Record energy and elapsed time together
A cumulative kWh value is meaningful only when its start and end times are known. Reset the compatible meter, note normal appliance settings, and record the complete measurement window. Instantaneous watts can describe a stable load, but cumulative energy is stronger for compressors, heating elements, variable-speed motors, and programme-based appliances.
Normalize the observation
Suppose a refrigerator uses 8.4 kWh over seven ordinary days. Daily energy is 8.4 ÷ 7 = 1.2 kWh. A planning month based on 365 ÷ 12 days is about 36.5 kWh, while the simple annual projection is 438 kWh.
Subtract cumulative readings before calculating cost
If a meter cannot be reset, record both readings. An illustrative start of 124.6 kWh and end of 133.0 kWh means the appliance used 8.4 kWh during the window. The end reading alone includes earlier use. If the reading decreases, check for a reset or rollover before treating the difference as consumption.
For exactly seven days at an illustrative $0.20/kWh, that window costs $1.68. The 1.2 kWh daily average projects to $7.20 for 30 days or $7.44 for 31 days. Those are projections from the same observation, not two different measured monthly bills.
Work through a time-of-use example
Suppose the same 8.4 kWh is recorded as 3.0 kWh in a peak band at $0.30/kWh and 5.4 kWh off-peak at $0.12/kWh. Cost is 3.0 × $0.30 + 5.4 × $0.12 = $1.548, displayed as $1.55. The simple average of the two rates would incorrectly give $1.76 because the energy was not split equally.
If only total kWh is available, report the limitation. In this example the energy cost falls between $1.008 and $2.52 under those two rates; a precise blended cost requires the time allocation. These rates are fictional, not current tariff quotes.
Match the window to the appliance
One full cycle can be informative for a kettle, dishwasher, washer, or dryer. Seven or more days is stronger for refrigerators, dehumidifiers, and household standby load. Heating and cooling need representative weather; one mild day should not be projected across an entire hot or cold season.
Use the rate that applied during the reading
For a flat tariff, multiply by the per-kWh energy charge. For time-of-use service, a single blended meter reading may conceal when the energy was consumed. Split peak and off-peak energy when the meter or charger provides timestamps. Fixed daily charges should remain outside an appliance-level marginal-cost estimate.
Keep context with the number
Record programme, thermostat setting, room conditions, occupancy, load size, season, and unusual events. A measurement without context can be accurate for its window yet misleading for another household or season. PowerCost Lab treats the observation and its context as one evidence record.
Verify a saving with before-and-after periods
Measure a baseline, make one change, and measure again under comparable conditions. Cleaning a filter, changing a schedule, adjusting a set point, or replacing equipment can then be evaluated using observed energy rather than a claimed percentage. Weather-sensitive comparisons should use similar conditions or a longer window.
Electrical and privacy safety
Use only equipment rated for the plug, voltage, current, and load. Do not place ordinary plug meters on hard-wired or unsuitable high-power equipment. If you save a meter image, exclude names, addresses, account numbers, faces, and reflections. PowerCost Lab does not need those details to learn from an energy reading.
Source register
Sources used for this guide
Sources establish definitions and market context. The explanations, calculations, examples, and limitations on this page are PowerCost Lab's original analysis.
- PowerCost Lab measurement protocol ↗The site protocol for representative measurement windows, context, evidence, and privacy.
- PowerCost Lab methodology ↗Definitions for estimated, label-derived, measured, and outcome-verified figures.
Who · how · why
How this page was prepared
PowerCost Lab publishes this page to help readers make a specific electricity-cost decision. Worked figures are recalculated from the inputs shown, assumptions remain visible, and market facts are separated from user-entered estimates.
Run your own scenario