How to Kill a Watt While Sizing a UPS

In my last two posts, I described some basic water conservation measures we’ve taken at the Poole Ancestral Homestead. We’ll return to that topic in the near future. But for now, I’ve turned my attention to optimizing our home’s use of electric power.

Utility-provided electricity is often an expensive form of energy. And it’s seldom efficient, because coal- and gas-fired plants (from which most of us get our electricity) actually waste a great deal of energy in the form of heat released during the generation process. Additional power is also required to transport generated electricity to the consumer. In fact, the so-called source efficiency of utility-provided electricity is only about 30%, on average. In other words, unless your electricity is generated on-site, it might require as much as 100 watts of power (of various kinds) to provide you with only about 30 watts of electrical power. So, it behooves most of us to reduce our daily demand for electricity, as much as possible.

Recently, a number of relatively inexpensive tools have emerged that can greatly help in measuring and reducing household use of utility-provided electricity. P3 International’s Kill A Watt™ EZ Power Meter is one such tool. Combining an electric power meter with a cost calculator, the Kill A Watt™ EZ can help you identify which of your appliances consume the most power, and determine how to reduce this demand. (Note that the Kill A Watt™ EZ is designed to work only with common household appliances in the 120V/15A range.)

The Kill A Watt(TM) EZ Power Meter

For example, you could use the Kill A Watt™ EZ to extrapolate the annual operating cost of an existing appliance, and then compare that against the published, estimated annual operating cost and purchase price of a newer replacement. Or, you could determine how to better manage an appliance that you don’t intend to replace; for example, by imposing certain daily limits on its use, so as to reduce its annual operating cost to some more tolerable level.

Using the Kill A Watt™ EZ requires first determining how much your utility company charges you for electricity, in dollars per kilowatt-hour (kWh). Usually, your billing rate can be found on your electric bill. For example, our electric company prints several separate rates on our bill: A generation rate (the cost of generating the electricity) of $0.106/kWh; a transmission rate (the cost of sending the electricity to a local substation) of $0.018/kWh; and a distribution rate (the cost of sending the electricity from the substation to our house) of $0.049/kWh. Also, there are several miscellaneous per-kWh charges and credits. When I add them all together, our total billing rate comes to $0.203/kWh. (We also have a fixed monthly connection charge of $15.85, but since it doesn’t vary by demand, I don’t include it in the billing rate.)

Next, you enter your billing rate into the Kill A Watt™ EZ. The meter needs to be plugged into a power source to do this, but it will retain the billing rate if you subsequently disconnect the power. I simply connected mine to a nearby outlet with an extension cord, then, using the SET button, entered and saved our billing rate of $0.203/kWh.

Setting the meter for our billing rate of $0.203/kWh

The Kill A Watt™ EZ is now ready to measure the power consumption of some appliance, and calculate its associated cost. The meter must be plugged into a power source, and the appliance plugged into the outlet on the front of the meter. With the appliance turned on, you press and momentarily hold the RESET button to clear any previous data and initiate a new power measurement process. The meter will then measure and accumulate the power consumption of the appliance, in units of kWh, for as long as the appliance is running.

After some reasonable amount of data collection time has elapsed, you can view the total quantity of power consumed by the appliance on the Kill A Watt™ EZ’s display. You can also view the equivalent, accumulated dollar cost, which is calculated using the entered billing rate. Furthermore, the Kill A Watt™ EZ can also estimate the total dollar cost of operating that appliance over an hour, day, week, month, or year.

My initial test of the Kill A Watt(TM) EZ using our cable modem ran for just over 24 hours...

Also note that, in general, the longer the Kill A Watt™ EZ is allowed to continuously monitor a particular appliance, the more accurate its measurements and estimates will become. This is because the accumulated, measured power consumption will have had a chance to settle to a longer-term, average value for the appliance. While this is especially true for appliances that regularly cycle on and off as part of their normal operation (for example, a refrigerator or small electric heater), it’s also the case for steady-state devices that might still experience variable loads over time (for example, a router/ gateway on your LAN).

...and measured an accumulated power of 0.14 kWh

As a first experiment with the Kill A Watt™ EZ, I decided to hook our cable modem up to it. Why the cable modem? Well, two reasons, actually. It’s the one plugged-in consumer appliance in the house (save for a few alarm clocks) that runs in a continuous, steady-state mode. So I felt it might be instructive to measure just how much power this device actually consumes over time.

