What Size Solar System Do I Need to Go Off Grid?

Your panel count answers a grid-tied question: how much array covers your usage over time. On a day the array makes more than you use, the surplus goes out through the meter. On a day it makes less, the grid makes up the difference.

That exchange is why an annual average is the correct input on grid. Days above and below the average cancel through the meter, and the array never has to carry any single day on its own. The timing problem belongs to the utility.

Cut the meter and that mechanism goes with it. Surplus above what the bank can hold is thrown away rather than banked, and every deficit comes out of storage that refills only from a later day's surplus.

So the same array is correctly sized on grid and badly undersized off it. What follows is the arithmetic, run on the defaults the calculators on this site already ship with.

The grid is the battery you already have

The grid-tied sizing guide makes the case for feeding your array the twelve-month average rather than one bill, and that advice is right for what it covers. Size to August and you buy panels the year never asks for; size to February and you fall behind every summer.

None of that survives disconnection. On grid, an average is a promise the meter keeps for you across the year. Off grid, an average is a number your house spends time on both sides of, and the low side has to be paid for at the moment it happens.

Three numbers, not one

The first number is the daily load. Everything here runs on 10 kWh a day, the solar battery bank calculator's own default. Every capacity below exists to serve that one figure.

The second number is storage. Ten kWh a day across one day of autonomy needs 10,000 Wh out of the bank, and dividing by the 0.8 depth of discharge gives 12,500 Wh of gross capacity — 12.5 kWh, or 260.42 Ah at the default 48 V.

Take the depth of discharge and the 48 V bus as given here. The battery runtime guide works through why a bank is rated twice, why only part of the nameplate is usable, and why large banks run at 48 volts rather than 12.

The third number is recharge, and it is the one most panel counts leave out entirely. The array cannot stop at producing what you consume each day. It has to produce that, and then produce more, to put back what the bank already handed over.

Those three are coupled, which is what makes the sizing awkward. A bigger bank carries the load longer but generates nothing, so extra storage without extra array is extra capacity sitting empty, waiting on a surplus that never arrives.

The array that covers the load has no headroom

Run the ordinary calculation first. A 10 kWh day is 300 kWh a month to the panel count calculator, and at 4.5 peak sun hours, 80% system efficiency and 400 W panels that asks for 2.778 kW of array.

Panels come whole, so it rounds up to 7 panels and an installed 2.8 kW. As a grid-tied coverage answer this is correct, and it is exactly what the tool was built to give you.

Push it back through the output calculator at the same assumptions: 2.8 x 4.5 x 0.8 comes to 10.08 kWh a day. Set that against the 10 kWh load and the entire surplus is 0.08 kWh.

That remainder is the whole recharge budget. The house eats almost everything the array makes, so once the battery has covered a shortfall, the array has to give the energy back at 0.08 kWh a day while still carrying the load in front of it.

The break-even sun hours

Turn the array around and it says the same thing more sharply. Divide the 10 kWh load by the 2.8 kW array and the 0.8 efficiency and you get 4.46 peak sun hours — the level below which this array cannot cover the day.

It was sized at an annual average of 4.5. The gap between the two is 0.036 sun hours, which is 0.79% of headroom between working and not working.

That figure is arithmetic, not weather. It makes no claim about how often your site sees any particular level of sun. It says only that this array needs very nearly the whole sizing assumption just to break even on the day's load.

A year contains conditions on both sides of its own average, by definition. On grid the meter reconciles them over months. Off grid, every day under 4.46 sun hours sends the bank to work, because at that level the array cannot finish the job alone.

Refilling the bank is the term that dominates

Say the bank does its job and delivers its 10 kWh across one day of autonomy. The 7-panel array now has 0.08 kWh a day to put it back, because the load still takes its 10 kWh first.

Divide 10 by 0.08 and the recovery time is 125 days. The bank covered a single day; the array needs four months of surplus to undo it, and any cloudy stretch in between starts the count over.

That is the sizing error laid bare. The bank calculation is right, the panel calculation is right, and bolting them together still gives you a system that cannot come back from a normal night's work.

Fix it by sizing for the recharge instead of the load. Choose a five-day recovery: putting 10 kWh back over five days needs 2 kWh a day of surplus, so total production has to reach 12 kWh a day.

That takes 9 panels, or 3.6 kW, which at the same 4.5 sun hours and 80% efficiency makes 12.96 kWh a day. After the load takes its share, 2.96 kWh is left over.

The bank refills in 3.4 days, and break-even drops from 4.46 to 3.47 peak sun hours. Two panels — seven to nine — is the difference between 125 days and 3.4.

