What Size Solar System Do You Actually Need?

Sizing a solar system comes down to one formula, and you can compute it before any installer sets foot on your roof:

System size (kW) = annual consumption (kWh) ÷ (daily sun hours × 365 × efficiency factor)

Assume a home using 11,000 kWh per year, in a region averaging 4.5 sun hours per day, with a 0.80 efficiency factor for real-world losses. That’s 11,000 ÷ (4.5 × 365 × 0.80) = 11,000 ÷ 1,314 ≈ 8.4 kW. At 420 W per panel, that’s twenty panels. That is the skeleton of every professional sizing exercise, and everything an installer’s software adds is refinement of those three inputs. The refinements matter, though — get the consumption number wrong and you buy the wrong system; get the sun-hours or loss assumptions wrong and your “100% offset” system delivers 80%. So the work is walking the formula input by input, then dealing with the constraints of roof, budget, and utility rules that push the final answer away from the mathematical one.

The three inputs, measured honestly

Everything starts with kWh per year, and it must come from your own bills, not an average. National figures span an enormous range — an efficient apartment-sized home might use 5,000 kWh a year while an all-electric house in a hot climate with a pool clears 20,000 — so a system sized for the “typical” home could be half or double what yours needs. Pull twelve consecutive months of bills (your utility’s online portal usually shows a usage history) and sum the kWh. Twelve months matters, because a summer-only sample overweights air conditioning, a winter sample overweights heating, and either skews the result badly. If you can only find a few bills, note that usage in most US homes peaks in summer, winter, or both, and interpolate cautiously before you verify. Then adjust for the future, because you’re sizing for the next 25 years and not the last 12 months. An electric vehicle typically adds 2,500–4,000 kWh per year depending on mileage — often the single largest addition a household ever makes. A heat pump replacing gas heat can add several thousand kWh annually while cutting the gas bill, and the economics usually favor the swap even though the electrical load is real. Heat pump water heaters, hot tubs, pools, home additions, and new household members all push usage up; kids leaving for college and efficiency upgrades push it down. A home using 11,000 kWh today that plans to add an EV should size against roughly 14,000 kWh, because undersizing for known future loads is the most common sizing regret and it’s expensive to fix — adding panels later means paying permitting, design, and mobilization costs all over again for a handful of modules.

If you can’t lay hands on a clean twelve months — you moved in recently, or the account changed names — reconstruct the number rather than guessing at it. Most utilities keep a year or two of usage history in the online portal even for a prior occupant’s account, and a call to customer service can often pull the address’s consumption record. Failing that, build the year from the months you do have by leaning on the shape of usage: note whether your home is cooling-dominated, heating-dominated, or both, scale the summer or winter bills you’re missing against the ones you have, and deliberately round up rather than down, since undersizing is the costlier mistake. Whatever method you use, the goal is a defensible annual kWh figure you can point to, not a round number a salesperson supplied — because every downstream calculation inherits the error in this one.

The second input, sun hours, measures how much usable solar energy your location receives daily, averaged across the year — not how long the sun is up. One sun hour equals one hour of full-strength (1,000 W/m²) sunlight, and US locations range from roughly 3.5 sun hours per day in the cloudiest Northwest and Northeast regions to more than 6 in the desert Southwest, with most of the country between 4 and 5.5. The consequence is proportional: the same 11,000 kWh target needs about a 10.7 kW system at 3.5 sun hours but only 6.3 kW at 6, so location alone changes the required system by 70%. Use a location-specific figure — the sizing calculators linked below apply one for your area — rather than a national default. Your roof then modifies the regional number. South-facing capture is the benchmark; east or west orientation typically costs 10–20% of production, which you compensate for by sizing up proportionally, and shading losses come off the top too. A home needing 8.4 kW on paper with a west-facing roof and light morning shade might genuinely need 9.5–10 kW to hit the same annual output.

