@mk4 Charge to 100%, discharge down to 10%.

Posted by: @bobflux

↑

@mk4 Charge to 100%, discharge down to 10%.

Since you seem to quite thoroughly investigate every step you take, how did you choose these levels? I am trying to understand what are the optimal levels to preserve battery health and lifespan.
Sungrow (our battery) documentation suggests a 10%-90% range. We are able to fully charge our 9.6kwh almost every day and now that we have the ASHP, we can also easily drain it on a daily basis. Thank you.

Pylontech US5000 stores about 4.3 kWh accounting for inverter losses and costs 1000€. Over 3000 cycles this means one kWh going through the battery costs 7.7c ; half that for 6000 cycles. Much cheaper than drawing from the grid (about 0.20€/kWh) even if I run them into the ground they will pay for themselves nicely. And the more I use them, the more they pay for themselves.

I did some research, but unlike NMC which are special delicate snowflakes, LFPs just don't seem to care about anything except heat (the most important), the usual overcharge/discharge (which the BMS should prevent) and cycling near full charge.

There's this study which says "Cycling near the top of charge (75%–100% SOC) is detrimental to LFP/graphite cells." However, charging the battery instead of using the energy directly is also detrimental to the bottom line because the inverters have 10-15% losses on the roundtrip... So using the energy in real time is both more efficient and saves wear on batteries. In other words, priority to the EV charger and other stuff, battery charging comes last.

There seems to be a linear relationship between total energy cycled and aging, no matter the SOC span of the cycles, except the point mentioned above. So with half-cycles, they will last twice as many cycles... but if I only use half the capacity, then I need to buy 2x as many batteries, so why bother?

They report losing 1% SOH per year which is in line with manufacturer specs. With lots of care and cuddling, the cells will probably last longer than the BMS lol.

Pylontech says do not go below 5% SOC so I set it to 10% to be safe. I will probably add 1-2 more and set it to 30% to have some juice in case of blackout.

I set max SOC to 100% so the BMS can do its balancing. They do age faster at 100% but they don't stay there for long...

I put them in the cellar, the internal sensor reports temperatures around 20°C idle to 30°C at end of charge.

Max charge power is about 10kW so 0.5C, but this only occurs if nothing else wants power. If the car is plugged in, or there are other loads running, home batteries will of course charge slower or just wait for their turn.

@bobflux 

Thank you for the detailed explanation!

@batpred 

I just came across an article stating that things are moving with plug in panels in the UK! The source is Bloomberg. The link has the english translation of said article.

https://en.protothema.gr/2026/03/26/britain-introduces-plug-in-solar-panels-for-homes-to-tackle-the-energy-crisis-caused-by-the-war-in-iran/

Installed in May 2011: 16*Sharp 245 W monocrystalline panels in 2 strings feeding 2* Sunny Boy 1700 Inverters, together with new Main Consumer Unit (Fuseboard). The 2 strings were used to minimise a shadowing problem.
Since then the system has generated  33726kWh, with the panels having to be re-wired once due to a pigeon problem.
This power generation has given an income of £19,213 from the UKGov FIT scheme as well, of course, as the savings on grid power usage.

3.84kWp of solar on our main South-East facing roof, installed back in 2020. Fitted with Tigo optimisers to account for the neighbours 150 year old oak tree that gets in the way!

Recently added 1kW of solar to our south-facing shed connected to an Ecoflow Stream.  This allows us to catch more of the late afternoon/evening sun.  Not plug in / balcony - I got a friendly electrician to wire in a fused spur and change the RCB to bi-directional.  Self-installed the panels to keep costs down.

12kWh of Pylontech batteries and a Sofar Hybrid inverter complete the set-up.