Methodology

How Willrun decides

Portable power station connected to a cooler at a campsite

Willrun's runtime is nameplate watt-hours times 0.85, divided by average draw, and a compressor, pump, or tool stays a surge risk until startup watts are published.

The inverter allowance lives in one place, src/site.ts, as 0.85. Changing it changes every page on the next build. Affiliate URLs live in src/data/links.json and are empty in this version.

What a verdict means

Runs means the published running watts fit the station's continuous AC output, and any published appliance peak also fits the surge figure Willrun is willing to treat as a motor start.

Runs, surge risk means the running watts fit, but startup watts were not published for a compressor, pump, or tool, or a published peak is above the motor-surge limit. A fan is marked surge risk only when its running watts are more than half the continuous output.

Won't run means the running watts are above continuous output and the load is not a simple heater that fits a documented resistive boost. Runtime is left blank in that case.

No run verdict means running watts were not published. An hour figure on those rows is only an energy budget from a published annual average. It is not a claim that the inverter can start the appliance.

Runtime

Usable watt-hours are nameplate capacity times 0.85. That 0.85 is a Willrun allowance for inverter loss. It is not a measured efficiency for any model, and there is no extra idle-draw term.

Runtime is usable watt-hours divided by average draw. Average draw is the stored average when one exists, otherwise running watts times a duty cycle, otherwise the running watts themselves.

A combined calculator load uses the worst individual verdict. If the sum of running watts is above continuous output, and the loads are not all simple heaters inside a resistive boost, the mix will not run. Combined runtime uses the sum of average draws.

Duty cycle

A fridge duty cycle is stored only when the maker published both a wattage and an annual kWh figure. Average watts are annual kWh times 1,000 divided by 8,760. Duty cycle is that average divided by the published wattage. The label on those pages calls it an annual average, not a hot-day measurement.

Window air conditioners and dehumidifiers are not given a duty cycle. Their watt figures are the draw at the rated test condition. A sump pump has no duty cycle because how often the float runs depends on the inflow, which was not published.

Derived watts

Window-unit running watts are Btu/h divided by the published CEER. That result matches the ENERGY STAR annual kWh figure at the DOE room-air-conditioner assumption of 750 hours, so it is used as rated cooling input. CEER is not a pure EER, and startup watts are still unpublished, so the verdict stays a surge risk.

Dehumidifier running watts come from pints per day times 0.473176 liters per pint, divided by 24 hours, divided by the integrated energy factor, times 1,000. That is the draw while the unit is removing water at the rated rate, not a whole-year average.

The Zoeller Model 53 running figure is 9.7 A times 115 V. The product page lists 4.8–9.7 A because the model is sold for 115 V and 230 V. The 9.7 A row is the 115 V automatic model in Zoeller's 50-series table.

Surge, X-Boost, Power Lifting, and SurgePad

A published surge is used for motors only when the maker states it separately from a heater-only mode. X-Boost, Power Lifting, and SurgePad are stored as a resistive boost. They apply only when the appliance is a simple resistive heater and the element is between continuous output and that boost rating. Voltage can sag in that mode.

Microwaves and induction cooktops are electronics. They are not boost-eligible. A fan, a fridge, a window air conditioner, a dehumidifier, and a pump are not boost-eligible either.

EcoFlow's RIVER 2, RIVER 2 Max, RIVER 2 Pro, and RIVER 3 Plus publish a surge figure that is the same as X-Boost and is described for heating devices, so it is not used as motor-start capacity. The DELTA 2's 2,700 W surge appears on a 230 V manual line and is not used for the 120 V model. BLUETTI's AC180 lists a 2,700 W surge that the FAQ treats as Power Lifting and tells owners not to use for an air conditioner or a washer, so it is not used for motors. The AC70's 1,500 W surge is kept for motors; its 2,000 W lifting mode is resistive only. Anker's C300 headline surge matches the FAQ's SurgePad description, so it is a boost. The Goal Zero Yeti 1000X page disagrees with itself on surge (3,000 W and 3,500 W), so surge is omitted.

What was left out

Specs that could not be read on a manufacturer page or an ENERGY STAR record were left blank or the model was dropped. Launch prices were not on the pages used, so every launch price is empty. Current shelf prices were not copied in.

Official pages for several smaller or newer stations did not resolve to a usable spec table, including Jackery Explorer 300 Plus, 600 Plus, and 500 v2, a Jackery 3000 Pro URL that opened a different product, and BLUETTI EB3A and Elite 200 V2 redirects. Those models are not in the set. The Anker F2000 marketing line of 3,600 W was not in the spec table, so it is not stored. A garbled DC voltage line on the DELTA Pro 3 page was omitted.

Common loads without a reputable running-watt source were omitted rather than filled from a generic watt chart: drip coffee makers, well pumps, corded power tools, oxygen concentrators, Starlink, garage-door openers, pellet stoves, box fans, kettles, toasters, slow cookers, rice cookers, hair dryers, air fryers, pressure cookers, electric blankets, and RV roof air conditioners. Full-size refrigerators and chest freezers in this set have ENERGY STAR annual kWh only, so they get no run verdict.

Laptop rows are an adapter maximum or an ENERGY STAR short-idle figure, not a measured full workload. Router watts are the TP-Link Archer AX73 networked-standby figure, not a marketing '3 W router' claim.

Generators: running vs starting watts

A generator lists two numbers. Running (rated) watts is what it can hold continuously. Starting (surge or peak) watts is the brief burst it can deliver for a second or two while a motor or compressor spins up. Willrun says a generator runs an appliance when the appliance's running draw fits within running watts and its startup draw fits within starting watts. When an appliance maker does not publish startup watts for a compressor or pump, Willrun only clears it when the generator's starting watts are at least 3× the running draw, a common planning allowance, and says so on the page. Dual-fuel models lose some output on propane; Willrun compares the gasoline rating and lists propane separately.

