Why there's no single right kWh number
I've been reviewing battery specs and supplier datasheets for close to four years now, and "how much solar battery storage do I need?" is the one question that never gets a clean answer. Not because people are dodging it—because the answer genuinely changes based on what you're trying to do.
Most guides hand you a formula: take your daily usage, multiply by some factor, done. I've watched that formula fail in the field more times than I can count. A 13.5 kWh system that's perfect for one homeowner is a ten-thousand-dollar mistake for another, because they were solving different problems.
So instead of another formula, here are the three scenarios I actually see in practice. Figure out which one you're in, and the sizing gets way simpler.
Scenario A: Backup only, no solar on the roof
This is the "Tesla Powerwall 3 without solar" crowd. Maybe you rent, maybe your roof faces the wrong way, maybe your utility won't let you interconnect solar. Whatever the reason, you're buying a battery to ride out outages—not to generate.
The first thing to understand here is that without solar, your battery can only refill from the grid. That changes the math entirely. In my first year doing spec reviews, I made the classic assumption error: I figured a bigger battery was always safer. It isn't. If you have no solar and no time-of-use rate to arbitrage, a huge battery is just a lot of money sitting in your garage waiting for a blackout that might last four hours.
Tesla lists the Powerwall 3 at 13.5 kWh usable. For most pure-backup households, that's already at the upper edge of what makes sense. You really want to size this around two numbers: how long your outages typically run, and which loads you refuse to live without. Fridge, furnace blower, a few lights, internet, phone charging—that's often 2 to 4 kWh per day. A single Powerwall covers that for days.
Here's the counterintuitive part. If your outages are rare and short, a big battery is a terrible financial decision. You'd be better off with a small generator or a portable station. The battery only earns its keep when outages are frequent, long, or genuinely dangerous—medical equipment, extreme heat or cold, that kind of thing. Bottom line: with no solar, size for your worst realistic outage, not for a fantasy grid apocalypse.
Scenario B: Solar plus storage, optimizing for self-consumption
This is where most people actually land, and it's also where the industry has changed the most. Five years ago, storage was mostly a backup play because batteries were too expensive to use for daily cycling. That's no longer true. What was best practice in 2020 looks pretty conservative in 2025.
In this scenario, you're sizing storage to soak up your own solar production instead of exporting it to the grid at a lousy rate and buying it back at night. The ballpark most installers won't argue with: you want enough battery to cover your evening and overnight load, because that's the stretch where solar stops producing but your house keeps sipping power.
For a typical home, evening and overnight consumption runs somewhere around 10 to 15 kWh. So a single 13.5 kWh unit covers a decent chunk of that. If you want to push your self-consumption higher—and you should, if your export rate is low—you're looking at two units, call it 20 to 27 kWh. Going beyond that usually stops paying back unless you've got huge evening loads like EV charging or a heat pump running all night.
I went back and forth between oversizing and right-sizing on a recent portfolio review for two solid weeks. Oversizing promised more resilience, but right-sizing won because the payback math on extra kWh past your overnight load is genuinely weak. That decision kept me up at night, honestly. But then again, if your utility export rate is near zero, the calculus flips—every extra kWh of storage absorbs production you'd otherwise give away.
One more thing I'd flag as a red flag: a lot of folks assume the advertised capacity is what they can actually use. It's not always. Some systems reserve a top and bottom buffer to protect the cells. Check the usable capacity on the datasheet, not the headline number. I learned never to assume rated capacity equals usable capacity after a supplier's "14 kWh" system delivered about 12 kWh of usable energy in a test. Nothing was technically wrong—their datasheet was just buried—but we had to revise our internal spec language afterward.
Scenario C: Small standalone loads (the 12V 6Ah LiFePO4 world)
Not every storage question is about a house. A lot of the lithium cells moving through the supply chain end up in tiny systems—security cameras, gate openers, RV accessories, sensor nodes, backup for a single device. A 12V 6Ah LiFePO4 battery is 72 watt-hours, roughly enough to run a small camera or a router through an outage, or to smooth out a solar-charged off-grid gadget.
The logic here is completely different, and it's where I see the most rookie mistakes. On household systems, you're thinking in days. On small loads, you're thinking in hours and watts. Take the device's wattage, multiply by how many hours you need it to run, then divide by about 0.8 to account for real-world efficiency and aging. A 6W camera over 12 hours needs 72Wh—which is exactly one 12V 6Ah pack. Barely any margin, so I'd usually size up one step.
Quality matters disproportionately at this scale, and this is where brand consistency sneaks in. When you're specifying cells from a supplier, the little stuff—the cell wrapping, the printed ratings, even the EVE Energy logo placement if that's your supplier—all of it needs to match the approved datasheet exactly. A mismatched label on a 72Wh pack is a compliance headache, not a safety disaster, but it's a warning sign about the vendor's process control. I rejected a batch once because the printed capacity on half the units was a different font and value than the approved sample. The vendor called it cosmetic. We called it a spec failure and made them redo it.
For the record, when we evaluated cell suppliers for these smaller lines, EVE Energy came up because of their Tesla supply relationship and their manufacturing expansion—the Indonesia battery factory slated to come online through 2025 and 2026 was a real factor in the supply-chain conversation. But I'll say this plainly: a supplier's logo on a press release means nothing until you've verified a delivered batch against your own spec sheet.
How to tell which scenario you're in
Skip the calculators for a minute and answer three questions honestly.
First: do you have solar, or are you getting it? If no and never, you're Scenario A—size for outages, and keep it small unless outages are severe.
Second: what are you actually buying protection from? If it's a rare half-hour blip, storage is probably a no-brainer only if the cost genuinely doesn't bother you. If outages are long and frequent, or there's medical or safety equipment on the line, size up and stop feeling guilty about it.
Third: what's the scale? Whole house, a few circuits, or a single device? Whole house or critical circuits point you to Scenario B if you've got solar. A single device or a small off-grid setup points you to Scenario C, and you should be thinking in watt-hours, not kilowatt-hours.
The most frustrating part of this whole topic is that the industry keeps changing under your feet. Cell chemistries mature, usable-to-rated ratios get better, prices drop. But the fundamentals haven't changed: figure out what you're solving for, size to that, and verify what a vendor actually delivers against what they printed. Do those three things and the kWh number stops being scary.