How we calculate

No black box. These are the exact assumptions the calculator uses, and what it doesn't cover yet.

Battery capacity

For each device: running watts × share of time it actually runs × hours per day × quantity. A fridge compressor, for example, runs about 35 % of the time.

We check these figures against published energy data where it exists. Our full-size fridge (150 W × 35 % × 24 h) comes to 460 kWh a year. The median of 3,033 ENERGY STAR certified full-size models is 457 kWh. If one of our assumptions drifts more than 15 % from its reference, our build stops.

AC devices go through the station's inverter. We count two losses: it converts 90 % of the battery energy into AC power, and it uses about 15 W for itself whenever the AC output is on, whatever you plug in. Manufacturers state anything from a few watts to about 30 W; when a station's own figure is published, we use that. Together this gives 89 % overall at 1,000 W, 79 % at 100 W and only 47 % for a 15 W router. Small devices running all day on AC are expensive.

Devices on the 12 V or USB ports skip the inverter and its idle draw. Their converters lose less: we assume 90 % of the energy taken from the battery reaches the device.

Stations don't let you drain every last watt-hour. Unless the manufacturer says otherwise, we count 95 % of the rated capacity as usable.

Not modelled: capacity loss in the cold and as the battery ages, and the fact that a fridge runs longer in a hot room. If that applies to you, add a day of reserve.

Output and surge

Continuous: the running watts of all AC devices added up, as if everything is on at once.

Surge: motors and compressors draw far more than their running power for a split second when they start. In our device list the start-up figure ranges from 1.2× to 10× the running watts. We add the largest single start-up spike on top of everything else that is running. That assumes two motors don't start in the same instant. When they could, for example two fridges when the power comes back, we show the all-at-once peak as a warning.

Two start-up figures for fridges. Published start-up values disagree a lot. A measured frost-free fridge peaked at about 13 A for half a second, roughly 1,500 W. Generator sizing guides assume up to 2,900 W for a fridge, which fits older units. We use the first figure for the check and the second as a warning: if a station only starts the typical fridge, it is marked tight. If you know your fridge's start-up watts, enter them and only your figure counts.

We only compare against a station's published surge or peak rating. Boost modes that let heaters and kettles run above the rating at reduced voltage don't count, because they don't help a motor start.

A check is marked tight (close to the limit) when it uses more than 90 % of the station's rating. Devices above 120 W on the 12 V car port get a warning: most car ports stop there.

Solar refill

We use monthly average solar yield from PVGIS (Photovoltaic Geographical Information System, European Commission Joint Research Centre). We calculated it in advance for points roughly 100 km apart and blend the nearest points for your location. The figure is kWh per kW of panels per day, also called "peak-sun hours".

PVGIS assumes a fixed, optimally tilted panel with 14 % system losses. Portable panels are moved around and rarely aimed perfectly, so we apply a further 85 % factor, and 92 % charging efficiency. The 14 % already includes some conversion losses, so this is on the cautious side. Panel power above the station's maximum solar input is ignored.

"Sun covers" compares an average day of sun with your daily use. "Empty to full" is the time to refill a station's whole usable capacity with nothing plugged in.

These are long-term monthly averages. A cloudy week can be far worse. And in frost most LiFePO4 stations don't charge at all: the ones we've checked charge only between 0 °C and 40–45 °C (32 °F to 104–113 °F).

How close is the grid? We compared it with exact PVGIS results for six cities (Denver, Seattle, London, Munich, Sydney, Perth) in November 2026. Over the year our figures were within 4 %. For a single month the largest gap was 9 %, in Seattle, where coast and mountains change the sun within a short distance.

Not modelled: most stations charge more slowly over the last 10 to 20 % of the battery, so a refill to 100 % takes somewhat longer than shown. Check that your panel's voltage fits the station's solar input range.

Solar data: PVGIS © European Union, Joint Research Centre (JRC), 2001–2026. Monthly averages for an optimally tilted fixed panel, 14 % system loss.

Station specifications

Every capacity, output and surge figure comes from the manufacturer's published specifications, stored with its source, the exact wording and the date we read it. A model only appears once every figure is sourced and passes automatic checks for impossible values, for example a surge rating below the continuous rating. When a manufacturer's page contradicts itself, we use the lower figure.

We recommend the variant sold in your market: 120 V models on the English site, 230 V models on the German site.

How this site is made

Code, calculations and copy are written with the help of AI tools. Every number comes either from the calculation core, which automated tests check against hand-worked values, or from a stored source. Manufacturer figures are never estimated by an AI: if a spec is missing, the model stays hidden.

The operator named in the legal notice is responsible for the content. If you spot a mistake, email us. We log corrections with a date.