First with the device in question turned on, then with it off, you time how long it takes the wheel on your electric meter to make one revolution. (If you have a digital meter, how long it takes the dashes to return to the same spot). Then using the simple equations below, calculate how much power the device is using when on. (This works because the meter records energy use and power is energy per second.) Second, for devices which cycle on and off automatically, determine percentage of on-time on by sampling.
If you already know how much energy is indicated by one turn of the wheel on your electric meter, skip step 4. The number is often printed inconspicuously on the face of the meter (Kh=3, for example, means 3 watt-hours/revolution). Otherwise...
You can now calculate Kh: the number of Watt-Hours per revolution on your meter:
Kh = ( Wk * T3 * T4 ) / (3600 * (T3 - T4))
W2 = ( 3600 * Kh ) / T2 W3 = ( 3600 * Kh ) / T3The device you wanted to measure in step 2 is consuming (W2 - W3) watts or:
DeviceWatts = (W2 - W3) = Kh * 3600 * (T3 - T2) / (T3 * T2)
If your baseline watts in step 3 (W3) seems higher than you expected, you may want to investigate further. Some devices consume a surprising amount of power even when turned off. You may be surprised how many you have and what their cumulative impact on your power consumption is.
Since there are 8760 hours in a year, each watt consumed continuously for a year contributes 8.76 KWh (kilowatt-hours or $0.60-$1.20) to your annual bill (at $.07-$.14 per KWh).
Many TVs, VCRs, radios, telephones, answer machines, calculators, etc. consume 3-10 watts continuously. You might be able to save $100 worth of electricity per year by unplugging devices you use infrequently and thought were already off!
To measure the small 3-10 watt devices using the technique above, you may need to get your baseline power consumption (W3) to about 100 watts (possibly by temporarily switching off more devices and/or circuit breakers). You can expect to accurately measure power differences (W2-W3) of as little as 5% of baseline (W3).
Appliances such as refrigerators and freezers turn on and off automatically. It takes special equipment to record exactly when they turn on and off, however you can get a good approximation by simply sampling (observing) whether the device is on or off at a few random times and averaging.
The key to having this work well is having enough observations and having your observation times be random (or at least not highly correlated with one state--for example, not after dinner when the door has been open a lot and not all in one hour). Good times would probably be when you first wake up; when you first come home; when you go to bed. The more observations you make, the more accurate the estimate. Two or three per day for a week should get you within 10%.
So, assuming you take 21 observations (3 per day for a week), just divide the number of times you found the appliance was on by the number of observations you made, to get the fraction of on-time (or duty-cycle). To determine the average power consumption of this appliance, multiply the power it uses when on by the fraction of time it is on.
OnTime = Times Observed On / Number of Observations AverageDeviceWatts = DeviceWatts * OnTime
Multiply the Average Device Watts by 8760 hours per year and divide by 1000, to compute the number of KWh/year your appliance is using. You can compare that to energy-guide numbers on a new one.