How this calculator works

Everything the model does, in plain language — followed by every fixed number it uses and where that number comes from. If a figure is a published source, it's linked. If it's my own engineering estimate, it says so.

Open the calculator →Read this alongside the tool, or come back to it when a result surprises you.

The core comparison

This tool treats rooftop solar as a long-term, bond-like investment and asks a single question: over your chosen horizon, does buying the system beat investing the same amount of cash? Both sides are measured the same way, so the comparison is apples-to-apples.

The cash side grows the system's price at your expected return, compounded, with tax applied to the gains. The solar side is the yearly bill savings — the electricity you no longer buy, plus the gasoline a solar-charged EV avoids — reinvested each year at that same after-tax return. Because avoided costs aren't taxed but investment gains are, the cash side clears a slightly lower after-tax bar, and the tool accounts for this.

The headline measures are payback (the year cumulative savings cover the system cost), the equivalent return (the steady annual return the solar earns on its price, to compare against your cash assumption), and the cost of doing nothing (the escalating utility and fuel bills you'd keep paying without solar). Over time, electricity savings grow with your utility escalation rate (which already includes general inflation), gasoline savings grow with a separate fuel-price outlook, and panel output fades with a small annual degradation. All figures are shown in nominal (future) dollars.

Capture: the heart of the estimate

Solar only earns full value for the energy you actually use or store; the rest is exported — at retail, at a low net-billing rate, or nothing, depending on your utility. The capture percentage is built from your net-metering regime, battery size (with a ~10% round-trip storage loss), roof orientation, and the timing of your usage, including whether an EV charges during peak sun. Production you can't use or store is valued at your export rate, and any remaining household demand is still bought from the grid.

How exported solar is credited

When your panels make more than you're using at that moment, the extra flows to the grid — and what your utility pays for it varies enormously by program and region. Instead of asking you to name a tariff, the calculator asks which of three mechanisms applies and lets you set the per-kWh rate.

At (or near) my retail rate — traditional net metering. Every exported kWh is credited at (or close to) your full retail rate, up to a year's worth of your own usage — your meter effectively runs backward, so a kWh sent at noon offsets a kWh drawn at night. Timing and batteries change your resilience, not the dollars. This is legacy NEM 1 / NEM 2 in California and still the norm across much of the US. Production beyond a full year of use is bought out at a low annual true-up rate (see below).
At a lower set export rate — net billing. Exports are credited at a rate below retail (an avoided-cost or value-of-solar figure) while the power you draw still costs full retail. That gap makes using or storing your solar far more valuable than exporting it. This is California's NEM 3 (Net Billing Tariff) and many newer programs. The calculator's default for this mode is $0.06/kWh — see the derivation below.
Little or nothing — non-export. The system never pushes power to the grid; anything it can't use or store is curtailed. Used where interconnection is limited, or to protect a legacy tariff on a separate system. Here, capturing your own production is everything.

The $0.06/kWh export-credit default

For the "lower set export rate" mode, the calculator pre-fills $0.06/kWh and lets you change it. That figure is a generation-weighted annual average — the mechanically defensible way to collapse a rate that actually varies by hour into one number: weight each hour's avoided-cost value by the share of a typical system's annual export that lands in that hour. For a system exporting without a battery, that weighting falls mostly on low-value midday hours, which is why every serious estimate lands in the $0.05–0.08/kWh band.

Cross-checks:

First-year net-surplus / export values across the three California utilities have run roughly 4.6–8.2¢/kWh (SDG&E low, PG&E high); the three-utility average is near 6.7¢, so $0.06 is slightly conservative — the right direction for a planning default.
Independent write-ups of NEM 3 economics consistently cite a blended $0.05–0.08/kWh for a no-battery residential system.
It is likely low if you're a PG&E or SCE residential customer receiving the ACC Plus adder (about $0.009/kWh, or $0.036 for CARE/FERA, fixed for nine years), or if a battery shifts your exports into summer evenings, where the rate spikes to $1–3/kWh in the top hours.
It is not the "value of solar." Grid-value studies put distributed solar's full worth 2–3× higher (Berkeley Lab's 2025 review: roughly $0.13–0.18/kWh). That gap is a policy debate; this field reflects what a utility actually credits, not what solar is worth.

