Right, a quick follow-up on my journey to checking I'm not mugging myself by buying a big battery.
Following advice from members here I thought I would change the heat pump settings to see how much benefit I can get from time-shifting consumption just by doing heating and hot water at mostly-cheap rates. The change happened on 7/4/2024. I basically set heating and hot water timings to be at cheap overnight rates and avoiding 16:30 to 20:00. I also slashed the flow temperature.
Without too much detail, I compared consumption per half hour from February to the change date, with consumption both from the change date to the end of April, and to the end of May. Obviously consumption is reducing a lot from May onwards and though the rest of April was fairly cool in Sheffield, average import per day went from 40kwh between February and 7/4/24, to 33kw from 7/4/24 to the end of April. It's just about comparable but it's more of a guide than a really scientific test.
Anyway, here's the findings for what happened after making the change to heat pump settings:
- Average import per half hour in April from 4:30PM to 8PM decreased a lot: between 30% and 80% per half hour slot. Still importing on average 3KW a day in this period so there's scope to time-shift more of that with a battery
- Average import per day during the high-ish morning rates (7AM - 9AM usually) didn't change at all - average import was still 3.4kwh in this period. So again, a lot of scope to time-shift this with a battery
- Import in the cheap tariff periods of 1PM to 2:30PM (often negative pricing when it's windy) actually reduced, partly because of more solar consumption as the sun came out. Obviously this is bad, it's a big missed opportunity to load up on cheap electrons
- And here's the kicker: rates in the evening are often high-ish after 8PM although much cheaper than the peak. Import was down about 30% - but on average I'm still importing 9kwh in this period. If I can fill up a battery at lunch time I can save a lot 4-8p a kwh in the evening, if it's big enough.
Obvious caveat: this is slightly rough analysis with some loose numbers. Even so, I think it's instructive. In April there's a lot of arbitrage available: 3+kwh in the morning, another 3+kwh at peak rates, and another 9kwh with less benefit in the evening - a total of 15kwh that could be time shifted with a battery.
So a 10kwh battery with a fast enough charge / discharge rate would have enough capacity to cover most or all of this consumption. And in the depths of winter there's probably even more benefit.
Here's what I think I want to do - let me know if the logic is flawed
I'd want to use a battery with integration with the Octopus smart scheduler API in their R&D system, and I'd probably want a high charge / discharge rate of 6kw, so that probably means a Givenergy battery, preferably the AIO for the high charge rate. These are more expensive than the Pylontechs etc so payback would be longer (probably 7 years) but it's all feasible through a simple web portal, with a reasonable level of support available. I can't DIY
.
Thanks for the helpful advice as ever!