We audited the internet's most-copied wattage table. It's wrong in both directions.
Almost every solar blog, prepper site and generator affiliate page shows the same appliance wattage table. We know, because we started from it too. Then we built a sizing calculator that had to defend every number to a real purchase — and the table started falling apart.

How it started
WattSizer answers a simple question: what size battery station or generator do you actually need? To answer it honestly we needed running watts, startup surge and duty hours for every appliance someone might plug in. The famous table — the one that lists a coffee maker at 900–1,200W and an electric blanket at 100W — was the obvious seed. It's free, it's everywhere, and it looks authoritative in a table.
Our discipline required something the table didn't have: a source for every number. So we went and got them — U.S. DOE appliance data, the ENERGY STAR certified-product APIs (which cover thousands of current models and can be re-queried by anyone), and manufacturer documents: Goodman condenser spec sheets, Franklin Electric's submersible-motor tables, Dometic RV-AC install manuals, InSinkErator engineering specs.
Over one verification round in October 2026 we machine-checked every row we had inherited or collected — 130 appliances in the end. The folklore table lost its authority in the first afternoon, and not in the direction we expected.
Wrong in both directions — and both errors cost money
The table doesn't just overstate or understate. It does both, and each failure mode burns a different buyer:
Understated: the loads that trip your station offline
- Corded circular saws listed at 900W. Walk into any store: 7¼" saws are rated 13–15A — 1,560–1,800W at the plug, and measured startups around 2,400W. The audited class value is 1,500W with a 3× startup factor. Someone sizing a jobsite station off the old table buys one that trips the first time the blade meets wood.
- Refrigerator "surge" listed at 2–3× running. Household compressors are split-phase motors: nameplate reports and repair documentation put locked-rotor draw at 9–12× running — a 125W fridge starts at roughly 1,200W. The famous table doesn't carry a surge column for fridges at all, and the startup number is exactly what decides whether a small station survives.
- RV rooftop AC surge quoted for the wrong variant. Dometic's install manual lists the reference compressor at 68A locked-rotor, giving approximately 7,800VA at the reference voltage. This is apparent power, not a measured watt peak. Variant, current and startup duration must be checked before matching a station.
Overstated: the phantom loads that double your battery bill
- Gas clothes dryer at 1,800W. A gas dryer's motor, drum and igniter draw 250–400W measured — the 1,800W figure is the circuit rating, not the appliance. Somebody copied an electrical-engineering number into a power table and everyone followed. Size a battery for 1,800W×hours and you pay for capacity you will never use.
- LED bulbs frozen in 2012. The table says a 60W-equivalent LED draws 13W — true of first-generation products. The current ENERGY STAR certified median across 219 models is 9W. Multiply by a whole house of bulbs and a day of runtime, and the error is battery you bought for a ghost.
- Box fans at 200W. Lasko's own spec sheet says 65–90W for a 20" box fan; independent metering agrees. The copied value is more than double.
The full accounting — every correction we made, old value, audited value and the source behind it — is published as the audited appliance wattage chart. It stands at 0 corrections: 0 loads corrected upward (your station was going to trip), twelve corrected downward (you were going to overpay).
Why the best-known errors are the oldest ones
There's an instructive pattern in which numbers survived. The overstated values tend to be legacy technologies measured on legacy products — the 13W LED, the box fan, the gas dryer. Nobody updated them because copying is free and checking is expensive. The understated values hide a different sin: the table publishes nameplate running watts and stays silent on startup — the one number that actually separates a working backup plan from a tripped one. Silence reads as safety.
And the incentive loop is brutal: the table ranks in search, so it gets copied, so it ranks higher. Nothing in that loop ever opens a spec sheet.
What we did differently (and what it costs)
Every row in our library carries running watts, startup surge, duty hours, a source URL, and an evidence-tier label — manufacturer-published LRA, physics-checked compressor value, cross-checked class estimate, or ENERGY STAR aggregate. Aggregates re-compute from the official datasets automatically every week, so certified values can't silently rot the way the folklore table did. Corrections get applied with the evidence recorded next to them.
This is slower than copying. It is also the entire product: if you're spending $1,500 on a battery station because a table said so, you deserve to know which table, which number, and which spec sheet.
If you publish a wattage table
Then this story is for you. Check the six figures above against your own copy — most tables we tested carry at least four of them. And if you want values you can defend, take ours: the audited chart is free to cite with attribution, every number links to its source, and the methodology page documents exactly how each tier is verified. Or run your own audit against the ENERGY STAR APIs and the spec sheets we list — we'd love to be proven wrong on any row. That's how the table gets fixed.
Putting the numbers to work? The chart is the ingredient list — the calculator is the recipe, and the soft-start screening tool handles the nastiest startup spikes.