The underlying data is public. In California, PG&E posts its Energy Export Credit value sheets (a month × hour heat map, one per interconnection-application year; full hourly data at pge.com/eecvalues) and SDG&E posts its export-pricing files; both derive from the CPUC's Avoided Cost Calculator. Outside California, look for your state's value-of-solar study or your utility's avoided-cost / feed-in schedule (DSIRE catalogs state policies). For the reasoning behind why a single number is defensible at all, see Berkeley Lab's A Review of Value of Solar Studies and One Year In: Tracking the Impacts of NEM 3.0.

Household savings vs. car savings

The system's yearly savings come from two different sources, valued on different terms. Household savings are the grid electricity you no longer buy: each kWh of solar you use at home is worth your retail electricity rate, so household savings = self-consumed kWh × that rate. Car savings come from replacing gasoline. I size the EV's yearly need as miles ÷ MPGe × 33.7 kWh — 33.7 kWh is the energy in one gallon of gasoline (the EPA gasoline-equivalent figure) — and value it at the gas it avoids: (miles ÷ MPG) gallons × your price per gallon. Spread across those kWh, a solar-charged mile is typically worth ~$0.50–0.70 per kWh — well above a ~$0.30–0.40 retail electricity rate.

Time-of-use rates

On a flat plan the grid rate is one number off your bill. On a time-of-use (TOU) plan the price changes through the day — an expensive "peak" in the late afternoon and evening (often 4–9pm) and a cheaper off-peak the rest of the time — and no single number captures that. The tool takes both rates off your bill and works out the blend: your panels produce during the day, but the pricey peak usually hits after the sun is low, so the value of your solar depends on how much of it lands in the peak window. A battery stores midday solar and releases it into the evening peak; a west-facing roof produces later in the day, so more of it overlaps the peak directly. The tool adds these into a peak share and blends the two rates by it. This is a planning estimate, not an hour-by-hour simulation.

On the escalation rate

I set a likely-conservative baseline escalation of 5%. The last 10 years have run an annualized increase of about 6.75% by my calculation. Starting in 2027, PG&E is officially requesting steady year-over-year revenue increases; analyses of the filing project roughly an 8% increase in 2027 (to ~38.4¢/kWh) and about 6.1%/yr through 2030 (to ~42¢/kWh). Your own utility and tariff will differ — enter what you can defend from your bill history.

Under the hood: fixed values and modeling choices

These are the values the form doesn't ask about. Where a figure comes from a published source it's linked; the rest are my own modeling estimates — reasonable engineering approximations, labeled as such. Verify anything material against a real quote and your own bill.

Capture — how much solar you use or store

25% / 35% / 45%Base share of output used as produced, for Low / Typical / High daytime use, before any battery. Studies put no-battery self-consumption near 25–40% (PV self-consumption research); the three tiers are mine.

10% – 85%Floor and ceiling on that direct-use fraction after adjustments. My modeling estimate.

±5% (capture effect)How much roof direction changes how much of your solar you self-consume vs. export — a small swing, since a home draws power across the whole day. A study of real consumption profiles found reorienting away from south moves the self-consumption index up to ~6 points (Vandevelde et al., Energies 2020); my ±5% sits in that range.

up to +30%Extra capture when the car charges during production hours, at 80% assumed coincidence. My estimate.

92%Maximum modeled capture — the model never reaches 100%. My modeling estimate.

Seasonality & battery

Monthly solar profileCapture is computed on a representative day each month, weighted by a normalized fixed-tilt profile — Jan–Dec: 5.5, 6.5, 8.5, 9.5, 10.5, 10.8, 11.0, 10.3, 9.2, 7.6, 5.6, 5.0%. Shape from NREL PVWatts.

90%Battery round-trip efficiency — energy stored and later drawn loses ~10%. NREL's Annual Technology Baseline puts lithium round-trip at ~85–90% (NREL ATB).

