The 12 Items a Solar Proposal Must Show, and Why

A proposal on the table says 8.40 kW DC and 11,900 kWh in the first year. Divide one by the other and you get 1,417 kWh per kW of DC capacity per year. That ratio is the only thing in the document that can be compared against another quote for a different-sized array, and it is not printed anywhere on the page.

Each of the twelve below is either printed in the document or has to be requested in writing. The test for putting one on the list was narrow: something specific stops being checkable when that item is missing.

The stakes moved in 2026. With section 25D closed on 31 December 2025, there is no 30% federal credit absorbing a third of a pricing error any more. What used to be a rounding difference is now the whole result.

Items 1 to 4: what goes on the roof, and what it is claimed to make

1. Both system sizes, DC and AC. Twenty-one modules at 400 W each is 8.40 kW DC. Paired with a 6.72 kW AC inverter that is a DC-to-AC ratio of 1.25. A proposal printing only one of the two numbers makes two things invisible: whether the inverter will clip production on the brightest hours, and which denominator the price per watt was computed against. Quoting dollars per watt on the AC figure instead of the DC figure makes the same system look 25% cheaper per watt.

2. Module and inverter model numbers, with quantities. Not "high-efficiency panels." A manufacturer, a model, a wattage, a count. Without them you cannot pull the datasheet, you cannot read the warranty that item 10 depends on, and you cannot tell whether the contract's "or equivalent product" clause lets a different panel arrive on installation day. If that clause is in there, ask what equivalent means and where the contract defines it.

3. The annual production estimate and the tool that produced it. California's disclosure regime is explicit about this. The CSLB Solar Energy System Supporting Information form, developed jointly with the CPUC under Business and Professions Code section 7169, carries a form field named Name of Calculator Used, alongside fields for annual solar generation, annual electricity usage, and an explanation if no annual usage data was available (CSLB Supporting Information form, read 21 August 2026). A production figure with no named tool behind it cannot be reproduced, and a figure you cannot reproduce is a claim rather than an estimate.

4. The shading number the model actually used. This is the item most often missing and the one that moves payback hardest. Energy Trust of Oregon's programme guide defines Total Solar Resource Fraction as the combined effect of shading, tilt and orientation, and writes it out as a product:

Shading = 100% - annual loss caused by shading
TOF     = 100% - loss due to sub-optimal tilt and orientation
TSRF    = Shading x TOF

Its incentive threshold is a lowest TSRF of 75% for an on-site evaluation, or an average TSRF of 80% across the plane of the array for a remote shade evaluation (Program Guide for Solar Trade Allies, Part 1, last updated July 2026, read 21 August 2026). You do not have to live in Oregon for the number to be useful. Ask which figure the model used and ask for the shade report it came from. A model run at 0.95 on a roof that measures 0.82 overstates production by 0.95 divided by 0.82, about 16%, before anything else is wrong.

Items 5 and 6: the cash price, and the price you would actually be financing

5. Total cash price, and the price per watt you compute yourself. $25,200 over 8,400 DC watts is $3.00 per watt. Do the division rather than reading the seller's figure, because item 1 decides the denominator. Then check what sits inside the price: a main service panel upgrade, roof repair, trenching, and battery hardware each belong in or out of the number explicitly.

6. Financing terms, and the gap between loan principal and cash price. This is the largest hidden number in residential solar. The CFPB's August 2024 issue spotlight found that lenders bake fees into the loan principal that typically run 10 to 30 percent of the cash price and can exceed 50 percent, that these fees are not included in the total cost of credit presented to the consumer, and that salespeople often do not explain the difference between the cash price and the loan principal at all (Issue Spotlight: Solar Financing, read 21 August 2026). Its worked example is a $30,000 cash price carrying a $9,000 fee.

Apply that same 30% to the system above:

cash price                      $25,200
dealer fee at 30%             +  $7,560
loan principal                  $32,760

$32,760 at 3.99% for 300 months   <- illustrative loan terms
  monthly payment              $172.74
  total repaid                 $51,822
  principal per DC watt          $3.90   (cash price was $3.00)
  total repaid per DC watt       $6.17

Year-one savings of $2,533.51 work out to $211.13 a month, which is more than $172.74. The pitch is arithmetically true. The system still costs $51,822 instead of $25,200. The CSLB Supporting Information form has separate fields for the name of the financing company, term in years and months, interest rate, monthly payment and total cost, plus a yes/no field asking whether the contractor is facilitating the solar loan. Ask for the cash price in writing before any financing conversation starts.

Items 7 to 9: the three assumptions that bend the savings line

7. The utility rate escalation assumption. California caps what a seller may assume at 10 percent per year, and the CPUC's consumer guide tells homeowners to ask which rate was used (California Solar Consumer Protection Guide, read 21 August 2026). A cap that high is not a benchmark. Measure it instead. EIA's Electric Power Monthly Table 5.3 puts the U.S. average residential price at 12.55 cents per kWh in 2016 and 17.30 cents in 2025, a compound annual rate of 3.63% (Table 5.3, data through May 2026, read 21 August 2026).