The third input is the one most back-of-envelope calculations skip, and skipping it undersizes every system by a fifth. A “7 kW” system never delivers 7 kW × sun hours of energy, because between the panel nameplate and your meter sit losses: inverter conversion of a few percent, wiring resistance, panel soiling and dust, temperature (panels lose output as they heat up, so hot climates give with sun hours and take back a little with temperature), module mismatch, and occasional downtime. Stacked together, real systems typically deliver 75–85% of the naive calculation. Using 0.80 as the derate is the standard planning assumption; drop toward 0.75 for hot climates or dusty areas, nudge toward 0.85 for cool climates with clean, well-ventilated installations. It’s one number, but it’s the sizing error hiding in most kitchen-table estimates, and if the gap between panel ratings and real-world output feels murky, solar panel wattage explained covers what nameplate numbers do and don’t promise.

Choosing an offset target, and the reality checks the roof imposes

The formula sizes for 100% offset — producing as much annually as you consume — and that’s the intuitive target, but the right target depends on your utility’s rules. Under full net metering, 100% is usually the sweet spot: every kWh has the same value whether you use it or export it, so matching production to consumption maximizes value without waste, and going meaningfully over 100% rarely pays because most utilities credit annual surplus at a low wholesale rate or wipe it annually. Under net billing or weak export rates, exported power is worth much less than avoided purchases, and each additional panel yields declining returns since the bigger the system, the larger the share of its output that exports at the discount rate — systems in these markets often pencil best at 60–90% offset, sized to cover your daytime baseline plus what a battery or load-shifting can absorb. Under tight roof or budget constraints, partial offset is still worthwhile, because solar economics are roughly linear: a system covering 60% of usage delivers about 60% of the savings, and thanks to fixed project costs, per-watt pricing gets worse for smaller systems, so cover as much as constraints sensibly allow, once. Some utilities also cap system size, commonly at 100–120% of your historical annual usage, so the paperwork itself may enforce a ceiling — ask before falling in love with a design. There’s a nuance on the other side, too: because the last few panels on a project are the cheapest ones, sizing modestly above today’s needs can be rational when a future load is likely but not certain. Assume the marginal panels cost $1.60/W installed while the system averages $3.00/W, and capacity bought now for a probable EV costs roughly half of what the same capacity costs as a retrofit later. The gamble only fails if the load never materializes and your exports earn little, so weigh it against your utility’s rules rather than treating it as a universal rule of thumb.

A battery reshapes the offset question rather than answering it. Storage lets a home use more of its own midday production after dark, which raises the effective value of each panel in a weak-export market and can justify sizing the array a little larger than a battery-less design would — but a battery neither generates energy nor removes the utility’s size cap, so it shifts when your solar gets used, not how much you’re allowed to install. Size the array to your consumption and export rules first, then decide whether a battery earns its place on top of that design.

The math produces a number in kW; the roof decides whether that number fits. Divide system size by panel wattage to get panel count and area — assume 420 W panels at roughly 21.5 square feet each, and the 8.4 kW example needs 20 panels and about 430 square feet of suitable, contiguous-ish roof. “Suitable” excludes north faces, heavily shaded areas, and the fire-code setbacks most jurisdictions require around ridges and edges, so a “2,000-square-foot roof” might offer only 500 usable square feet. How roof size affects solar works through this constraint in detail; the short version is that when space runs short, higher-wattage panels are how you buy capacity back. Inverter sizing surprises people next: installers commonly pair panels with an inverter rated below the array’s DC total, and a DC-to-AC ratio of 1.1–1.3 is normal and deliberate, since arrays rarely hit nameplate simultaneously, so don’t be alarmed by a 7.6 kW inverter under a 9 kW array, though you should ask about the ratio if it exceeds ~1.4. Electrical panel capacity is the last reality check — larger systems backfeed more current, and older 100 A panels sometimes can’t accept a big system without an upgrade, which is a cost question more than a sizing limit but can make the last kilowatt disproportionately expensive.

Pull it all together for a concrete case. Assume 12,600 kWh/year of current usage from twelve real bills, an EV arriving next year (+3,000 kWh), a west-facing roof (apply a 12% orientation penalty), 4.8 regional sun hours, and full net metering, so target 100% offset:

1. Future consumption: 12,600 + 3,000 = 15,600 kWh/year.
2. Base sizing: 15,600 ÷ (4.8 × 365 × 0.80) = 15,600 ÷ 1,402 ≈ 11.1 kW.
3. Orientation adjustment: 11.1 ÷ 0.88 ≈ 12.6 kW.
4. Panel count at 420 W: 12.6 kW ÷ 0.42 = 30 panels, needing roughly 645 square feet of usable roof.
5. Reality check: if the west face holds only 24 panels, the options are higher-wattage modules, using a second roof face, or accepting ~80% offset — all legitimate, with different price tags.

