Experimental PV calculator

[LesNewell]

I'm in the process of getting a solar installation and I got bored while waiting for the G99 to go through. To pass the time I thought I'd build a simulator to experiment with tariffs and battery options. It pulls in historical solar, tariff and usage data and simulates a solar installation at 30 minute increments.

It's quite interesting to see how different tariffs and battery capacities affect the payback. So far Octopus Flux seems to be the winner by a big margin. The simple strategy of buy as much as possible when it's cheap and sell as much as possible at peak rate seems to generate the best payback. After playing around with the numbers, 2 - 3 times the battery's maximum charge/discharge rate seems to be the best battery capacity.

My current planned system is 8kWp solar with a 6kW inverter (5kW battery charge/discharge rate) and around 9kWh usable from the battery (3x Pylontech US3000).
Here are the estimated battery payback figures for a few battery capacities with my setup on Octopus Flux at current rates:
5kWh: £330/year
10kWh: £590/year
15kWh: £800/year
20kWh: £863/year
25kWh: £872/year
This is just the effect of installing batteries and does not include installation cost or profit from solar generation.
From these figures it's pretty clear that around 10kWh usable is the sweet spot. Anything over 15 is pointless.

Interestingly solar generation has surprisingly little effect on battery payback. For instance with zero solar 10kWh pays back about £520. It all comes down to the difference between low rate import and peak rate export. Of course it's risky to base everything on one tariff. If that tariff goes away so does the payback.

If anyone wants to have a play with it I stuck it on Github. It's written in Python so it runs on Linux or Windows. For Windows users I also created a release package that you can run without needing to install Python. Any feedback would be welcome.

Does anyone have detailed consumption data, ideally for at least a year? In particular I'm looking for someone with air source heat. My current usage data is based on massaged figures from another property using storage heaters, combined with some extra guesses and estimates. My heat pumps haven't been in for long enough to gather meaningful data.

[Stinsy]

Have a you modelled intelligent Octopus? Charging the batteries overnight on 7.5p and using the power during the day…

[richbee]

My feeling is about the same, based on 6 months of having only 1x 5kWh battery with Solar PV. There is a lot of time where the battery is full and exporting, but then the battery is empty the following morning. Also the discharge rate of 2.5kW is not enough to run things like a kettle fully.

In April I imported 140kWh and exported 188kWh - which isn't great, with 71% of my electricity coming from Solar.

I have now fitted a 2nd 5kWh battery and it seems to have made a huge difference - in May so far I have inly imported <10kWh and usage is well over 90%, even though we have had less PV generation on average compared to April.

My feeling is that adding a 3rd battery wouldn't make so much difference - but I think it will also depend on the house usage - we are on around 13kWh average load. If you have ASHP and/or BEV with much higher usage, I would imagine that more batteries might be worth it.

I haven't setup any Octopus off peak tariff yet, but I want to for winter usage - I reckon (not backed up by any actual calculations), that having the 2 batteries would make it worthwhile to fill the batteries overnight

[LesNewell]

Have a you modelled intelligent Octopus? Charging the batteries overnight on 7.5p and using the power during the day…

I just ran a sim using intelligent with my setup. As I don't have an EV, Flux is better.
Total yearly bill on Intelligent import, 15p fixed export is £-588 (i.e they pay me £588)
Total yearly bill on Flux: £-720

If you have an EV the balance probably changes. As the low rate is below the solar export rate EV charging should only be done on low rate.
Currently the sim doesn't handle EV loads properly. I will need to think on that one.

To get anything approaching realistic numbers you need to run a sim based on your figures. My usage may be very different to yours.

[LesNewell]

Richbee, depending on your tariff optimising for minimum import may not be the best strategy. For example with my flux simulations, I have a very high import and export as I'm effectively trading electricity between low rate and peak rate. This is hard on the battery though - around 400 cycles/year.

What tariff are you on at the moment? If you are on a single rate tariff you may be leaving a lot on the table.

[richbee]

Richbee, depending on your tariff optimising for minimum import may not be the best strategy. For example with my flux simulations, I have a very high import and export as I'm effectively trading electricity between low rate and peak rate. This is hard on the battery though - around 400 cycles/year.

What tariff are you on at the moment? If you are on a single rate tariff you may be leaving a lot on the table.

I am currently on a single tariff with SoEnergy - who are not great!
My idea is to move to Octopus and swap onto one of their 'economy 7' type tariffs - Cosy or Go probably - It shouldn't make much difference over the summer as we generate almost everything we need, but come autumn and winter I want to be charging the batteries on off-peak and using them during the day.

[nowty]

My current planned system is 8kWp solar with a 6kW inverter (5kW battery charge/discharge rate) and around 9kWh usable from the battery (3x Pylontech US3000).
Here are the estimated battery payback figures for a few battery capacities with my setup on Octopus Flux at current rates:
5kWh: £330/year
10kWh: £590/year
15kWh: £800/year
20kWh: £863/year
25kWh: £872/year
This is just the effect of installing batteries and does not include installation cost or profit from solar generation.
From these figures it's pretty clear that around 10kWh usable is the sweet spot. Anything over 15 is pointless.

Whilst financially your probably right, your view may change when you've been up and running for a few months.

Have you put down any cash on your system yet ?

[LesNewell]

Whilst financially your probably right, your view may change when you've been up and running for a few months.

I'm a tight arse so I'll probably just go with the numbers ... probably I'm pretty sure I'm not going to rival your setup any time soon!
It will be interesting to see how close my simulations are to reality.

Have you put down any cash on your system yet ?

Not yet. The G99 application has been in for about a month and I have accepted the quote from my installer. As soon as the application comes through it will be all hands on deck.

[Caesium]

This is hard on the battery though - around 400 cycles/year.

I would like to suggest this is less of a concern than some people think.

The LiFePO4 cells are rated for 6000 cycles or so, and even after doing that, they should still operate at 80% of their full capacity.

6000/400 = 15 years. By that point, we will almost certainly have an order of magnitude better battery technology and we'll be onto the next big thing. Even if we don't, there's also calendar ageing of these cells to consider, which is put at 20 years or so. Although I can't remember offhand what is expected to happen after 20 years, probably just a gradual decline of capacity again - so they're still useful after that.

In short, I've come around to the way of thinking of actually put the cells to use and get value out of them while they're good, don't baby them. I wouldn't call your 400 cycles/year hard, I'd say thats a pretty easy life

[LesNewell]

I would like to suggest this is less of a concern than some people think.

That's pretty much the conclusion I came to, hence going for the strategy of buying as much as possible then reselling it.