92%Battery availability factor — imperfect dispatch and capacity held back for backup. Deliberately not a depth-of-discharge derate: manufacturers' headline capacities (a Powerwall's 13.5 kWh, say) are already the usable figure, so discounting for depth-of-discharge again here would double-count it. My modeling estimate.

0.5%/yrDefault panel output loss per year, applied on the solar side of the comparison. NREL's review of field data across thousands of systems finds a median degradation near 0.5%/yr (NREL — PV degradation rates); you can change it in the tool.

Flat monthly loadConsumption is assumed uniform across months; seasonal load (summer AC) is a future refinement. My simplification.

Roof direction → when solar is valued (peak-timing effect)

Peak window ≈4–9pmAssumed high-price evening window; check your plan (example: PG&E).

Peak overlap 0% – 33%Of the solar you use directly, the share landing inside that window — east ≈0%, west ≈33% — following a cosine centered on WSW (≈255°). My modeling estimate.

Battery → 100% peakStored solar is assumed to be discharged entirely into the peak window. My simplifying assumption.

Where it breaks down. A west array is still producing at 5pm while an east array finished by early afternoon, so orientation strongly shifts how much output lands in a fixed evening peak. The "100% to peak" battery assumption only holds if the battery is large enough to cover the whole peak, is cycled daily rather than reserved for backup, and your peak usage is big enough to absorb it. When those don't hold, real peak capture is lower than the effective rate suggests.

Exports & tariff mechanics

$0.06/kWhEditable default export credit for the "lower set export rate" mode — a generation-weighted annual proxy for California net billing. It is a planning midpoint, not your number: research your specific utility's net-billing or export-rate schedule to determine whether it's right for you, and enter that figure instead. See the derivation and sources above.

$0.03/kWhAnnual true-up rate applied only in the "at/near retail" mode, and only to production beyond a full year of your own usage. Real net-surplus / avoided-cost buyout rates run roughly 2–4¢/kWh (some programs pay $0); $0.03 is a mid estimate. This only affects deliberately oversized systems on legacy (retail-rate) net-metering agreements — for a system sized at or below your annual use, it never comes into play.

Credit capped at annual useIn the "at/near retail" mode, the retail-rate export credit is limited to a year's worth of your consumption; anything beyond that gets the true-up rate. This is the standard net-metering true-up rule (EIA rates). The "lower set export rate" mode applies its rate to all exports with no cap.

Cheapest-to-export firstWith a second roof surface credited differently from the main system, self-consumption is drawn from the lowest-export-value array first. My dispatch assumption.

EV & gasoline

33.7 kWh/galEnergy in a gallon of gasoline (EPA gasoline-equivalent), used to size the EV's kWh need — see fueleconomy.gov.

Car before homeCaptured solar is credited to the highest-value use (gasoline avoided) first, then the home. My allocation choice.

1% / 3% / 5%Gas-price outlook presets (Low / Base / High) per year. Base ≈ flat in real terms per EIA's long-run outlook (EIA AEO); the exact figures are mine.

Financial engine

Nominal dollarsFigures aren't inflation-adjusted; rate escalation already includes general inflation. Stated convention.

After-tax returnBoth sides measured at your return after tax; solar savings reinvested at that same rate. Stated convention.

≥ 25-year scheduleThe model always runs to at least 25 years internally (a typical panel lifespan); the headline compares at your chosen horizon.

Federal tax creditNot applied automatically. Enter your net system cost to reflect an incentive you expect, or your gross cost for the unsubsidized picture. Incentive rules change; entering it yourself keeps the result honest.

Defaults

Pre-filled figuresSeeded from a specific Tesla quote (6,423 kWh, $26,806, 10 panels, 4.6 kW, 13.5 kWh battery) and PG&E rates ($0.39 grid; $0.52 / $0.40 peak / off-peak; $0.06 export). Replace with your own quote and bill.

