question

sharpener avatar image
sharpener asked

ESS refuses to make the coffee!

Hi all,

I am having trouble with my MP II 48/5000 ESS setup with 2 x Pylontech Force-L2 74 Ah modules.

The problem occurs first thing in the morning. Even if they still show over 20% SoC the batteries are not capable of powering a 2.2kW electric kettle to make the coffee. The battery stack rating is 75A "normal", 110A for 10 secs, so should be more than sufficient for this load.

I thought ESS had a kind of reverse Power Assist function (possibly something @Alexandra said?) whereby if the batteries were insufficient for the load then the Multi is supposed to draw additional power from the grid as required.

However turning on the kettle pulls the battery voltage down immediately to below the Pylontech BMS 47.0V low limit and it then sets the allowable discharge current to 0A so stops inverting. This also means that the system stays in idle mode until there has been enough solar PV generation to raise the terminal voltage above the ~49.0V threshold. So even if there is enough energy left in the batteries for smaller loads they will not supply it (the base load in the house is only 400W).

As the battery approaches 100% charge the Pylontech BMS gradually reduces the charging current in 10% steps. So why does it not similarly reduce the allowable discharge current as it gets depleted so that the grid is called to assist at an earlier stage?

Dynamic Cutoff does not seem to be any help either, the recommended settings are all 46.0V so will not operate before the BMS does (and would have the same result, inverter shut-down).

Is there a good workaround for this problem? I do not want to timetable a Scheduled Charge period to help the batteries recover, because after breakfast we are on peak rate electricity which rather defeats the purpose of the ESS.

Thanks in advance for any assistance!

Pylontech
1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Jason - UK avatar image Jason - UK commented ·
Love the title of this post. lol


0 Likes 0 ·
6 Answers
shaneyake avatar image
shaneyake answered ·

It sounds like you don't have enough battery current capacity for your load.
With Pylontech I tend to take the recommended number as the max, they are very over ambitious with their ratings.

On the ESS menu on the GX, you can limit inverter power, which will then only allow the inverter to cover X watts and use the grid for the rest. If you set that to like 2kw that might help or lower if it keeps turning off since the BMS is causing the shutdown if it is commanding 0A.

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

sharpener avatar image
sharpener answered ·

Thank you @shaneyake. Is it possible to control the inverter power via the Aux 2 input? [I plan to use Aux 1 to limit charge current to 20A at night so we are not disturbed by the fan noise].

Otherwise can I automate this process using the OS Large and Node Red, is that now sufficiently stable for me to try it out?

Seems I will have to do the job of the BMS for myself, why can't it control the discharge current in 10% steps like the charge current?

Thanks for your help.


3 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

klim8skeptic avatar image klim8skeptic ♦ commented ·
Buy a lower power kettle?
0 Likes 0 ·
nickdb avatar image nickdb ♦♦ commented ·

Aren't those limited to 25A per battery? So two would be 50A, hence the victron sizing guide stating 4 for a 5kVA.

You can use node red to limit the inverter output but it does seem you are way underspecced.

May the Force be with you..

0 Likes 0 ·
shaneyake avatar image shaneyake commented ·
For both settings you are trying to do, I would recommend using NodeRed on the large OS.

I use it at a lot my installs and it is very stable and now you can just update to latest version and then switch to large image in the firmware menu in the GX.

The assumption made by Victron is that you would still enough batteries to run the inverter at full load while off-grid, if this is the cause then you will never have a problem. For the 5kva that is around 110A continues rating. This the one of the reasons I don't use plylontech but rather FreedomWon if available in area. Their packs have very high discharge current ratings.

0 Likes 0 ·
bm97ppc avatar image
bm97ppc answered ·

On your GX, if you go into the Battery > Parameters (down the page) , in there you can see the CCL and DCL values. (Charge and Discharge limits). The L2 are limited at 100A per Stack due to the cable (approx 20mm2) a 2 Module stack should be 75A standard and 30A normal. Check what discharge limit is showing in there.

75A X 48V = 3600W so that should be fine.

less than 45A X 48V = 2160W so that would not cut it, the inverter will be restricted by the bms.

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

sharpener avatar image
sharpener answered ·

Maybe @Daniël Boekel (Victron Energy Staff) can help here with his expert knowledge of BMSes?

The CCL and DCL are showing up correctly as 74A as per the Pylontech manual.