Then notice how much the window matters. The same table gives 1.71% a year across 2016 to 2021, and 5.64% a year across 2020 to 2025. A seller who fits the escalator to the second window rather than the first has not said anything false. Ask which years it was fitted to, and rerun the savings on your own state's series instead of the national one.

8. The annual degradation rate. Modules lose output slowly, and every proposal models that loss with a percentage. The number to test it against is not an industry average. It is the performance warranty in the module's own datasheet, which you can only locate once item 2 has given you a model number. Manufacturers publish a guaranteed output curve, usually a larger first-year step down followed by a fixed annual decline. If the proposal's degradation figure is more optimistic than what the manufacturer is willing to sign, the proposal is claiming something nobody has underwritten.

9. The tariff and export assumption, before and after. The CSLB form carries two adjacent fields: Rate schedule before install and Rate schedule after install. They are separate for a reason. Going solar frequently moves you onto a different rate plan, and under a net billing structure an exported kWh is not worth what an offset kWh is worth. Sizing interacts with this. California's net billing customers are limited to oversizing by no more than 150 percent of the past 12 months of usage without filing an attestation, and the CPUC guide notes that a system is typically sized to around 80 to 85 percent of the prior year's use. A proposal that values every generated kWh at one retail rate has skipped the export tariff entirely. Rebuilding the four numbers on your own bill settles what the offset kWh is worth; the export tariff is a separate document and has to be read separately.

Items 10 to 12: what the paperwork owes you after the crew leaves

10. Three warranties, named separately. Product, meaning the hardware is not defective. Performance, meaning the module still makes a stated percentage of rated output in year N. Workmanship, meaning the installer's own labour, including roof penetrations. They come from different parties and expire on different dates. The CSLB form does not ask for the terms at all. It asks the contractor to fill in a blank: Information about warranty coverage is in the contract on page: ______, with a second line doing the same for service and maintenance. Borrow that. Make the proposal point at a page number rather than at an adjective. The CPUC guide's question list adds the one people forget: if the inverter fails after its warranty period, what does replacement cost?

11. Removal and reinstall terms for future roof work. Panels have a design life north of 20 years and a roof covering may not match it. The CPUC guide tells homeowners to ask roughly what it will cost to remove and re-install the panels if the roof needs replacing, including inspection fees. Get that as a figure rather than a reassurance, and get it before signing rather than in year 12.

12. The cancellation window, written into the contract. California prints it on the cover. Business and Professions Code section 7169 requires the Solar Energy System Disclosure Document in boldface 16-point type on the front or cover page of the contract, and the form carries one filled-in line for the TOTAL COST for the solar energy system (including financing and energy / power cost if applicable) plus a Five-Day Right to Cancel for buyers 65 or older where the contract was not negotiated at the contractor's place of business (CSLB Solar Requirements and the disclosure form, both read 21 August 2026). The CPUC guide states the floor for everyone else in the state: at least three business days to cancel for any reason, five if you are 65 or older. The FTC's blunter version applies everywhere: never deal with a company that pressures you for a quick decision or tells you to sign without time to review (Solar Power for Your Home, read 21 August 2026).

If a battery is in the quote, two more fields on the CSLB form start to matter: battery capacity and minimum state of charge. Neither means much until you have sized the battery against your own interval data.

One proposal, three payback numbers

Take the 8.40 kW system. $25,200 cash, 11,900 kWh claimed, and an offset kWh worth 21.29 cents. That last input is not a national figure and is not offered as one: it is the July marginal rate rebuilt tier by tier from one Georgia Power R-31 bill, where the same bill's annual average came to 17.65 cents and its January marginal rate to 13.39. Substitute the marginal rate off your own sheet before any row below means anything. Year one saves 11,900 x $0.2129 = $2,533.51.

Cumulative savings over N years, where e is escalation and d is degradation, follow a geometric series with ratio r = (1 + e)(1 - d):

total(N) = 2533.51 x (r^N - 1) / (r - 1)

Three runs of one formula. The first keeps the dials the proposal set, 5.00% escalation and 0.25% degradation, which are sales-document inputs rather than measured quantities. The second replaces the escalator with the EIA-measured 3.63% and moves degradation to 0.5% — itself a placeholder, and one that stands only until item 8 sends you to the warranted curve in your own module's datasheet. The third additionally corrects production for a shade report showing TSRF 0.82 against a model run at 0.95, cutting year one to 10,272 kWh and $2,186.82.

Two things fall out of that table. Escalation and degradation together moved payback by less than six months while knocking $22,894 off the 25-year figure, so they are levers on the headline total rather than on the break-even year. The shading correction, one number that was never printed anywhere in the document, moved payback by 1.2 years on its own. And none of the three rows describes the financed version, where $51,822 leaves the account instead of $25,200.

Run the top row on your own proposal first. Divide claimed annual kWh by DC kilowatts. If the answer sits far from what your shade report and your tariff sheet support, the remaining eleven items tell you where the difference went.