Location & the panel-spec production estimate

If you don't have a quote, the calculator can estimate first-year production from a panel count, a panel wattage, and a location. The location list is 14 real coordinates — one per major US solar market, chosen for market size and geographic spread, not to stand in for a "typical" climate. Each was run individually against the real NREL PVWatts v6 API at a fixed 20° tilt, standard modules, and PVWatts' default 14% system loss, at eight compass orientations — 112 runs, saved as a static table so the tool stays a single file with no live API dependency. For an azimuth between the eight reference points the tool interpolates along a curve fitted to the real PVWatts output, not a plain cosine (a cosine over-penalizes east/west roofs and ignores the diffuse skylight they still collect).

This is separate from the capture / time-of-use engine: this model sets how much total energy an orientation produces in a year; capture sets when that energy lands relative to peak pricing. A west roof produces less total than a south roof but lands more of it near the evening peak — both real, for different reasons. The table has only 14 points, so anywhere else is an approximation to the nearest one, and a metro's own microclimates (coastal fog, inland valleys, elevation) can still differ from its anchor. For a real number, run PVWatts with your exact address or get an installer quote and paste it into "Annual production" with estimate mode off.

$3.00/W, $1,150/kWhCost-estimate defaults when no quote is entered. Per-watt sits between competitive marketplace averages (~$2.60–2.75/W, early 2026) and Berkeley Lab's national median for cash purchases (~$3.50/W); the battery figure comes from marketplace data putting a 13.5 kWh install near $15,600 before incentives. Budgeting placeholders, not prices — roof complexity, panel upgrades, trenching and permitting move real quotes more than system size does. The federal residential credit (Section 25D) expired at the end of 2025, so the estimate shows what you'd pay with no credit netted out.

460W / 430WDefault panel wattage for the primary system / added panels. Starting points — use your panel's actual spec sheet.

Planning for future electricity use

Sizing solar against last year's bill quietly assumes your usage stays flat — usually wrong for people considering solar, since electrification and solar tend to arrive together. The calculator lets you add expected loads (heat-pump water heater, heat-pump space heating, induction range, electric dryer, pool pump, spa, added AC, an always-on machine) so the modeled shortfall reflects where you're heading. The presets are mid-range starting points, each of which varies widely with climate, equipment and habits: HPWH ~1,200 kWh/yr when replacing gas (real-world reports cluster 900–1,500; DOE's category average runs ~2,195); heat-pump space heating 2,500 / 4,000 / 7,000 kWh/yr for mild / moderate / cold climates (whole-house figures run 3,000–8,500); induction range ~500 kWh/yr; electric dryer ~600 kWh/yr; pool pump ~2,500 kWh/yr (single-speed; variable-speed 50–70% less); spa ~1,800 kWh/yr. Always-on machines are computed from the draw and hours you enter. One deliberate conservatism: for gas-replacement rows the model adds the new electricity but doesn't credit back the gas you stop buying, so your real net cost is better than shown. (The EV is the exception — avoided gasoline is counted there.)

The stress test

Every headline number rests on assumptions, and some matter far more than others. The stress-test panel re-runs the entire model — the same code path as the headline — with one assumption pushed against you at a time, ordered by impact, and reports what the verdict becomes. These are deliberate adverse cases, not forecasts; the final row stacks them all at once, which is genuinely pessimistic and unlikely in full — if the answer survives that, it's robust. The headline sentence solves for the lowest utility escalation rate at which solar still wins, so you can see how much slack there is behind the escalation figure you entered.

Two rows worth naming. Rates → fixed fees: solar only offsets the per-kWh part of a bill, so a utility shifting revenue into a fixed monthly connection charge erodes solar savings even as bills rise — not hypothetical, PG&E restructured this way in 2026 (~$24/month into a Base Services Charge). Battery end of life: the one equipment risk modeled (inverters are excluded — warranties have lengthened to ~25 years and replacement costs fallen). At end of life the model takes whichever is cheaper over the remaining window: replace the battery, or run solar-only without it. The default failure year (12) and a 2026 replacement price are both deliberately harsh and both adjustable; the comparison doesn't price the loss of backup power, so "cheaper" isn't automatically "better."

These are planning estimates, not a guarantee or financial advice. Rates, incentives, taxes, degradation, and interconnection rules vary — verify the specifics of any quote before relying on these numbers.