Does a better solar forecast save money with a battery?
A little. On a real 27 kWh battery, planning on a weather forecast instead of the last three days’ sun cost £2.26 less over 26 days of September 2026. After that, even a perfect forecast of the sun would have saved only £1.85 more over 56 days.
What these figures rest on: Orza’s plans replayed half hour by half hour on one real home’s readings (a 27 kWh battery and 13 kWp of panels in the North West of England), September to November 2026, with the prices Octopus published and each forecast as it was issued. How Orza is tested.
Most of the gain was one sunny day after cloudy ones.
A battery plan has to guess tomorrow’s sun. Guess too little and the battery fills from the grid overnight, then the sun arrives with nowhere to go. Guess too much and the evening runs short. A forecast only earns its keep on the days the weather changes.
The real home, September 2026
- 11 September: £4.37 down to 15pThe last three days said 9.5 kWh of sun. The forecast a day ahead said 29.1 kWh, and 33.3 kWh came.
- 10 September: £1.60 up to £4.06The forecast did worse here. The two days together were still £3.08 better with it.
- 17 to 29 September: about levelOn ordinary days, the last few days’ sun is already a fair guess.
Orza plans your battery every half hour around your roof’s sun forecast and your prices.
Free, with no card: an email and a password, and you look round an example home first. Connect your battery through its maker’s cloud when you’re ready, or never. It starts in Monitor only, and nothing is sent to it until you choose.
A forecast took £2.26 off £30.08.
The same 26 days of September 2026 replayed four ways, changing only what the plan was told about the sun. The forecast is the one that was issued before each day, adjusted to what this roof actually makes. Lower is better: it is what the home paid for its electricity, less what it was paid for export.
| The plan was given | What it is | 26 days |
|---|---|---|
| The last three days’ sun | No weather forecast at all | £30.08 |
| A weather forecast | As issued, adjusted to the roof | £27.82 |
| Perfect sun | The sun that actually came | £28.46 |
| Everything known | Every price and the sun, in hindsight | £27.22 |
Measured on 30 September 2026. Forecasts from before 17 September came from the forecaster’s own archive of runs as issued, so no day used weather it could not have known. The home was on Octopus Agile.
- The gap to the best possible fell from £2.86 to 60pMeasured against a plan that knew every price and the sun in hindsight.
- Perfect sun did not beat the forecast£28.46 against £27.82, because how the home would use power and the prices not yet published were still guesses.
- One home is one homeA roof that faces another way, or a sky that changes more often, will give its own figure.
With a forecast in, perfect sun was worth £1.85 in 56 days.
Once the plan used a forecast, the same home’s 56 days to 2 November 2026 were replayed again with each guess swapped for what really happened. Even knowing all three perfectly would have saved £7.68: about 14p a day, £50 a year. That is the most any better forecast could ever win on these days.
Saved more, over 56 days, if it had known
- The sun: £1.85The most a sun forecast, paid or free, could have added on top.
- How the home used power: £2.78The largest single part, more than the sun.
- Prices not yet published: £1.48Octopus publishes the next day’s Agile prices each afternoon, so a plan made before then guesses the rest.
- All three: £7.68More than the three added up, because the guesses work together.
Measured on 3 November 2026, over 56 days to 2 November: Octopus Agile to 17 October, then Flux, with the export tariffs that went with them. Each line is what the plan would have saved beyond what it did.
Before you pay for a forecast.
Two months of one home are evidence, not a promise. These are the questions that come next.
All our guidesGet more from your battery
How Orza is tested
Is a paid solar forecast worth it for a home battery?
Weigh its price against what is left to win. On this home, once a free weather forecast was in use, knowing the sun perfectly would have saved £1.85 more over 56 days, and no forecast can do better than perfect.
What matters more than the forecast?
Planning the battery at all. In a year of modelled homes, the maker’s self use mode saved £170.99 and the battery planned by Orza every half hour £431.13. How the ways of running a battery compare.
Does the forecast matter more in winter?
These days ran from September to 2 November, so they don’t show a winter, when there is less sun to guess. What a battery saved month by month.
Should I charge from the grid before a dull day?
Only when a forecast for the day is low enough to be worth acting on. On 11 September the opposite happened: the last three days said 9.5 kWh, 33.3 kWh came, and a battery filled from the grid had nowhere to put the sun. A solar battery in summer.
Why do prices not yet published matter?
On Agile, tomorrow’s prices come out each afternoon, so a plan made in the morning guesses part of the day ahead. On these 56 days, knowing them would have saved £1.48. How Agile’s prices ran over a month.
What does Orza cost?
Nothing right now: everything is free while we spread the word, and no card is needed to join. Paid tiers will come later, and there will always be a free version.
Plan for the sun that is coming, not the sun that came.
Free, with no card. Look round an example home first, and connect your battery when you’re ready.