With the undesirable combination of load shedding and overcast weather. I have several times recently, used ECO mode to ensure a full battery. I have a default rule set up to charge from grid from 15:00 to 16:00, but sometimes I want to top up EG if the load shed is early in the morning - my batteries SOC, under usual conditions, will be at it's lowest about 7:15. So if the weather's not so lekker in the early morning, I get twitchy.

The thing is, I can't see any pattern to how eco mode works. Clearly I have misunderstood it. This is a problem because now I am having to go back to the office, I have to teach my wife to use it pre-emptively when the double whammy of bad weather and load shedding is upon us.

I'll add that I don't believe that Goodwe load shedding in mind when they designed this feature. If there was no load shedding I wouldn't concern myself with it, because the system will eventually recover. But we do have load shedding, and this seems to be the best tool that I have to mitigate against the aformentioned double whammy.

So some examples (all figures as per SEMS portal)

In all cases there are some constants.

1. The Rated Power % in eco mode is always set to 80% (whatever "rated power" is).
2. The charge current limit is set to 125A by the BMS
3. But the inverter can only charge at 100A anyway
4. Battery is a 200aH Freedom Won 10/8
5. The "rule" for charging was always 1 hour in duration - 9:00 to 10:00, 15:00 to 16:00 etc

 

And so I find
Monday 17th - SOC at 15:00 was 86%. At 15:25 it was 100%. So SOC increased by 14% in 25 minutes = 33.6% per hour.
Tuesday 18th - SOC at 9:00 was 65%. At 9:50 it was 100%. So SOC increased by 35% in 50 minutes = 42% per hour.
Wednesday 19th - SOC at 1:00 was 76%. At 1:50 it was 100%. So SOC increased by 24% in 50 minutes = 28.8% per hour.
Thursday 20th - SOC at 15:00 was 76% at 15:45 it was 100%. So SOC increased by 24% in 45 minutes = 32% per hour.

So every time the battery does get to 100%. Every time the charge cycle started SOC was > 50%. But the rate at which the battery charged is never constant.

I expected that the 80% setting would result in a uniform rate of charge across the hour. But the rate of charge varies, the charge has never taken exactly an hour, and the time taken does not seem to correlate to the SOC at the beginning of the cycle.

Clearly I do not understand how this works. Can somebody offer some clarity please?

Hi @Bobster.

I hope to be able to know, in advance, some fundamental data:

a) What is the desired Dod ON GRID?
b) What is the nominal Energy of your batteries / 48 Volts (SOC at 100%, 4.16 kWh ?)
c) What is the Total Nominal Power of your photovoltaic strings?
d) What is the REAL maximum charging power with which Goodwe GW5048 charges its batteries in full sun or with full mains power to charge? (This is a key and fundamental piece of information, because it will tell us the minimum time in which the batteries can reach 100% of the SOC, starting from the minimum SOC of their DOD)
e) What is the average daily hourly consumption of essential loads?

f) The BMS of your Batteries communicates directly with the goodwe ?

Edited October 24, 20223 yr by Cef

BMS comms will limit the charge current once battery charges beyond 90%...

I.o.w. you should have a uniform charge rate up to 90%, thereafter a (reduced) uniform charge rate up to 98% I think, at which point there's a further reduction up to 100%

Edited October 25, 20223 yr by markus_m2

8 hours ago, Cef said:

Hi @Bobster.

I hope to be able to know, in advance, some fundamental data:

a) What is the desired Dod ON GRID?
b) What is the nominal Energy of your batteries / 48 Volts (SOC at 100%, 4.16 kWh ?)
c) What is the Total Nominal Power of your photovoltaic strings?
d) What is the REAL maximum charging power with which Goodwe GW5048 charges its batteries in full sun or with full mains power to charge? (This is a key and fundamental piece of information, because it will tell us the minimum time in which the batteries can reach 100% of the SOC, starting from the minimum SOC of their DOD)
e) What is the average daily hourly consumption of essential loads?

f) The BMS of your Batteries communicates directly with the goodwe ?

Hi @CefI have to physically commute today, so I can't answer all of these questions now because I need PV Master to get some of this data. But
a) 40%

Actually I'm not sure I can answer all of these. The best I can do for E is (daily average - swimming pool)/24. Because the swimming pool is the only big load that should run every day and which isn't backed up. That's about 0.42 kWh. 

