Discount Rates, Energy Inflation and the Honest Payback Number

Every solar quote I have ever seen leads with one number: “payback in 7.4 years”. It sits there looking precise and friendly. It is also close to meaningless, because it hides three assumptions you never got to vote on: what electricity will cost in 2035, what your money could earn elsewhere, and whether your battery survives long enough to matter.

If you are searching for a solar panel payback period uk figure, here is my position: stop asking for one number. Ask for a range, with the assumptions printed next to it. Then use NPV and IRR to check whether the project beats the alternatives for your cash.

What simple payback gets wrong

Simple payback is cost divided by annual saving. Take a typical UK system: 4 kW of panels plus a 5 kWh battery, installed for £8,400. Say it saves £780 a year (self-consumption bill savings plus Smart Export Guarantee income).

£8,400 ÷ £780 = 10.8 years.

That calculation ignores timing completely. £780 in year 10 is treated as equal to £780 today. It ignores panel degradation (roughly 0.4% to 0.5% a year on decent modules). It ignores an inverter replacement, typically £600 to £1,000 somewhere around year 12 to 15. And it quietly assumes that your savings stay flat, or rise at whatever rate the salesperson had in mind.

Discount rates: what your money would otherwise do

A discount rate is the return you give up by spending £8,400 on panels instead of something else. For a UK homeowner the honest candidates are:

• Paying down a mortgage at, say, 4.5% (a guaranteed, tax-free “return”).
• A cash ISA or savings account at around 4%, before you consider that rates drift.
• A stocks and shares ISA, where long-run expectations of 5% to 7% are common but not guaranteed.

I use 4% as a base case for people with a mortgage or decent savings. Use your own. The point of the exercise is that a pound of saving in year 15 is worth less to you than a pound now, and NPV prices that in.

NPV: the number that tells you if it’s worth doing

Net present value discounts each year’s cash flow back to today. Positive NPV means the system beats your alternative. Negative means your money was better off elsewhere, even if it eventually “pays back”.

Here is a simple 25-year model for the system above. Assumptions: £780 saving in year 1, panel degradation 0.5% a year, inverter replacement of £800 in year 13, discount rate 4%.

I’ll run it three ways, changing only energy price inflation:

(Figures are from my own spreadsheet run on the assumptions above, rounded. Reproduce them rather than trusting me; the formulas are below.)

Look at the first row. With flat energy prices the system still “pays back” in about 11 years, yet it loses to a 4% savings account over 25 years. A single payback figure would have called that a success. NPV calls it a marginal miss.

Now look at the spread. Moving inflation from 0% to 6% swings the NPV by over £10,000 on an £8,400 purchase. That is the real story. The panels did not change. The forecast did.

Doing it in a spreadsheet

You do not need anything exotic. In Google Sheets or Excel, put annual net cash flows in column B, with year 0 (the negative install cost) in B1 and years 1 to 25 below.

Year 0 cash flow (B1):   -8400
Year n saving (B2..B26): =780 * (1-0.005)^(A2-1) * (1+infl)^(A2-1)
Year 13 adjustment:      subtract 800 for the inverter

NPV:   =B1 + NPV(0.04, B2:B26)
IRR:   =IRR(B1:B26)

Note that Excel and Sheets NPV() assumes the first value arrives at the end of period one, which is why year 0 sits outside the function. People get this wrong constantly, and it flatters every result by a year’s worth of discounting.

Put infl in a single named cell. Then build a one-way sensitivity table down the side: 0%, 2%, 4%, 6%. In Excel use Data > What-If Analysis > Data Table. In Sheets, just copy the formula block four times pointing at different inflation cells.

Where AI tools actually help

I would not ask ChatGPT, Claude or Gemini to “calculate my payback”. They will give you a confident single number, sometimes wrong, and you will not see the working. Use them instead for what they do well:

1. Building the model. Paste your quote and ask for a Google Sheets layout with named assumption cells, degradation, inverter replacement and a sensitivity table. Then check the formulas yourself.
2. Cleaning your data. Export half-hourly consumption from your smart meter via Octopus’s Home Mini or the n3rgy data portal, or pull it through the Octopus API, and have the model write a short Python script (pandas) to calculate how much of your usage falls inside daylight hours. That is your real self-consumption ceiling.
3. Stress-testing the quote. Ask it to list every assumption in the installer’s projection and flag the ones that are unstated. If the quote says “£950 a year saving” and does not say what unit rate it used, that is the first question to put to the installer.

For the quote-checking side in more depth, the Cost Models and AI Quote Analysis guide walks through how to pull those hidden assumptions out of a PDF.

Why energy price inflation is the variable that matters

Historic UK electricity prices are a terrible guide. The Ofgem price cap moved from about 21p/kWh in early 2021 to roughly 34p/kWh in late 2022, then drifted back down. Anyone who modelled 2022 as the new normal overshot. Anyone who modelled 2019 as normal under-called the following years.

So I do not pick one rate. I pick three and say why:

• Low (1%): policy costs shift off bills, wholesale prices soften as renewables grow.
• Central (3%): roughly in line with long-run inflation.
• High (5%): network costs and standing charges keep climbing and you keep buying more of your own electricity from the sun, which is valued at the full import rate.

Note one subtlety. Savings from self-consumed power track the import unit rate. Savings from exported power track the SEG rate, which for many tariffs sits between 4p and 15p/kWh and does not necessarily inflate at all. If you export a lot, blend the two inflation rates instead of applying one to the lot. Many installer calculators skip this.

What I would put on a quote

Instead of “payback 7.4 years”, write this:

System: 4 kW PV + 5 kWh battery, £8,400
Discount rate: 4%   Degradation: 0.5%/yr   Inverter: £800 in yr 13

Scenario    Inflation   Payback    NPV (25y)   IRR
Low         1%          10.7 yrs   +£300       4.6%
Central     3%          9.6 yrs    +£3,300     6.7%
High        5%          8.7 yrs    +£7,100     8.9%

That reads less reassuringly than a single year. Good. It tells the reader the project is probably worth doing, and it tells them what would have to go wrong for it not to be. A homeowner who sees “low case still roughly breaks even against savings” can make a decision. One who sees “7.4 years” has been handed a sales line.

A rule I now follow

If a payback figure arrives without a discount rate, an inflation assumption and a stated export share, I treat it as marketing. If it arrives with all three, I rebuild it myself in twenty minutes and see whether the installer’s number lands inside my range. Usually it lands at the optimistic end. Occasionally it is conservative, and that installer gets my attention.

Run your own numbers, publish your own range, and keep the spreadsheet. In five years you will want to compare your smart meter data against the forecast, and the version with the assumptions written down is the only one that lets you learn anything.