Methodology

The base relationship converts power and time into energy, then energy into money. Duty cycle scales the running hours to reflect the share of that time the appliance actually draws rated power, and standby power is added for the remaining hours of the day.

One month means 30.44 days everywhere on this site — 365.2425 ÷ 12, the average length of a calendar month once leap years are included. A year is 365.2425 days, so twelve months add up to exactly one year and a daily, monthly and annual figure can never contradict each other. A 30-day month is deliberately not used: it understates every monthly estimate by about 1.5% and makes twelve months fall five days short of a year. The only exception is a tool that asks you for a specific number of days; that number is then used literally.

Nominal capacity is reduced twice before it reaches an AC load: once by the depth-of-discharge limit and any reserve you set, and once by inverter conversion efficiency.

Sizing runs the same relationship backwards, dividing the required AC energy by the same two factors. Power constraints are checked separately: continuous load against the inverter’s continuous rating, and motor start surge against its surge rating. Failing a power check produces a warning regardless of how large the battery is.

Charging cost is always based on energy drawn from the wall, since that is what the meter records. Time is based on energy reaching the pack at the sustained charging power. Real DC fast charging tapers steeply above roughly 80% state of charge, so times in that region are flagged rather than extrapolated.

Both systems are compared at identical delivered heat. Resistance heating is treated as COP 1.0. The heat pump figure should be a seasonal average (SCOP), and results are labelled a scenario estimate because that single input dominates the outcome. Equipment, installation and incentives are excluded entirely.

Specific yield comes from regional irradiance data rather than a global constant, and system losses default to the 14% used by NREL PVWatts, covering soiling, wiring, mismatch, inverter conversion and availability. Monthly distribution uses a seasonal shape for the region; it is indicative, and a site-specific simulation with real shading data will differ.

Rounding happens only when a number is displayed. Currency shows two decimals, or three when a rate per kWh needs them. Runtime is expressed in hours and minutes below a day and in days above it. Percentages are shown as whole numbers unless a fraction of a percent changes the decision.

Confidence is stated on every result: a deterministic calculation when the output follows directly from your inputs, a modelled estimate when a dataset stands between input and output, and a scenario estimate when a dominant assumption such as SCOP or duty cycle drives the answer. Where a real range exists, the range is shown rather than a false midpoint.