One month with a solar battery - real statistics

August is meant to be full of gloriously hot days. An endless parade of sunshine and drinks in the park. This year it seemed mostly grey, miserable, and prone to pissing it down at a moment's notice.

We all know that solar panels' efficiency wilts in the heat, but do they get a tan work standing in the English rain?

At the beginning of August we installed a 4.8kWh solar battery to supplement our 5kW of solar panels.

The battery provides a CSV of readings taken every 15 minutes. It measures solar power, household usage, and battery usage. August, looked like this:

A graph of the month covered in lines showing solar power and electricity usage.

Not massively helpful. But, with a little bit of 🐍 Python and 🐼 Pandas, I worked out the following:

🏠 Our home used 290kWh

We're tracking pretty close to the UK average of about 10kWh per day. Our average of 9.4kWh each day is perhaps slightly higher than normal for a 2-person household. But we work from home regularly and have a lot of hungry smarthome gadgets.

🔌 Only 15% of our electricity came from the grid

The sun doesn't shine at night. Duh! But the battery usually provides most of our power after sunset. The battery can only discharge at a maximum of 2.4kW, I think. So if we use the electric shower, oven, or other high power appliances, then we draw from the grid.

So, what did that 44kWh cost us?

💷 Normally, we'd pay £90 for August's electricity. We only paid £14!

The price you pay for electricity depends on where you are in the UK and what tariff you're on. With a mix of solar and battery, we cut our August bill by 85%.

But what's the mix between solar direct and solar delayed?

🌞 Solar gave us 46% of our electricity

About 135kWh of our month electricity needs was met directly from solar. That means photons hit the panels, they bounced down into the inverter, and then straight into the wires in our walls, where they were gobbled up by laptops, TVs, and toasters.

That just leaves the battery…

🔋 Battery storage gave us 39% of our electricity

We used about 113kWh from stored solar. An average of 3.6kWh per day. Perhaps that means our 4.8kWh battery is over specced? I'm not so sure. Some days we use over 100% of our solar battery capacity.

Of course, not all of the solar power can get used or stored. Once the battery is full, that electricity has to go somewhere…

🔙 We sold 140kWh of solar back to the grid

Our solar power feeds into our local grid for our neighbours to use. We sell the electricity at market rates - which change every 30 minutes. This made us £13.

📈 Total cost for August's electricity? £1.

Yup! For the whole month of August, our electricity bill was £1.

(Plus the standing charge, of course!)


OK, time for a little bit of a sanity check.

Firstly, these data are drawn directly from the battery. It has clamps over our import and export wires to monitor what the household is doing. These are broadly accurate - I estimate less than 2% different from what our smart meter and inverter report.

Secondly, the battery groups up the stats every 15 minutes. So, again, that's likely to introduce some errors into the data.

Thirdly, prices for both import and export can vary massively. Our export price in particular varies depending on demand.

Fourthly, these data were gathered in South-East London on an East / West split solar site. Your panels will be in a different location and will perform differently.

Fifthly, the price of panels and battery storage is high. If you can afford the up-front capital costs of an installation, I think it makes sense to do so. The payback period is usually under 10 years. But can be considerably shorter during a time of rising energy costs.

I want more stats!

Every day at sunset, my solar panels publish their generation stats to GitLab. You can download all the data from 2020 and see how much solar generation we've had.

If you need more, I published 5 years of minute-by-minute solar generation as Open Data from our previous house. This dataset has been cited in several academic papers.

I'm considering whether to release my daily usage statistics. At the moment, it feels a little invasive. You can tell when I put the kettle on in the morning, see when I load a tumble-dryer, and calculate just how long I use the oven for. Perhaps you can even analyse the overnight fluctuations and work out what model of fridge I have. I don't think you can tell what video content I'm watching because it's hidden in the noise of my other appliances. But you could probably tell if I was home or not.

Here's a typical daily graph. What do you think you can figure out from this?

Graph with multiple lines. There's a spike about 6AM which is probably a kettle being boiled. Another near lunchtime which might be a microwave. The evening has a couple of hours of high use - which is probably a washing machine.

So I think I'll release it in a year's time. That's a decent balance between openness and privacy.

I hope you've found this blog post useful. If you have, you can support me by:

  • Switching to Octopus Energy - if you join, we both get £50. They do dynamic pricing for import and export. And, even better, they have an API so you can query your energy usage.
  • Supporting OpenBenches - it's a crowdsourced site of memorial benches run by me and my wife.
  • You can also buy me a book to read.


Share this post on…

6 thoughts on “One month with a solar battery - real statistics”

  1. said on

    Ta. This has encouraged me to do similar analysis.

    We have a small #SolarPanels installation (6 panels) facing a mixture of east and south, in #Bristol, UK. No battery.

    August 2023.
    Solar generated: 183 kWh.
    Exported to grid: 121 kWh.
    Imported from grid: 93 kWh.

    I have a standard fixed tariff for both incoming and outgoing power, so I think that the net electricity cost to me is around £8 for August.

    Reply | Reply to original comment on
  2. said on

    Thanks for your informative solar posts, they've inspired me to experiment with an Anker Powerhouse 767 & a 200w Renogy solar panel, adding another soon. I'm just powering my home lab & networking. This itches a scratch instead of going whole house, but also teaches many interesting skills. I'm trying to get data about power usage & solar generation, so it's great to see there are some tolerances to the stats, which I was concerned about as I'll be rolling my own data collection. Thanks!

    Reply | Reply to original comment on
  3. Tommy Long says:

    Your returns are similar to mine. Has anybody been able to calculate how many years it will take to recover their outlay? Octopus Flux calculates your usage based on the average amount of Kw used across the 3 consumption rates; the Exported income is calculated in the same way. However, all of this is distorted by the extortionate standing charge rate. I have derived that both my gas and electricity usage, from the grid, has halved in my first year. I have a 4.8Kwh system, both panels and battery. Is there a mathematician who can help?

    1. @edent says:

      You're going to be paying a standing charge no matter which provider you go with - so you can ignore that for your calculations.

      I've made the simplified assumption that I'm saving 35p for every kWh which passes through the battery. Perhaps sometime is costs more or less, but it will do as a simplified calculation. After a year, add up the total use.

      For solar and export, it is slightly trickier. I wrote a post about how to calculate solar savings.


What are your reckons?

All comments are moderated and may not be published immediately. Your email address will not be published.Allowed HTML: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong> <p> <pre> <br> <img src="" alt="" title="" srcset="">