Compare two solar quotes

How a verdict between two quotes is earned, not eyeballed

Cost per nameplate watt is where every comparison starts, and where careless ones end. It strips system size out of the two prices, which is necessary — a bigger quote is not a worse quote — but it treats every watt as equal, and watts are not equal: a watt behind a mediocre inverter on a shaded roof delivers less than a watt in a well-engineered installation, at the same nameplate.

The performance ratio is how quality re-enters the arithmetic. Multiplying each quote’s nameplate by its expected ratio yields an effective capacity — the fraction of the nameplate the installation is actually expected to make good on — and re-dividing each price by its effective watts re-prices both quotes in delivered terms. The pack’s own declared vectors include a pair where the verdict flips at exactly this step: the quote that is cheaper per nameplate watt turns out dearer per effective watt, because its lower expected performance costs more than its lower price saves.

The primary output is the signed difference between the two effective rates, and the pack declares the reading rather than leaving it to the reader: a warning names the better value on either side, and fires only when the difference clears a hair’s-width dead band. Inside that band the honest verdict is a tie — a difference smaller than the uncertainty in either performance ratio is no verdict at all.

What the verdict covers is value per effective watt, and nothing else. Warranty terms, installer track record, financing, equipment brand and the credibility of each quoted performance ratio all sit outside the closed form — and that last one matters most, because an installer who flatters their own ratio buys the verdict. The ratios entered here should be the reader’s honest expectations, informed by the measured-performance page of this cluster, not the brochures’.

When this calculation is used

• Choosing between two real quotations that differ in size, price and expected installation quality all at once.
• Testing whether a cheaper headline price survives adjustment for the performance the installation is expected to deliver.
• Re-running the comparison with pessimistic ratios, to see whether the verdict is robust or hangs on an installer’s optimism.
• Running the declared reverse workflow: the price at which the first quote would tie the second, as a negotiation target.

Worked example

Take the pack’s own verdict-flipping vector: quote A at four thousand dollars for three kilowatts-peak with an expected performance ratio in the mid-eighties of percent, against quote B at four and a half thousand dollars for three and a half kilowatts-peak expected to perform at eight tenths.

On nameplate cost per watt, B looks cheaper — more capacity per dollar on paper. But once each quote is re-priced per effective watt, the signed difference comes back on A’s side of the dead band and the pack’s declared warning names quote A the better value: its higher expected performance more than pays for its higher rate. That reversal is the whole argument for this page — the headline metric and the honest metric can disagree, and the decision belongs to the honest one.

Every figure is computed by the verified engine at load, and the scenario is one of the declared test vectors carried in the signed pack. The verdict is a value comparison between two estimates, not a prediction: both performance ratios are expectations, and whichever installer is chosen, the real installation will settle the question — this comparison is where the conversation with them starts, not where it ends.

What each input represents

Assumptions and limits

• Both quotes describe the same scope of work on the same incentive basis; the closed form normalises size and performance, never scope.
• Each performance ratio is a single honest expectation, although real performance varies with season and settles only after a metered year.
• Value per effective watt is the only axis judged — warranty, financing, brand and installer record are deliberately outside the form.
• The same arithmetic runs on both legs, so any bias in the inputs passes straight through to the verdict undamped.

What the guards protect against

• Both prices and both system sizes must be positive — the pack declares these as explicit domain guards, and ships the zero-price refusal as a test vector, because a free or empty quote has no rate to compare.
• Each performance ratio must sit above zero and no higher than one, guarded on both legs; a percentage typed where a decimal belongs lands far outside the interval and is refused rather than silently crowning the wrong quote.
• The verdict warnings fire only when the effective-rate difference clears a small declared dead band on either side, so a near-tie is reported as what it is instead of being called for whichever quote noise favours.

Provenance

Standard quote-normalisation metrics: cost per nameplate watt, performance-ratio-adjusted effective capacity, and cost per effective watt, differenced between quotes

One closed form evaluated identically for both quotes — price over nameplate watts, nameplate times performance ratio for effective capacity, price over effective watts — returning both legs and the signed difference in cost per effective watt, with dead-banded declared warnings naming the better value; the declared reverse workflow solves the first quote’s price for a target difference.

Educational reference for comparing quotations, not financial or professional advice; the verdict covers value per effective watt only. The signed pack carries its own citation, and the page reports the verification state of the release it mounted.