Open-frame generators can have high total harmonic distortion. For CPAPs and sensitive electronics Willrun prefers inverter generators, which publish under 3% THD.

Whole house via a 30 amp inlet

A 30 A, 120/240 V power inlet (L14-30) can carry at most 30 × 240 = 7,200 W, split into two 120 V legs of 3,600 W. Only generators with a 120/240 V 30 A outlet (L14-30R or equivalent) can feed both legs. A 120 V L5-30R tops out at 3,600 W and cannot run 240 V loads. A TT-30R is an RV outlet. Willrun's essential-circuit list totals 4,440 W running and 6,790 W at the worst start, using typical chart values.

Generator safety

Questions

What a verdict means

Runs means the published running watts fit the station's continuous AC output, and any published appliance peak also fits the surge figure Willrun is willing to treat as a motor start. Runs, surge risk means the running watts fit, but startup watts were not published for a compressor, pump, or tool, or a published peak is above the motor-surge limit. A fan is marked surge risk only when its running watts are more than half the continuous output. Won't run means the running watts are above continuous output and the load is not a simple heater that fits a documented resistive boost. Runtime is left blank in that case. No run verdict means running watts were not published. An hour figure on those rows is only an energy budget from a published annual average. It is not a claim that the inverter can start the appliance.

Runtime

Usable watt-hours are nameplate capacity times 0.85. That 0.85 is a Willrun allowance for inverter loss. It is not a measured efficiency for any model, and there is no extra idle-draw term. Runtime is usable watt-hours divided by average draw. Average draw is the stored average when one exists, otherwise running watts times a duty cycle, otherwise the running watts themselves. A combined calculator load uses the worst individual verdict. If the sum of running watts is above continuous output, and the loads are not all simple heaters inside a resistive boost, the mix will not run. Combined runtime uses the sum of average draws.

Duty cycle

A fridge duty cycle is stored only when the maker published both a wattage and an annual kWh figure. Average watts are annual kWh times 1,000 divided by 8,760. Duty cycle is that average divided by the published wattage. The label on those pages calls it an annual average, not a hot-day measurement. Window air conditioners and dehumidifiers are not given a duty cycle. Their watt figures are the draw at the rated test condition. A sump pump has no duty cycle because how often the float runs depends on the inflow, which was not published.

Derived watts

Window-unit running watts are Btu/h divided by the published CEER. That result matches the ENERGY STAR annual kWh figure at the DOE room-air-conditioner assumption of 750 hours, so it is used as rated cooling input. CEER is not a pure EER, and startup watts are still unpublished, so the verdict stays a surge risk. Dehumidifier running watts come from pints per day times 0.473176 liters per pint, divided by 24 hours, divided by the integrated energy factor, times 1,000. That is the draw while the unit is removing water at the rated rate, not a whole-year average. The Zoeller Model 53 running figure is 9.7 A times 115 V. The product page lists 4.8–9.7 A because the model is sold for 115 V and 230 V. The 9.7 A row is the 115 V automatic model in Zoeller's 50-series table.

Surge, X-Boost, Power Lifting, and SurgePad

A published surge is used for motors only when the maker states it separately from a heater-only mode. X-Boost, Power Lifting, and SurgePad are stored as a resistive boost. They apply only when the appliance is a simple resistive heater and the element is between continuous output and that boost rating. Voltage can sag in that mode. Microwaves and induction cooktops are electronics. They are not boost-eligible. A fan, a fridge, a window air conditioner, a dehumidifier, and a pump are not boost-eligible either. EcoFlow's RIVER 2, RIVER 2 Max, RIVER 2 Pro, and RIVER 3 Plus publish a surge figure that is the same as X-Boost and is described for heating devices, so it is not used as motor-start capacity. The DELTA 2's 2,700 W surge appears on a 230 V manual line and is not used for the 120 V model. BLUETTI's AC180 lists a 2,700 W surge that the FAQ treats as Power Lifting and tells owners not to use for an air conditioner or a washer, so it is not used for motors. The AC70's 1,500 W surge is kept for motors; its 2,000 W lifting mode is resistive only. Anker's C300 headline surge matches the FAQ's SurgePad description, so it is a boost. The Goal Zero Yeti 1000X page disagrees with itself on surge (3,000 W and 3,500 W), so surge is omitted.

What was left out

Specs that could not be read on a manufacturer page or an ENERGY STAR record were left blank or the model was dropped. Launch prices were not on the pages used, so every launch price is empty. Current shelf prices were not copied in. Official pages for several smaller or newer stations did not resolve to a usable spec table, including Jackery Explorer 300 Plus, 600 Plus, and 500 v2, a Jackery 3000 Pro URL that opened a different product, and BLUETTI EB3A and Elite 200 V2 redirects. Those models are not in the set. The Anker F2000 marketing line of 3,600 W was not in the spec table, so it is not stored. A garbled DC voltage line on the DELTA Pro 3 page was omitted. Common loads without a reputable running-watt source were omitted rather than filled from a generic watt chart: drip coffee makers, well pumps, corded power tools, oxygen concentrators, Starlink, garage-door openers, pellet stoves, box fans, kettles, toasters, slow cookers, rice cookers, hair dryers, air fryers, pressure cookers, electric blankets, and RV roof air conditioners. Full-size refrigerators and chest freezers in this set have ENERGY STAR annual kWh only, so they get no run verdict. Laptop rows are an adapter maximum or an ENERGY STAR short-idle figure, not a measured full workload. Router watts are the TP-Link Archer AX73 networked-standby figure, not a marketing '3 W router' claim.