This is exposing how much I don't know! 🤣

6 hours ago, Bobster. said:

a) 40%
 

OK! 😄 @Bobster.. Short version:
When the batteries reach the set DOD (40%), their SOC reaches its lowest point of 60%. It used approximately 3,840 kW from its batteries. That is what needs to be filled.
The theory says that Goodwe can charge up to 100 amps, but in practice I really don't know with which batteries this is possible.
I give you my own example. I have 4 Pylontech US3000C, that is to say 296 ah that make approx 14.208 kWh nominal. Goodwe's communication with Pylontech BMS reports one thing, but in reality they only charge between 2.4 and 2.5 kWh maximum, that is, approximately 52 Amper.

The answer point D) is essential. In the worst case with a completely cloudy day and contribute "0" of your PV, I will give an example:
If your batteries communicate through the BMS with the Goodwe, let's assume that they are will charge at a maximum of 40 amps, that is, about 1.92 kWh.
You need to recharge to reach 100% SOC 3.840 kWh, so 2 (two) hours of charging that could provide 1.92 Kw each should be enough to fully charge them.
Continuing with the Example, in the Economic Mode a single charging period of 2 hours or 2 periods of one hour of an hour could be established with a Nominal % equal to or greater than 42% (1932 Watts), but I have noticed that this % is not exact. I would add from 50% as floor.
But without knowing exactly how much the Goodwe charges its batteries per hour, it is not possible to get any closer.

Edited October 25, 20223 yr by Cef

@Cef not sure why you're only charging max 52amp? Are you sure this isn't a PV available restriction under normal mode?

My setup, Goodwe with one US3000 and two US2000s. 

Max charge I've seen is around 82amps (under ECO mode, i dont have enough panels to get that high in normal mode). This works out close enough based on battery specs: 

US3000 - max 37A

US2000 - max 25A each (*2)

Totals max charge of 87A

Edited October 25, 20223 yr by markus_m2

Charge rate will also depend on battery state, as it approaches the set voltage, the charging current will taper down. Depending on your BMS, your SoC could also drift.

9 minutes ago, P1000 said:

Charge rate will also depend on battery state, as it approaches the set voltage, the charging current will taper down. Depending on your BMS, your SoC could also drift.

Correct, this is what I alluded to earlier....I'm sure different batteries / different BMS' do it differently, in the case of Pylontech it will charge at max 0.5C until 90% SOC, whereafter it drops significatly, until I think around 98%, at which point it drops again.

Off the top of my head - Pylontech US3000:

Max charge 37A.
After reaching 90%, charge reduced to max 10A
After reaching 98%, charge reduced to max 2A

(my exact numbers may be off, but you get the point...)

This is just what I picked up in the early days when I was addicted to watching my SEMS portal

 

Edited October 25, 20223 yr by markus_m2

@ markus_m2 Yes ! Minute to Minute current hour (10 am): Yellow: PV, Red: Network, Orange: Battery Charge (2220 to 2565 Wh approx), Blue: Essential and non-essential loads.

 

The automatic BMS configuration does not allow you to configure anything.

Yesterday, Between 9:00 a.m. to 6:00 p.m. Partly cloudy day. Graph by hour of the day. The energy loaded in the batteries is clearly observed:

 

 

When we talk about Maximum Charge for Lithium, using Lead Acid battery terminology, it is the Bulk charge. Indeed, the amperage decreases when it approaches 100% of the SOC, but in lithium the absorption period is very short.

 

Edited October 25, 20223 yr by Cef

8 hours ago, Cef said:

OK! 😄 @Bobster.. Short version:
When the batteries reach the set DOD (40%), their SOC reaches its lowest point of 60%. It used approximately 3,840 kW from its batteries. That is what needs to be filled

Ag! Is 60% DOD. Which they never get down to. The Revovs did, the FW don't. It's usually mid 50s SOC in the morning, after the heat pump has run but before we have useful PV. (the heat pump was killing the Revovs. 

8 hours ago, Cef said:

The theory says that Goodwe can charge up to 100 amps, but in practice I really don't know with which batteries this is possible.
 

