New Growatt SPH + Li6532 Installation queries

[David_1224]

That's interesting and maybe worrying.


This is a typical SOC plot for my system.
The first overnight charge started at 00:20 and stopped 01:20.
It then discharges back to the 12% level and then you can see it slowly looses charge until 05:55 when the cycle is triggered all over again.
Following this second event the SOC is still at 12% when at 09:10 there is some excess from the PV and solar charging has kicked in.

Does this suggest there's an issue with the battery??

I have to note however that on your plot there appears to be structure in the minimum SOC section of your plot which looks like the charge IS being adjusted.
Can't really tell due to the scale being dominated by the 100% charge.
Might be interesting to look at the raw data between 12:30 and 23:30 to see what that activity really is.
Shame that it's not possible to pick scales and time periods for the plots so detail like that can be blown up.
Would be Soooo much easier.

 

[David_1224]

Have had to wait a while for a decent PV day in order to get an SOC plot with a 100% charge just to compare with that from John in post 40 above.
Although this doesn't have an overnight charge event, yesterdays charge was sufficient to carry over to the early hours, it does illustrate that on this scale the variations of a few % between 04:10 and 8:20 appear relatively small but comparable to those visible at the lowest SOC level in the post 40 plot.

I've also attached a screen shot below of SOC data of one of the forced charge events, taken from the shine phone app, as illustrated in post 41 above. This was taken when I had briefly changed the minimum SOC discharge level to 12% hence the data shows it going up to 14% on forced charge, before falling back to 12%.

John if you get back to this post at all I know you don't tend to use the phone app but it would be interesting to know what the detail SOC variations are on your system during a minimum SOC period such as between 12:30 and 23:00 in your post 40 image above.
What I'd like to check is if you are in fact getting a force charge of some degree during that state and if it actually goes back above your "discharge stop level" of 10%.
I can't unfortunately find a way to get the data below from the web interface that you prefer.

 

[johnb2713]

Have had to wait a while for a decent PV day in order to get an SOC plot with a 100% charge just to compare with that from John in post 40 above.
Although this doesn't have an overnight charge event, yesterdays charge was sufficient to carry over to the early hours, it does illustrate that on this scale the variations of a few % between 04:10 and 8:20 appear relatively small but comparable to those visible at the lowest SOC level in the post 40 plot.
View attachment 16708
I've also attached a screen shot below of SOC data of one of the forced charge events, taken from the shine phone app, as illustrated in post 41 above. This was taken when I had briefly changed the minimum SOC discharge level to 12% hence the data shows it going up to 14% on forced charge, before falling back to 12%.

John if you get back to this post at all I know you don't tend to use the phone app but it would be interesting to know what the detail SOC variations are on your system during a minimum SOC period such as between 12:30 and 23:00 in your post 40 image above.
What I'd like to check is if you are in fact getting a force charge of some degree during that state and if it actually goes back above your "discharge stop level" of 10%.
I can't unfortunately find a way to get the data below from the web interface that you prefer.

View attachment 16709

Unfortunately I cant get on the Growatt APP, it says my account is locked. I'm told it's because my home assistant is polling the Growatt server for data and this is how they respond. What I can tell you 100% is that my system doesnt import from the grid at all when the SOC is at minimum unless it's a scheduled charge time. I know this from independent CT's connected to my Home Assistant.

 

[David_1224]

Thanks again John.
Have to hold my hand up to wanting to fit your data to my observations without thinking it through before engaging fingers to keyboard.
Blindingly obvious point to make is that even if the structure at the SOC minimum is the BMS doing some degree of battery protection, during daylight hours at least it would probably scavenge from PV excess .... so wouldn't draw from the grid anyway.

As I've still got a installer review to take place I'll take this up with them and hopefully Growatt.
Not going to hold my breath though and in the grand scheme of things probably not significant.
Putting it into perspective the system has run for 3 days now without a forced charge event due to accumulated battery charge in the day being sufficient to cover background consumption through the night.
Just want to try and get some reassurance that the charge drain at the minimum SOC isn't symptomatic of a battery issue.

Observation from this morning was a very sudden discharge at around 07:20.
Battery had held up through the night and was getting down to the low ~20% when there was a sudden discharge to ~14% in 5 mins


Raw data over this period doesn't show a spike in demand from the battery or from the load.
Server however only gives updates every 5 mins and I'm beginning to think these are just spot readings, rather than an average, so short duration events could be missed.
Maybe as simple as that.
Has been noted on a few occasions, but not consistently, and there has to be sufficient excess residual charge left in the battery.