Run your own inputs through the how many solar panels do I need calculator to get the panel count for your usage and region, and if you’re working from a monthly figure, the walkthrough for a 1,000 kWh/month home shows the same math at a common consumption level.

The two confusions that produce the wrong system, and how to vet a proposal

Most sizing errors trace back to one of two confusions, both worth naming plainly. The first is sizing from the dollar bill instead of the kWh figure. “My bill is $220 a month” is not a sizing input, because the same $220 represents wildly different consumption at different rates — about 1,830 kWh/month at $0.12/kWh but only 730 kWh/month at $0.30/kWh. A salesperson who sizes from your dollar bill in a high-rate market will propose a system more than twice as large as one sized from the same bill in a cheap-power state, and one of them is wrong. The kWh number on the bill is the input; the dollar number is the output you’re trying to shrink. The second confusion is mixing up kW and kWh, and the units trip up nearly everyone at first. Kilowatts measure power — the system’s instantaneous capability, its engine size — while kilowatt-hours measure energy, power sustained over time, the thing your meter counts and your utility bills. An 8.4 kW system doesn’t produce 8.4 kWh per day; it produces its rating times sun hours times the loss factor, roughly 30 kWh on an average day in the example above. When a proposal says “10 kW system” and your bill says “1,100 kWh per month,” the formula in this article is precisely the bridge between those two numbers, and anyone quoting one unit as if it were the other — which happens in sales conversations more than it should — is telling you they don’t understand the product. A third, smaller trap sits alongside these: sizing to your peak month instead of the annual total. A system sized to fully cover August’s air-conditioning bill will overproduce for the other ten months, and under most export regimes that surplus earns little, so annual consumption divided by annual production is the honest match, with monthly imbalances left for net metering credits or a battery to smooth.

When the roof or the budget won’t stretch to the number the formula produces, the decision is about what to give up, and it helps to know the levers in order. Higher-wattage modules buy back capacity within the same footprint first, since a switch from 400 W to 430 W panels lifts a space-constrained array by roughly 7% for a modest price bump. A second or third roof plane comes next, accepting the orientation penalty on a west or east face to add panels the south face couldn’t hold. Only after those should you settle for partial offset, and partial offset is a perfectly rational outcome rather than a failure — a system covering 75% of a large consumption is often a better investment than one covering 100% of a small one, because the fixed project costs are spread across more producing panels. What you should not do is quietly shrink the consumption estimate to make the design fit the roof, which is the reverse of honest sizing and guarantees a system that disappoints against the bills that prompted it.

When quotes arrive, the sizing should survive three questions. Ask what consumption number they sized against, and whether it matches your bills plus your stated future loads — proposals based on “typical usage for your area” instead of your meter history are guessing. Ask what production they project, and what derates for orientation, shading, and losses that projection includes; a figure above roughly 1,600 kWh per kW per year deserves skepticism anywhere outside the sunniest states. And ask what offset the design achieves under your utility’s actual export rules, not under net metering rules your utility may no longer offer. An installer who answers all three crisply is doing engineering; one who answers with monthly-payment talk is doing sales, and the difference will live on your roof for 25 years. A fourth question sorts the careful from the merely competent: did they ask about loads you plan to add, and did they check whether your electrical panel can accept the system they’re proposing? A designer who never inquired about a possible EV or heat pump has sized for your past, and one who ignored a dated 100-amp panel may have quoted a number that grows by thousands once the required upgrade surfaces at the site survey. Neither omission is necessarily dishonest — a rushed sales visit skips these routinely — but both put surprises in your future that a few questions now would surface today, when you can still weigh them against another quote rather than after a contract is signed. Before the appointment, spend ten minutes with the solar panel size calculator — walking in with your own number changes the entire conversation, because you stop evaluating whether the proposal sounds reasonable and start checking it against a figure you derived yourself.