LED lights, from what I've read. Therefore a successful TED install typically requires two filters - one for the data going into the TED at the panel, and the other for the branch circuit used by the TED Gateway. I've had better luck with TED using filters, but it is annoying.

    I have several Kill-a-Watts, and often loan / give them to energy audit clients and assign them "home work" to monitor plug loads such as media gear and refrigerators.

    I believe the ductless thread was at GreenBuildingAdvisor. One issue that arose, but was later found to be overstated, IIRC, was standby power. I dimly recall a figure of 90 Watts being bandied about, which would really impair real world efficiency of a home with several minis.
  • Curt wrote: "..a successful TED install typically requires two filters - one for the data going into the TED at the panel, and the other for the branch circuit used by the TED Gateway."

    Seems like the filter would need to be on the offending load, no?

    BTW, my PV system (enphase) uses PLC to communicate between micro-inverters and the gateway device and its never given me problem one, although some enphase customers have reported issued in their forum. A good protocol can deal with noise produced by most switching power supplies, but there are some types of loads that make PLC com virtually impossible.
  • The filter only needs to be on the branch circuit that is feeding the TED readout unit, keeping the noise from that line. If other loads on that branch are causing the noise, then you need to also isolate them from the line, using a plug type filter. If it is hardwired, might be better to find a different branch to feed the readout. You are correct that the power line communications is a weak point, also being found with many x10 users. I've had about 50/50 success installing TEDs without filters, and for $15 for the starting unit not a bad way to get people interested in monitoring their power.
  • The filter in the panel is interposed between the panel and the power feed to the MTU(s). That provides the MTU(s) with clean power and voltage to measure. The Gateway is typically plugged into a receptacle convenient to a router for ethernet / internet connection. If the Gateway receptacle is on a branch circuit completely free of switching power supplies etc, then a plug in filter may not be needed. Of course, the router probably has a switching power supply...

    It is interesting that Enphase uses same PLC method without nearly as much difficulty.