Buying
Reading a solar proposal: eleven numbers that decide whether it is honest
Most proposals are designed to be skimmed. Here is what to look for line by line, including the figures that are quietly missing.

We are regularly asked to review competitors' proposals. It is not a comfortable position and we try to be fair about it, but there are patterns, and the patterns are consistent enough to write down.
A good proposal is checkable. Every headline number should be reconstructable from other numbers on the same page. When it is not, that is usually deliberate.
1. System size in kW DC
The sum of module nameplate watts. Should be stated plainly, not only as a module count.
2. Module count and wattage
These two multiplied must equal the system size. It sounds trivial. It does not always reconcile.
3. Inverter AC rating and DC to AC ratio
A ratio between roughly 1.15 and 1.3 is normal. Above 1.4 you are clipping real production on clear days. Below 1.05 you are paying for inverter capacity you will never use.
4. Modelled year one production in kWh
Divide it by system size. If the result is far above about 1,400 kWh per kW for this region, ask what assumptions produced it. Optimistic yield is the single easiest way to shorten a payback figure on paper.
5. Shading assumption
Look for total solar resource fraction per roof plane, or at least a shade loss percentage. A proposal with no shading figure on a wooded lot has not modelled your roof, it has modelled a roof.
6. Degradation rate
Should appear in any model that runs past year one. A rate of 0.4 to 0.55 percent per year is typical. A model that ignores degradation overstates 25 year output by roughly 5 to 6 percent.
7. Utility rate escalation
The assumed annual increase in what you pay per kWh. Around 2 to 3 percent is defensible. We have seen 6 percent used, which roughly doubles the modelled 25 year saving and is not a forecast, it is a wish.
Escalation is the most abused input in the industry. Ask what rate was used. If the answer is vague, the number it produced is worthless.
8. Export treatment
How surplus generation is valued. If the model assumes every exported kilowatt hour is worth your full retail rate, check that against your actual rider. Export credited below retail changes the economics of oversizing substantially.
9. Net cost and how the credit is treated
The contract price, then the credit shown separately, then the net. A proposal that shows only the net price is hiding what you are actually signing for. And the credit is a reduction in tax owed, not a discount applied to the invoice.
10. Payback arithmetic, shown
Net cost divided by first year saving gives a simple payback. A proper model runs cumulative cash flow year by year with escalation and degradation applied. Both should reconcile within a year or so. When the headline payback is far shorter than net cost divided by first year saving, something in between has been assumed generously.
11. Workmanship warranty term
Equipment warranties come from manufacturers and are broadly similar between installers. The workmanship warranty is the installer's own promise about the holes they put in your roof. Ten years is common. Twenty five is meaningfully better. This is the number that distinguishes companies, and it is usually in the smallest type on the page.
What is often missing entirely
- Roof age and remaining shingle life, which decides whether the array will need removing in eight years.
- Service panel condition and whether an upgrade is included or excluded.
- Interconnection timeline, which is outside the installer's control but should still be stated.
- Whether the battery figure is usable or nameplate capacity. Those differ, sometimes by 10 percent or more.
- What happens to the production guarantee if you sell the house.
Compare like with like
The single most useful comparison across quotes is dollars per modelled kilowatt hour over twenty five years, not dollars per watt. Dollars per watt rewards cheap modules on a shaded plane. Dollars per modelled kilowatt hour rewards a design that actually produces.
Work it out yourself: net cost, divided by the twenty five year production figure after degradation. Do it for every quote. The ranking frequently changes.
More notes
Related reading
kW or kWh: the unit mistake that ruins solar quotes
Rate and quantity are different things. Mixing them up is how people end up with a battery that cannot start their well pump.
Read itStorageWhat one battery actually runs during a Triangle ice storm
We pulled telemetry from eleven of our storage systems through a 2026 winter outage. Here is what the numbers show.
Read itDesignShade is a number, not an opinion
Total solar resource fraction turns a conversation about trees into arithmetic you can check. Here is how we measure it and when we say no.
Read it