It’s also the one plugged-in consumer appliance that I want to keep running continuously, because it provides our land-line telephony service, something that we’re not yet ready to dispense with in favor of a wireless-only solution. In the event of a power failure, I don’t want to lose the modem, so I’m planning to back it up with a UPS. But I’d like to accurately size the UPS to the measured demand of the modem, so the modem will at least keep operating for some reasonable amount of time following a (hopefully short-lived) power outage. So I need to have an accurate measurement of how much power the modem requires during its normal, steady-state operation (regardless of what’s stamped on the side of its adapter), and the Kill A Watt EZ seems like the ideal tool for measuring this.

...for a total cost of $0.02

With all that in mind, I plugged my newly configured Kill A Watt™ EZ meter into the power strip feeding the cable modem, plugged the cable modem’s adapter cord into the meter, and reset the meter to begin collecting data.

After 10 hours, the Kill A Watt™ EZ showed that the modem consumed 0.06 kWh of power, at a cost of $0.01. The Kill A Watt™ EZ also measures and displays a number of other power-related metrics beyond accumulated power and cost. I was able to determine, for example, that the current being drawn by the adapter was 0.10A, that the line voltage was 119V, and that the adapter’s apparent and real power demands were 12VA and 6W, respectively.

After 24 hours, the Kill A Watt™ EZ displayed a consumed power and cost of 0.14 kWh and $0.02, respectively. It also revealed that the adapter was now drawing 0.08A, and, as a result, showed an apparent power demand of 10VA. My guess is that either the measured value for amperage had settled on some steady-state average by this point, or the modem was just experiencing less of load now, than it had been previously. The adapter’s real power was still measuring at 6W, however.

My Kill A Watt(TM) EZ estimates that, for one year, our cable modem costs just over $10 to run

With regard to sizing a UPS to the modem, the particular UPS I’m considering is specified as supporting a full real load of 450W for about two minutes, and a 100W real load for about 36 minutes. I have no idea how to compute this curve (not yet, anyway; I’ll find out in time), but it seems like a safe bet that the 6W real power requirement of the adapter (as measured by the Kill A Watt™ EZ meter) is well within the capacity of this UPS to provide for some reasonable length of time (certainly more than 36 minutes, anyway). You can be sure, though, that once I actually configure the UPS, I’ll do a live test to find out exactly what that length of time is.

And one final note: Although devices like the Kill A Watt™ EZ power meter do a great job measuring and costing the electric power demands of individual appliances, they don’t readily provide a profile of the electric power demands of the entire house over time. A whole home energy monitor is required for this, and in a forthcoming post, I’ll discuss the specific tools I’m experimenting with in that regard.

About John Poole

My interests include historic homes, architectural preservation and restoration, improving the energy performance of old houses, and traditional timber frames.
This entry was posted in Energy Efficiency and tagged , , , , , , , , , , , . Bookmark the permalink.

4 Responses to How to Kill a Watt While Sizing a UPS

  1. Er, well, that all involved higher math so I got lost somewhere in the Kill A Watt as I started to think of the books I read in high school (it involved a mockingbird and Gregory Peck). But, I can comment intelligently on one aspect of this post – you are personally generating enough electricity to power all of CT. Maybe you should consider selling some of that! Buzzzzzzz

    • John Poole says:

      Well, good thing I left Maxwell’s equations out of this one…maybe for the next post! And glad to hear that I’m so “electric”, but are you sure I’m not just plain old shocking? :-)

  2. Joe Lopes says:

    I have Kill-o-watt devices for a long time and also find them especially useful for various electronics. Most devices have 2 or three stages. Stage 1 is “off”, which may also have residual or “vampire” loads, like DVR’s (13 watts). Stage 2 is “on”, where the device is turned on (DVR=25 watts). Then there may be specialized operations (for recording TV, 29 watts). During an extended period of time, there is a combination of those. The kill-o-watt gives you total kWh and total time and by dividing usage by time, you can get average usage per hour, then multiply by 8760 to get annual consumption, then by average cost to get annual $. No Maxwell’s equations required.

    • John Poole says:

      Hi Joe. I noticed how you referred to an “extended period of time”, and I similarly said “adequate amount of data collection time”. I found part of the art of doing this is making sure that you find how long it takes for a particular device to settle to a steady average state before making cost predictions. Which is sometimes why I’ll just throw one on and let it run for 24 or 48 hrs (less time for more predictable, non-cycling devices). Of course, that all assumes that Maxwellian demons aren’t just messing with me! :)