The internal resistance of the batteries seems to be about 12 - 15 mohm so a 2.4kW (50A) load drags the terminal voltage down by about 0.6 volt. This is not a problem above 20% SoC, but below that the voltage drop hits the 47.0V limit in the BMS.

As you can see from this picture, the DCL has no tapering off, it falls abruptly to zero. This shuts the inverter down completely until there is sufficient PV to bring the voltage back up to 49V, which means the remaining 20% charge in the battery cannot be used.

1652865516036.png

In contrast, you can see in the screenshot below how the BMS reduces the CCL in steps as the battery approaches 100% SoC

1652864313233.png

All I ask is that it does the same thing from 20% down to 0% SoC, that cannot be any more difficult so it is a mystery why Victron/Pylontech have not implemented it.

And since it doesn't seem possible to control the DCL or inverter power easily by means of the Aux 2 input I suppose it means I will have to find out how to do it myself with Node Red.


screenshot-2022-05-18-115232.png (78.3 KiB)
screenshot-2022-05-18-115553.png (68.8 KiB)
2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Alexandra avatar image Alexandra ♦ commented ·
@sharpener

Pylons run at lower voltages.

This has the disadvantage of having a higher current draw for a given load.

When a high current is drawn, there is always voltage sag.

As an example. (Ignoring cable and conversion losses)

Your 2500W kettle draws at full charge 2500w ÷ 52v = 48A

Now when you do that you will see voltage sag on the system. How much depends on battery bank size, cables, current sharing and cell balancing.

2500W at 49V = 51A except your voltage sag will be greater now at lower voltages. Especially if the bank is out of balance (one or two low cells) or undersized.

The system will try and power assist on a big load start depending on how your input current limit is set and if you battery soc is down to the minimum set level.

Your batteries are shutting down the system. They are unhappy.

0 Likes 0 ·
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ commented ·
Your battery is not up to the task, so get a bigger battery or different battery.
0 Likes 0 ·
sharpener avatar image
sharpener answered ·

Yes @Alexandra but if Victron/Pylontech had implemented a stepped reduction to the DCL (as they clearly have done for the CCL) it would provide better protection for the batteries and enable me to use the last 20% of the stored charge which is inaccessible at present. Is there any possibility it can be changed?

2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Alexandra avatar image Alexandra ♦ commented ·

Not unless Pylontec do that. I see the same behaviour in a number of different compatible battery manufacturers. There is maybe one or two steps that is it in charge current.

Discharge current limit, I have only seen the BYD flex have theirs dynamically related to load. (There could be others.)

So it is really down to the battery manufacturers and what they decide for their batteries. The victron really does what it is told to do with a CAN managed battery. How often do you get a good absorption charge on the battery pack?


0 Likes 0 ·
nickdb avatar image nickdb ♦♦ commented ·

DCL/CCL/CVL are hard-set by the battery manufacturer in their BMS and managed by their internal algorithms, the victron device just does what it is told.

It has nothing to do with Victron.

The BMS will step down the draw limits, if it goes hard off it is because you have a problem, usually a cell imbalance - you can track this via battery view or the min/max cell voltages being reported.

You can't configure your way out from undersized/imbalanced batteries.

0 Likes 0 ·
sharpener avatar image
sharpener answered ·

OK so I now understand that it is Pylontech who could improve their BMS as that seems to be the issue. Ironically I wanted BYD in the first place but only the Pylontech were available at the time.

It is evident from this 7-day chart that depending on the weather it goes into Absorption most days and we are consistently achieving 97-98% SoC.

1652877428056.png

In this condition the min and max cell voltages are typically 3.47 - 3.49, as I write the real time readings are 3.742 - 3.946. Immediately prior to applying the kettle load they were 3.24 - 3.25, afterwards they were 3.14 - 3.16, I do not think these figures show a significant imbalance problem at any stage of charge or discharge.

The batteries did give some high voltage warnings when first installed 10 days ago, but they had been manufactured more than 6 months previously and had (only) 75% charge when delivered. The warnings went away as they balanced themselves over the first couple of days in service and have always shown 100% state of health.

So I think with a more sophisticated BMS solution I could utilise the last 20% without causing any harm, I plan to pursue this using Node Red. In the mean time my preferred workaround is to schedule a charge period to say 25% SoC immediately before the end of cheap rate electricity at 0800, that will cost less than being on the grid at peak rate until the sun comes out.

Thanks for all your helpful comments.


1652877428056.png (97.6 KiB)
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.