Read the surplus column, not the production column. Production has to feed the house before a single watt-hour reaches a flat bank, and it is what survives that subtraction that sets the recovery time. Production climbs 43% down that table while the surplus climbs 55-fold.

The two failures happen on the same days

Array margin and battery margin look like two separate safety nets. They are not, because the bank starts discharging at precisely the moment production stops covering the load. One condition triggers both.

Make it concrete with a stretch of days at 3.5 peak sun hours — a figure chosen for the example, not measured anywhere. The 7-panel array makes 2.8 x 3.5 x 0.8, which is 7.84 kWh against a 10 kWh load, short by 2.16 kWh every day.

At that rate the bank's 10 kWh is gone in 4.63 days, and then the house is dark until the sun comes back. The 9-panel array on the identical day makes 10.08 kWh and never touches its battery, because its break-even sits at 3.47 rather than 4.46.

This is the opposite of independent failure. With independent risks, margin in one component is still there when the other weakens. Here the shortfall that empties the bank is the same shortfall that removes the surplus which would refill it.

A full bank hides an undersized array for a while, which is what makes the mistake expensive. Everything looks adequate as long as storage is covering the gap, and the problem only shows once the bank is low and the array still has nothing spare.

What more autonomy actually costs

Autonomy is how long the bank can carry the load while the array cannot. Hold the load at 10 kWh a day, the bus at 48 V and the depth of discharge at 0.8, and storage rises in a straight line with the days you ask for.

Going from one day to two doubles gross capacity from 12.5 kWh to 25 kWh, and five days wants 62.5 kWh. The battery watt-hours calculator turns any of those into the amp-hour ratings printed on real cells.

None of that added storage reduces the load or adds a watt of production. The house still takes 10 kWh a day, and every kilowatt-hour drawn out of the larger bank still has to come back through array surplus later.

So the array does not shrink when the bank grows — if anything the opposite. A bank sized for five days can run up a five-day deficit before conditions improve, which makes the recharge rate more consequential, not less.

Storage is the heavy part of an off-grid design, but capacity is not resilience. Pick autonomy and recovery together, or you end up with a system that discharges impressively and has no credible route back to full.

What going off grid adds to a normal house

Take the panel count tool's default house at 900 kWh a month. Grid-tied, that asks for 21 panels at 400 W and an 8.4 kW array, with no storage at all, because the coverage calculation hands the timing problem to the utility.

Treat the same house as off grid and the daily load is 30 kWh. At 48 V and 0.8 depth of discharge, one day of usable energy needs a 37.5 kWh gross bank — 781.25 Ah — before any recharge margin exists.

The array grows to 25 panels, or 10 kW, producing 36 kWh a day at the same 4.5 sun hours and 80% efficiency. It exceeds the grid-tied array because it has to make recharge headroom after carrying the house, not merely match it.

The whole difference is 4 more panels and 37.5 kWh of storage. The panels are the part that looks like the original answer; the storage is the part the panel count never showed you, and it is doing the balancing work the grid used to do for free.

Getting your daily load right matters more than any of this, and how much electricity costs covers reading real usage off a bill rather than guessing from nameplates. Feed the arithmetic a wrong load and every number above is wrong with it.

Equipment selection, protection, wiring, controls and site conditions need a licensed installer or electrician. An off-grid system is not a do-it-yourself grid substitute. The arithmetic defines the energy problem; a qualified professional turns that into a safe installation.

Common mistakes to avoid

• Taking the panel count as an off-grid system size. The 7-panel result covers the average load correctly, and its 0.08 kWh of daily surplus gives you nothing to recharge a battery with.

• Sizing the bank without timing its refill. A 12.5 kWh bank delivers the 10 kWh it promises, and the load-sized array then needs 125 days to put that energy back.

• Counting all production as charging energy. Of 10.08 kWh made in a day the house takes 10 first, so only the last 0.08 kWh ever reaches storage. The load is subtracted before the bank sees anything.

• Reading panel rounding as headroom. Rounding 2.778 kW up to 2.8 kW feels like margin, and it buys 0.79% above the break-even the array actually needs.

• Adding autonomy without revisiting recovery. A 62.5 kWh bank rides out five days, and every kilowatt-hour it spends still waits on the same array surplus to come back.

• Treating array shortfall and battery drain as separate risks. The bank empties on exactly the days the array falls short, so whatever causes the deficit is also what prevents the recovery.

Your grid-tied array has one job: cover usage over time. An off-grid system has three — carry the load, bridge the days it cannot, and make enough surplus to refill the bank afterwards. Size all three together, then hand the design to a licensed installer or electrician.