How Many Batteries to Power a House: The Complete Sizing Guide

A customer called me last winter, frustrated. He had bought four 12V 100Ah deep-cycle lead-acid batteries, wired them for 48V, and connected them to a 5,000W inverter. His plan: power the furnace, fridge, lights, and router through a 2-day outage. The math on paper said he had 4.8 kWh of usable capacity, and his daily critical load was around 4.5 kWh. He should have had just enough to scrape through 24 hours.

He lasted four hours.

The mistake was not in his arithmetic. It was in what his arithmetic was measuring. He sized for the whole-house load from his utility bill — a number that included a 4-ton central air conditioner, an electric water heater, an EV charger, and a clothes dryer he forgot about. The batteries drained before his refrigerator’s first defrost cycle finished.

This is the sizing mistake I see over and over. People ask “how many batteries to power a house” as if a house is a single appliance with a fixed wattage label. It isn’t. A house is a collection of circuits, and most of them do not need to run during an outage.

Here is how to get the number right.

Step 1: Stop Thinking “Whole House” — Think “Critical Loads”

Your main electrical panel is probably 200A at 240V — that is 48,000 watts of theoretical capacity. In a real outage, you do not need 48,000 watts. You need maybe 1,500-3,000 watts continuous, and only on the circuits that matter.

The first decision when sizing a battery system is which circuits get backed up. This is not a shopping list. It is a triage list. Every circuit you add costs roughly $500-1,000 in additional battery capacity.

Here is what belongs on the list, in priority order:

Tier 1 — Non-negotiable:

• Refrigerator: 1.0-1.5 kWh/day (modern Energy Star; older units can be 2-3 kWh)
• A few LED lights in key rooms: 0.3-0.5 kWh/day
• Internet router + modem: 0.2-0.3 kWh/day
• Phone/laptop charging: 0.2-0.3 kWh/day

Tier 2 — Climate and safety:

• Gas furnace blower: 2-4 kWh/day (winter only)
• Sump pump: 0.5-2 kWh/day (rain-dependent, but when it runs, it runs)
• Chest freezer: 0.8-1.2 kWh/day
• Well pump: 1-3 kWh/day (high surge on startup — 3-5x running watts)

Tier 3 — Comfort:

• Window AC unit (small): 3-6 kWh/day
• Microwave: ~0.1 kWh per use (high power, short duration)
• TV/entertainment: 0.5-1 kWh/day
• Garage door opener: negligible for sizing (seconds of use)

Tier 4 — Do not back up:

• Central air conditioning (3-6 kW running, massive startup surge)
• Electric water heater (4.5 kW resistive load — battery killer)
• Electric dryer (5 kW resistive load)
• EV charger (7-11 kW)
• Electric stove/oven

A typical Tier 1+2 critical load total lands between 5 and 10 kWh per day. That is the number you design around — not your utility bill’s 30-50 kWh daily average.

Step 2: Calculate Daily kWh — Accurately

There are two ways to get this number. One is free and wrong. One costs $25 and is right.

The free way (estimates): Look up each appliance’s nameplate wattage, estimate how many hours per day the compressor or motor actually runs, multiply. For a refrigerator, the compressor runs roughly 30-40% of the time — so a 150W fridge uses about 150W × 8 hours = 1.2 kWh/day. For a furnace blower, estimate based on outdoor temperature: a cold day might mean 10-12 hours of runtime, a mild day 4-6 hours.

This gets you in the ballpark. If you design to 150% of the estimate, you will probably be fine. But you might also be 50% off in either direction.

The right way (measured): Buy a Kill-A-Watt meter or similar plug-in energy monitor. Plug each appliance through it for one full week during normal use. Write down the total kWh for the week, divide by 7. Now you have your actual daily consumption — not an estimate, not a nameplate fantasy, but the real number.

I have seen refrigerators rated at 200W that pull 350W when the door seals are shot and the condenser coil is caked in dust. I have seen “efficient” furnace blowers that run at 800W instead of the nameplate 500W because the filter hasn’t been changed in two years and the ductwork is undersized. The $25 meter pays for itself in correctly-sized batteries.

Step 3: Factor in Depth of Discharge — The Capacity Math

Battery capacity is not the same as usable capacity. This is where chemistry matters.

Lead-acid (flooded, AGM, gel): Discharging below 50% depth of discharge (DoD) permanently damages the plates. A 100Ah lead-acid battery gives you 50Ah of usable energy. Discharge it to 80% DoD a few times and you might get 200 cycles instead of 1,000.

LiFePO4 (lithium iron phosphate): Safely discharges to 80-90% DoD with minimal cycle life impact. A 100Ah LiFePO4 battery gives you 80-90Ah usable. Most manufacturers warrant to 70% capacity after 4,000-10,000 cycles at 80% DoD.

Here is the math that converts your daily kWh into battery count:

Required rated capacity (kWh) = (Daily kWh × Backup Days) ÷ DoD
Battery count = Required rated capacity ÷ Rated capacity per battery

Example 1: 2-bedroom apartment (gas heat, city water)

Daily critical load: 3.5 kWh (fridge, lights, router, furnace blower in winter) Backup target: 24 hours

• With LiFePO4 at 90% DoD: (3.5 × 1) ÷ 0.9 = 3.9 kWh rated capacity needed One 5 kWh EG4 server rack battery at $1,300. Done.
• With lead-acid at 50% DoD: (3.5 × 1) ÷ 0.5 = 7 kWh rated That is six 12V 100Ah AGM batteries ($200 each = $1,200) — and they weigh 400 lbs and die in 5 years.