Freedom Won say the 10/8 can be charged at 200a, and they ship it with impressive, big ass cables to carry lots of current. But the BMS limits the Goodwe to 90.23a

I'll work thru your example in the morning, and thank you for the time taken. 

Edited October 26, 20223 yr by Bobster.

5 hours ago, Bobster. said:

But the BMS limits the Goodwe to 90.23a

I think that is the inverter determining the limit for itself - you can check what the batteries request in semsportal on the PC (click on devices, bottom left).

(51.2V*90.23A ≈ 4600W)

 

1 hour ago, P1000 said:

I think that is the inverter determining the limit for itself - you can check what the batteries request in semsportal on the PC (click on devices, bottom left).

(51.2V*90.23A ≈ 4600W)

 

Indeed, the BMS is requesting 200A. Good luck with that.

16 hours ago, Cef said:

The answer point D) is essential. In the worst case with a completely cloudy day and contribute "0" of your PV, I will give an example:
If your batteries communicate through the BMS with the Goodwe, let's assume that they are will charge at a maximum of 40 amps, that is, about 1.92 kWh.
You need to recharge to reach 100% SOC 3.840 kWh, so 2 (two) hours of charging that could provide 1.92 Kw each should be enough to fully charge them.
Continuing with the Example, in the Economic Mode a single charging period of 2 hours or 2 periods of one hour of an hour could be established with a Nominal % equal to or greater than 42% (1932 Watts), but I have noticed that this % is not exact. I would add from 50% as floor.
But without knowing exactly how much the Goodwe charges its batteries per hour, it is not possible to get any closer.

Thanks for the time that you took with this.

I now think that the only example that I gave that is truly useful is this one:

Wednesday 19th - SOC at 1:00 was 76%. At 1:50 it was 100%. So SOC increased by 24% in 50 minutes = 28.8% per hour.

because there was zero solar involved at 1 in the AM. The best figure I can find for the Freedom Won is a nice, round 10kWh, so with just grid it delivered 2.4 kWh in 50 minutes. That's charging at 2.88 kW per hour. 

The rest is not easy to calculate exactly. FW's data sheet says nominal voltage is 52V. SEMS portal and the parameter view on PV master say 53V. Advanced settings on PV master says charge voltage is 55.8. In any event it is around 52A.

I suppose with so many inverter brands and so many battery brands, this must be an inexact science. Unless you buy an integrated system like a Power Wall where everything is made by the same manufacturer and you have no choices. THEN everything is very exact.

23 hours ago, Cef said:

@ markus_m2 Yes ! Minute to Minute current hour (10 am): Yellow: PV, Red: Network, Orange: Battery Charge (2220 to 2565 Wh approx), Blue: Essential and non-essential loads.

 

The automatic BMS configuration does not allow you to configure anything.

Yesterday, Between 9:00 a.m. to 6:00 p.m. Partly cloudy day. Graph by hour of the day. The energy loaded in the batteries is clearly observed:

 

When we talk about Maximum Charge for Lithium, using Lead Acid battery terminology, it is the Bulk charge. Indeed, the amperage decreases when it approaches 100% of the SOC, but in lithium the absorption period is very short.

 

Hello @Cef,

Interesting graphs...your hour graph does indeed show roughly 50A charge rate...

You running in ECO mode? Or some other mode which allows charging from the grid? Correct?

I'm no expert, but I suspect that, under your basic settings, you've set something other than 3 * US3000; while the battery BMS is saying "gimme more" your inverter has been setup presumably with 2*US2000s (or similar), hence the inverter itself (and not the BMS) is limiting to only 50A charge? 

If you login to SEMS, what value does it give you at "Charge current limit (BMS)" - when batteries are below 90% SOC this should show 111A (with 3*US3000s installed).

8 hours ago, Bobster. said:

because there was zero solar involved at 1 in the AM. The best figure I can find for the Freedom Won is a nice, round 10kWh, so with just grid it delivered 2.4 kWh in 50 minutes. That's charging at 2.88 kW per hour. 

The rest is not easy to calculate exactly. FW's data sheet says nominal voltage is 52V. SEMS portal and the parameter view on PV master say 53V. