← Guides·READY··5 min read

Size a backup power station with one math equation

Most people buy backup power based on horsepower marketing. The right question is watt-hours, and the equation is simple.

A backup power station is one of the highest-stakes purchases in home preparedness. Get it wrong and you have a $900 box that runs your fridge for four hours and dies in the middle of a multi-day outage. Get it right and you ride out a 72-hour grid event without thinking about it.

Most marketing for portable power stations leads with peak watts. That number does not tell you how long anything will run. The number that matters is watt-hours, and the math takes thirty seconds.

The one equation that runs everything

Watt-hours needed = (running watts of the appliance) × (hours you want to run it)

That is the entire formula. A 100-watt appliance running for 10 hours needs 1,000 watt-hours of stored energy. A 1,500 Wh power station can do that, with about 200 Wh left over for inverter inefficiency.

Memorize that equation and you can spec any system in your head.

What actually draws power in a blackout

Forget the air conditioner. Forget the electric oven. Those are not blackout loads. The real loads in a 72-hour outage are smaller and longer-running.

LoadRunning wattsHours per dayDaily watt-hours
Refrigerator (modern)1508 (compressor cycles)1,200
Modem + router2524600
Two LED lamps16696
Phone charging (two phones)15460
Laptop604240
Total per day~2,200 Wh

A modern fridge is the dominant load. It is also the load you most want to keep running, because the food in it is worth more than the power station. The compressor only runs maybe a third of the time, so 8 hours of compressor time per 24 hours is realistic.

Two thousand watt-hours per day is the practical baseline for a household running just essentials. Triple that for medical equipment, well pumps, or anything with heating elements.

How long do you need to last?

The next question is duration. The grid in most US suburbs goes down for a few hours a year. Once or twice a decade, it goes down for two to four days.

Set your target accordingly:

  • 24-hour buffer (~2,200 Wh): Covers most outages. A single mid-size station like our 1,500 Wh kit plus careful load management will get you through a one-day event.
  • 72-hour buffer (~6,600 Wh): Stacks two stations or one large 6,000+ Wh unit. This is the "named storm coming" tier.
  • Indefinite: You need solar input that matches or exceeds your daily draw. See the section below.

A 1,500 Wh station does not run your house for 72 hours by itself. Anyone who tells you otherwise is selling you something.

LFP vs NMC: pick LFP for backup duty

Power stations come in two battery chemistries:

  • LFP (Lithium Iron Phosphate): 3,000 to 6,000 cycles to 80% capacity. Tolerates partial charging. Stable in heat. Lasts 10+ years sitting at 50% charge.
  • NMC (Nickel Manganese Cobalt): 800 to 1,200 cycles. Lighter and energy-dense. Degrades faster when stored at high charge. Common in older or budget stations.

For backup power that sits 95% of the time waiting for an event, LFP is the only correct chemistry. The cycle count rarely matters in backup duty (you only deep-cycle it during actual outages), but the storage degradation does. A bargain NMC station at 50% charge in a hot garage will lose 20% of its capacity in three years. An LFP station in the same conditions loses maybe 5%.

Pay the LFP premium. It pays back over the life of the unit.

Solar input is for duration, not for daily use

If you want to run beyond what a single charge gives you, you need solar.

The math is the same: solar panel watts × hours of usable sun = watt-hours added per day.

A 200-watt folding panel in good sun gives you about 1,200 to 1,400 watt-hours on a clear August day. In December, half that. In rain, 10% of the rated output.

For a 2,200 Wh daily draw, you need at least 400 watts of panel capacity to keep up in summer. In winter, 800 watts. Most people undersize solar by a factor of two.

The right way to think about solar in backup mode: it is not the primary energy source. It is what extends your storage from 24 hours to 72 hours of meaningful operation. The station does the heavy lifting; the panel buys you extra runway.

Watt-hours are the spec; peak watts are the trigger

We have been talking about watt-hours (storage). The other number on every spec sheet is peak watts (output). Here is when peak watts matter:

  • Refrigerator startup surge. A 150-watt fridge draws 800 to 1,200 watts for the first second when the compressor kicks on. The station's continuous output rating must exceed that, or the inverter trips offline.
  • Anything with a motor. Sump pumps, well pumps, microwaves, power tools. All have surge loads two to five times higher than running watts.
  • Coffee maker, kettle, toaster. Resistive heating loads pull a constant 1,000 to 1,500 watts. Some smaller stations cannot handle them at all.

The rule: pick the station for your watt-hours requirement, then verify the peak output covers your highest single load. A 1,500 Wh station with a 1,200-watt continuous inverter will not run a 1,500-watt kettle. Read the spec sheet, not the marketing.

When to size up

Three situations push you to a larger station or a stack:

  1. Medical equipment. CPAPs, oxygen concentrators, refrigerated medication. These are non-negotiable loads. Size for the daily draw plus 50% margin.
  2. Well pump. Most residential well pumps are 1,000-watt to 2,000-watt motors with massive surge loads. A 1,500 Wh station with a 1,200-watt inverter will not start a typical well pump. Plan a separate dedicated supply.
  3. Multi-family or shared loads. If you are the household everyone in the neighborhood comes to during outages, your draw is not just yours.

What to do this weekend

Three steps to spec your system:

  1. Walk the house with a Kill A Watt meter. Note the running watts on every device you actually want to keep running. Add the daily hours.
  2. Multiply through. Get your daily watt-hours.
  3. Compare to a station's rated watt-hours. Subtract 15% for inverter inefficiency. That is your real runtime budget.

If you do not own a Kill A Watt, the load tables online for common appliances are close enough to make a decision. A 30-second search per device is the difference between a useful purchase and a $900 mistake.

If you want a unit with the math already worked out for the daily-essentials use case, our 1,500 Wh LFP station is sized for the table above with about 8 hours of margin. For multi-day runtime, plan to add solar input on top.

The goal is not to power your whole house. The goal is to keep the food cold and the comms up. Those are the loads that actually matter.

Other pillars

All guides →