Solis AC Coupled Inverter Pylontech batteries charging profile pauses for 30+ mins at 90%?

[agsolar]

Ive been trying to optimise the charge/discharge process with a cheaper overnight TOU tariff. 00:00 - 07:00.

3 x Pylontech batts US5000C @4.8KWH. 3KW inverter.
I want to charge to 100% and have my Overdischarge at 10%.

My desire is to minimise the time the batts are charging in the cheap TOU. Obviously 7 hours gives plenty of time to charge, but I reckon they should be ready in about 5 hours (=about 13.5KWh usable / about 2.5KW charge power). My charge current is limited to 52A before I get grid interference error messages.

With a 6 or 7 hour charging window, I notice that the SOC pauses at about 90% for 30-45 minutes before ‘topping off’ to 100% in the last hour. But according to the Solis graphs, energy is still seemingly being pumped in during that pause.

Is this the BMS managing the charge for the last 10%? I could understand if the power was backed off in this phase if so, but it still seems to be going in at full strength, so whats happening to it?

If I reduce the charging window to the expected 5 hours (+/- 30 mins) I don’t get about 90%.

 

[SeaGal]

My thoughts / questions...

My desire is to minimise the time the batts are charging in the cheap TOU.

Why, exactly?

Obviously 7 hours gives plenty of time to charge, but I reckon they should be ready in about 5 hours (=about 13.5KWh usable / about 2.5KW charge power).

It would be normal for a battery's BMS to reduce charge current as the SOC nears 100%, possibly temperature dependent too.

My charge current is limited to 52A before I get grid interference error messages.

Not familiar with that one... what is the exact error message you get? Is this on the Solis screen?

With a 6 or 7 hour charging window, I notice that the SOC pauses at about 90% for 30-45 minutes before ‘topping off’ to 100% in the last hour. But according to the Solis graphs, energy is still seemingly being pumped in during that pause.

Is this the BMS managing the charge for the last 10%? I could understand if the power was backed off in this phase if so, but it still seems to be going in at full strength, so whats happening to it?

The BMS will always 'manage' the charge - it sends the Solis the charge current request in a CANBus message every second. I heard of Pylontech's showing a "pause" before, but around 98%, not 90%.

I have a Solis, but not Pylontechs.. I suspect the pause is just the Pylon's BMS internally managing the SOC vs. cell voltages to get itself re-synchronised to 100% when it is fully charged.

If I reduce the charging window to the expected 5 hours (+/- 30 mins) I don’t get about 90%.

That sounds as expected to me.

View attachment 192807

From the graph, I don't see the sudden jump from 90% to 100% you mentioned?

Also, I don't see why battery would start discharging before 07:00 - your system appears to start using battery after 06:00 which is a waste of energy if that is still within your cheap economy 7 time period.

So, it's unclear to me as the graph doesn't seem to match your words? Sorry if I missed something

 

[SeaGal]

Further thought... my posting here (and in fact that whole thread) may be of interest.

Solis Hybrid Inverter Charging Parameters with a Pylontech battery

I've been doing a deep dive into my pylontech battery (three US5000 connected in parallel to a solis RHI-6K-48ES-5G-DC hybrid inverter). I had always assumed that the pylontech battery would communicate the correct charge voltage settings to the inverter through the CAN bus. But I am now not...

diysolarforum.com

 

[agsolar]

My thoughts / questions...


Why, exactly?


It would be normal for a battery's BMS to reduce charge current as the SOC nears 100%, possibly temperature dependent too.


Not familiar with that one... what is the exact error message you get? Is this on the Solis screen?


The BMS will always 'manage' the charge - it sends the Solis the charge current request in a CANBus message every second. I heard of Pylontech's showing a "pause" before, but around 98%, not 90%.

I have a Solis, but not Pylontechs.. I suspect the pause is just the Pylon's BMS internally managing the SOC vs. cell voltages to get itself re-synchronised to 100% when it is fully charged.


That sounds as expected to me.


From the graph, I don't see the sudden jump from 90% to 100% you mentioned?

Also, I don't see why battery would start discharging before 07:00 - your system appears to start using battery after 06:00 which is a waste of energy if that is still within your cheap economy 7 time period.

So, it's unclear to me as the graph doesn't seem to match your words? Sorry if I missed something

My thoughts / questions...


Why, exactly?


It would be normal for a battery's BMS to reduce charge current as the SOC nears 100%, possibly temperature dependent too.


Not familiar with that one... what is the exact error message you get? Is this on the Solis screen?


The BMS will always 'manage' the charge - it sends the Solis the charge current request in a CANBus message every second. I heard of Pylontech's showing a "pause" before, but around 98%, not 90%.

I have a Solis, but not Pylontechs.. I suspect the pause is just the Pylon's BMS internally managing the SOC vs. cell voltages to get itself re-synchronised to 100% when it is fully charged.


That sounds as expected to me.


From the graph, I don't see the sudden jump from 90% to 100% you mentioned?

Also, I don't see why battery would start discharging before 07:00 - your system appears to start using battery after 06:00 which is a waste of energy if that is still within your cheap economy 7 time period.

So, it's unclear to me as the graph doesn't seem to match your words? Sorry if I missed something

Why minimise the charge time? Its probably an OCD thing, but is related to your later question about why discharging before 07:00 - I only want to charge them to capacity as soon as possible, then set a discharge power of 0 between then and 07:00 (not yet implemented). But because the time of full charge was not as expected, my only option is to set the start-of-0-discharge to say, 06:00 rather than 05:00

GRID-INTF alarm - this occurs when my charge current is set to more than 52A - I suspect something to do with the 3KW limit being reached and/or the exisiting mains background load adding to the total detected load >3KW. https://usservice.solisinverters.com/support/solutions/articles/73000560430-grid-intf-alarm talks about harmonics on the line or heavy plant; none of which apply. The error is generated in the Solis log, the inverter backs off the charge for a few minutes, then applies it again for a few minutes again and so on until the charge period is ended. c53V * c52A = c2750W ?
TBH, unless Ive missed something, Ive accepted it as a non-issue as 52A works.