 

[David_1224]

Time for an update.
As noted in a separate post, inverter and battery enclosures have now been fitted with grounding straps.
The generation meter has also been swapped over for a ECA2.nz (my purchase). See initial posts if interested.
All completed by my installer and changes documented.

Battery charging is however throwing up some interesting behaviour.
Firstly in terms of what I was seeking to achieve the system does what's expected of it.
Charges the battery from excess PV and runs the property of the charge once the sun goes away.
So far so good.

However in addition to the forced charge cycle around SOC 10%, discussed above in this thread, I've noted another behaviour once charge reaches 100%.
At this point charging stops, as expected, but the battery begins to loose charge in a similar way to what I see when it stops discharging at 10%.
In this case although SOC falls away from 100% the recorded data shows no utility input being derived from the battery.
Property is being exclusively supported by the PV and excess exported to grid.
This continues until SOC hits ~95% at which point, providing the suns still out, excess again routes to the battery and takes the SOC back to 100%.
Guess if the day was long enough, and I failed to over consume, it would continue to cycle like this until PV output stopped.
So is my battery just not holding charge and falls away until a reaches a trigger level or is there something else at play?

On a separate forum I picked up on a thread discussing Growatt inverters, at 100% SOC, actually starting to draw charge from the battery to "self power".
I should add that from recollection this was not for SPH series inverters.
The specification sheet for the SPH only lists a night-time power consumption of <13W, which seems next to naff all but probably simply covers "stand-by".

Any thoughts on if a similar behaviour may be occurring on the SPH and is the night-time power figure the whole story when it comes to the self power rating. (there's nothing on the ratings label as far as I can see)
Anybody got a handle on what it might consume while running or is that just folded up in the power conversion from PV and / or battery? (probably)

Could obviously be barking up totally the wrong tree.
(Trying to build a model that fits the observation but not having all the relevant info)

 

[binky]

I suspect the inverter is running off the battery. Solar Inverters run off the DC from the panels, so I suspect hybrid inverters would do the same running either off the battery or dc. If the battery is fully charged, and there's no demand from the house, then it must shut the DC inputs down, leaving the battery as the only source of energy.

 

[David_1224]

Ah. A missing piece of information.
Thanks binky.
Goes some way to explaining a little of the battery drain, though I think there's a tad more going on here than spec sheet would lead one to believe.
Much will obviously have to do with the Growatts algorithms which sadly seem to be vailed in mystery.
Shame there's not more detail available on the system as a whole.
At least a high level overview of what to expect from a default configured system.
I'll try getting some answers from installer / manufacturer but I'm not going to hold my breath.
If anybody else has some insight / thoughts on the matter please add.
Otherwise I'll look to update if I hear anything.................. could be a while.

 

[johnb2713]

I suspect the inverter is running off the battery. Solar Inverters run off the DC from the panels, so I suspect hybrid inverters would do the same running either off the battery or dc. If the battery is fully charged, and there's no demand from the house, then it must shut the DC inputs down, leaving the battery as the only source of energy.

Why would they not use grid power? Why use the battery and then switch over to charge mode when the SOC has gone below some predetermined threshold? My Inverter has a standby consumption of 13w from the grid when it's not running (as in not generating or charging).

 

[binky]

Why would they not use grid power?

my bad wording, grid power is last resort. Solar inverters use a little grid power on standby mode overnight too, my own is 7Watts, part of the reason I chose it as it's one of the lowest figures.

 

[David_1224]

From observation I'm in agreement with you both.
What triggered the debate was the behaviour at SOC 100% as noted in post 45 above. (Nice sunny day and plenty of PV)
Once it hits 100% Battery charge begins to drop but the available data "suggests" that its NOT going to load.
However the PV is still more than adequate for load and excess is being exported to grid.
When SOC finally drops below 95%, which can take several hours, the PV excess starts charging the battery again.
So either the inverter switches over to use battery during this time (but why would it not use the PV!!) OR my battery is just not holding charge.
Obviously I'd rather establish a reason for discharge than have to prove a "leaky" battery.
Again from recollection the statement on the other forum (which I can't locate right now) was that Growatt had said it's "normal behaviour" for the inverter in question.
As I said before I'll try and get a view on this from Growatt with regard to the SPH.