Example 2: 4-bedroom suburban house (gas heat, sump pump)

Daily critical load: 8 kWh (2 fridges, chest freezer, furnace blower, sump pump, lights, router) Backup target: 24 hours

• With LiFePO4 at 90% DoD: (8 × 1) ÷ 0.9 = 8.9 kWh rated Two EG4 5.12 kWh server rack batteries ($2,600 total). Or one Enphase IQ 10P ($5,000-6,000 installed with controller). Or one Tesla Powerwall 3 at 13.5 kWh covers it with room to spare.
• With LiFePO4 for 48 hours: (8 × 2) ÷ 0.9 = 17.8 kWh Four EG4 server rack batteries ($5,200). Or two Powerwall 3 units ($15,000-20,000 installed).

Example 3: Rural property with well pump, 2 freezers, no city water

Daily critical load: 12 kWh (well pump at 2 kWh/day with high surge, 2 freezers, refrigerator, furnace blower, lights, router) Backup target: 48 hours (rural outages run long)

• With LiFePO4 at 90% DoD: (12 × 2) ÷ 0.9 = 26.7 kWh rated Six EG4 rack batteries ($7,800) + a 6,000W inverter-charger ($1,500-2,500). Or two FranklinWH aPower 2 units (30 kWh total, 10 kW continuous each — handles well pump surge) at $20,000-28,000 installed.

This is the point where many people realize their backup target was the wrong number to optimize. A 48-hour rural backup at 12 kWh/day is a serious system. If the budget does not stretch to 27 kWh of battery, you have two levers: reduce daily load (do you really need both freezers on backup?) or add a generator to cover the extended outage tail. A $1,000 inverter generator that runs 4 hours a day can cut your battery requirement in half.

Voltage Matters: Why 48V Is the Home Backup Standard

If you are building a DIY system or evaluating pre-built options, you will encounter 12V, 24V, and 48V configurations. For home backup at any meaningful scale, 48V is the answer.

Ohm’s Law does the explaining: Power (W) = Voltage (V) × Current (A). A 5,000W load on a 12V system pulls 416 amps. That requires 4/0 AWG cable — roughly the diameter of a garden hose — and the resistive losses are significant. On 48V, the same 5,000W pulls 104 amps. Standard 2 AWG cable handles that with room to spare.

There is also a regulatory reason. NEC Article 480 and related sections tighten significantly above 60V DC. At 48V nominal (which charges to about 56-58V), you stay below the threshold where arc-flash requirements, disconnect rules, and enclosure standards jump in complexity and cost.

12V has its place: RVs, boats, small off-grid cabins with loads under 2,000W. 24V is a middle ground rarely worth the trouble — if you are going above 12V, go to 48V and get the full benefit. Every major home battery manufacturer has standardized on 48V for exactly these reasons.

Real-World Sizing Table

Here is a quick-reference table based on real installations I have done or specified:

The EG4 rack battery is my reference for DIY because it is the best value in server-rack LiFePO4 right now: 5.12 kWh, 48V, UL-listed, ~$1,300 each. The pre-packaged systems (Powerwall, Enphase, Franklin) cost more but include the inverter, transfer switch, monitoring, and an installation warranty. Your choice comes down to whether you value dollars per kWh or having a single phone number to call when something goes wrong.

If you want the dollars-per-kWh route and are comfortable with electrical work, a DIY LiFePO4 rack system is the cheapest path to 20+ kWh of backup.

Adding Batteries Later: Plan Ahead or Pay Twice

Battery systems are modular — up to a point. The critical constraint is the inverter. If you buy a 3,000W inverter today and later decide you want 10,000W of output, you cannot just add batteries. You need a new inverter. So size the inverter for your eventual target, even if you start with fewer batteries.

The second constraint is battery age. Lithium batteries degrade gradually, and their internal resistance increases over time. If you add a brand-new battery to a bank that has been cycling daily for 3 years, the new battery will be pulled down to the old bank’s effective capacity — the weakest cell dictates the pack performance. The rule of thumb: add batteries within 12-18 months of the original installation, or plan to replace the entire bank when you expand.

Across brands, expansion is generally not possible. A Tesla Powerwall talks to Tesla’s Gateway. An Enphase IQ battery talks to Enphase’s System Controller. A FranklinWH aPower talks to the FranklinWH aGate. These ecosystems do not interoperate. If you start with Enphase and later want to add a cheaper EG4 rack for extra capacity, you will need a separate inverter and AC-coupling — which is possible but adds cost and complexity.

The Bottom Line

The number of batteries you need is not a fixed answer. It is the output of a simple equation: (daily kWh of critical loads × backup days) ÷ depth of discharge. The hard part is getting daily kWh right, and most people overestimate by including loads that should not be backed up at all.

Start with a Kill-A-Watt meter and a weekend of measuring. That $25 investment will probably save you $2,000 in oversized batteries.