Ive added the inverter Temp to the graph - it shows the temp of the invrter increasing steeply during the pause period, which reinforces to me that the power-still-being-used is just being wasted as heat.

Sudden jump question - can you see the increase at the end of the flat period, second arrow? Thats 10% in 20 minutes..... I wouldnt say 'steep' but its really the question of why the flat period is there.
I can actually see there is a flattening around 98% which would reflect your BMS comment.

Im also pulling data directly from the batteries, and it reflects the same thing, apart from the battery temps - they remain +/- 3 degrees; so it feels like the inverter side.

 

[SeaGal]

If you set the charge period from 00:00 to 07:00 won't that stop it discharging until 07:00?

Ive accepted it as a non-issue as 52A works.

Yes, strange error - you wouldn't think they could let you set the charge current to more than the inverter can charge at!

Strange about the temp increase at that point of time - but if the charge current remains the same into the Pylontechs, presumably that's where the power is going... will mull that over.

Out of interest, how do you get the inverter temp?

 

[agsolar]

If you set the charge period from 00:00 to 07:00 won't that stop it discharging until 07:00?


Yes, strange error - you wouldn't think they could let you set the charge current to more than the inverter can charge at!

Strange about the temp increase at that point of time - but if the charge current remains the same into the Pylontechs, presumably that's where the power is going... will mull that over.

Out of interest, how do you get the inverter temp?

Inverter temp is the IGBT Chip Temp field in Soliscloud. I cant remember if you were using the Solax Canbus/not using Soliscloud, in which case (for my HomeAssistant) its just showing as Inverter Temperature


On setting the charge time to 07:00, will that cause any draw from the battery after its reached 100% and then charge it again/keep topping (i.e hysterisis) ?

 

[rpdom]

Your battery will be doing internal balancing somewhere between 90% and 95%. It's quite common for it to "stall" for some time and then jump to 100%. Thirty minutes seems slightly long, but possible.

Oddly, I've noticed something odd while monitoring my batteries (2xUS2000C) as they charge. They get to around 92-94%, then I can see they show they are balancing. The inverter seems to show the same SOC for a long time, but the actual batteries report the charge creeping up until all cells show 100%. I wonder if the BMS stops reporting to the inverter during that balance stage? @SeaGal any thoughts?

 

[agsolar]

Your battery will be doing internal balancing somewhere between 90% and 95%. It's quite common for it to "stall" for some time and then jump to 100%. Thirty minutes seems slightly long, but possible.

Oddly, I've noticed something odd while monitoring my batteries (2xUS2000C) as they charge. They get to around 92-94%, then I can see they show they are balancing. The inverter seems to show the same SOC for a long time, but the actual batteries report the charge creeping up until all cells show 100%. I wonder if the BMS stops reporting to the inverter during that balance stage? @SeaGal any thoughts?

I pulled the BMS SOC directly from the battery and added the graph above - it also stalls there as well.

 

[SeaGal]

Inverter temp is the IGBT Chip Temp field in Soliscloud.

Ah! OK, I don't send my data to China, so don't get that info

The inverter seems to show the same SOC for a long time, but the actual batteries report the charge creeping up until all cells show 100%. I wonder if the BMS stops reporting to the inverter during that balance stage? @SeaGal any thoughts?

What you are describing makes perfect sense to me - just with a little caveat that the BMS won't stop reporting the SOC to the inverter. If it did that, the Solis would trigger a CANBus failure message with a few seconds, causing the battery to go offline.

What I suspect is happening is pretty much what I've done in my own BMS to Solis comms software in order to allow the Solis to trickle charge the batteries to 100%, whilst giving the BMS time to do some balancing and get the battery and BMS to 100%. I had the issue that my JDB BMS would report the batteries had reached 100% (from its internal coulomb counting) before the cells got to a real 100% voltage (well, when I say 100% I actually mean the 3.45V that I have defined as being 100%). Obviously if I just have my software tell the inverter what the BMS says the SOC is, then once it hits the Solis's OverChargeSOC% value (which for me is set to 100%) then the Solis stops charging. So, what I do is, if the JBD reports 100%, but the highest cell's voltage is < 3.45V, then I my software just tells the Solis that the SOC is only 99%. It keeps doing that until I get one cell to 3.45V. Before then (in fact from 90% SOC) I start to reduce the charge current. This has the effect to trickle charging up to full charge, but if you were just monitoring the SOC as received by the Solis, my graphs would "stall" at 99% for a while. I am guessing that Pylontech does something similar.

 

[agsolar]

Ah! OK, I don't send my data to China, so don't get that info

Its on the CANBUS as well....

 

[jgh]

Mine also stalls at 90%. I think it's just lying because SOC is a guesstimate until it declares 100% from the
voltage peaking.

The BMS is still monitoring change during that period; watch the charge-current-limit - it steps down in stages.

 

[SeaGal]

Its on the CANBUS as well....

Wow - that sounds useful Do you have any details about this, e.g. which CANBus frames / data format etc., TIA

 

[SeaGal]

Do you have any details about this, e.g. which CANBus frames / data format etc., TIA

Any feedback, thanks!