Pytes V5 SOC stalls at 20% and 90% on every cycle (~20 Ah / ~30 Ah), "15%" is almost empty

DIY
, , , , ,
1

Hi all,

I’m seeing odd SOC reporting from my Pytes V5 bank and would like to know if others see the same, and whether a firmware update fixes it.

System

  • EasySolar-II GX + MultiPlus-II 48/3000 in parallel (VE.Bus fw 556), Venus OS v3.80
  • ESS, DVCC on, Pytes on CAN-bus (can-bus-bms)
  • 2× Pytes V5 (200 Ah total), battery fw shown on the GX as v24.1 (= 1.10.24)
  • SOH reported 100%, cells balanced (~40 mV spread even at the bottom)

What happens

On discharge the SOC freezes at 20% until the lowest cell drops to ~3.17 V. On charge it freezes at 90% until the cells reach ~3.45 V, then jumps to 93%. A normal 1% step is ~2 Ah; these holds are 10–15× that. Data from VRM (1-min log), 18–25 Sep:

Date / time SOC Held for Ah while held Cells (min/max) at start → release
18.09 04:00 20% (dis) 89 min 19.5 Ah out 3.21/3.22 → 3.17/3.18 V
20.09 03:08 20% (dis) 81 min 20.5 Ah out 3.20/3.22 → 3.17/3.17 V
21.09 05:19 20% (dis) 98 min 21.0 Ah out 3.21/3.22 → 3.17/3.18 V
25.09 02:33 20% (dis) 74 min 19.8 Ah out 3.22/3.23 → 3.17/3.18 V
19.09 12:10 90% (chg) 29 min 31.5 Ah in 3.39/3.40 → 3.45/3.47 V, then 93%
20.09 12:03 90% (chg) 24 min 31.2 Ah in 3.38/3.40 → 3.45/3.47 V, then 93%
24.09 14:33 90% (chg) 65 min 30.3 Ah in 3.36/3.37 → 3.44/3.45 V, then 93%

Discharge current during the holds was ~15–19 A (~0.1C), charge current 50–70 A.

Capacity itself looks fine. From 100% (cells at 3.54 V) down to the ESS minimum (“15%”) the bank delivered 182 Ah / 9.5 kWh (19→20 Sep) and 180 Ah / 9.4 kWh (20→21 Sep).

The practical problem is the bottom end. At “15%” the cells rest at 3.18–3.19 V. This morning the grid disconnected while the SOC was at “15%”. The MultiPlus drew only ~5 Ah (~18 A for 17 min) before it shut down on low battery at 49.0 V (cells 3.04–3.08 V under load). During those 5 Ah the SOC dropped 15% → 3%. So the ESS “15%” reserve was really only a few percent.

Also, on days without a full charge the 20% hold doesn’t appear and the SOC drops quickly instead. Between “21%” and “15%” I measured anywhere from 4 Ah to 32 Ah depending on the day. That makes the ESS minimum SOC an unreliable reserve.

Questions

  1. Is this the expected behaviour of the Pytes BMS (voltage-anchored SOC correction at 20% / 90%)?
  2. Has anyone seen this improve with newer V5 firmware (1.10.27 / 1.10.29)?
  3. Until then, what min SOC would you recommend with this BMS, so the ESS reserve is actually there when the grid drops?

Thanks!

4

It seems to be quite normal. Pylontech does show the same behaviour (at 90 / 11%). How long these plateaus will last depends on how long your Battery was in the “free-flight-zone” between the voltage-anchors.

Each module seems to be quite different in guessing its soc. The longer they guess, the bigger the deviation is. This causes single modules to hit the 90%/20% borders way earlier then, and the overall system soc then stucks way longer, until all modules pass the respective voltage threshold.

The combined soc the bms reports is not just the average of all socs, it seems to be weighted, tending towards the more extreme 90/20 values. Thus, bigger deviations make the system soc stuck longer at this point.

Here for example on my pylontechs (16 Modules):

image
image972×462 44.8 KB

image
image490×264 10.7 KB

But if you look at the cell voltages during that times, you can clearly confirm that its just “false-positive” soc-deviations that build up and all modules pass the same min/max cell-voltages quite simultaneously.

That is the most important conclusion I made from looking into details here as well: Especially along with DESS, that mens: If I would use a min soc of 15%, the system may run out of energy on a close schedule - because we just did not really discharge 85% of energy but less.

Seeing it being able to discharge another full hour, staling at 11%, running with 10% min soc now, when needed :wink: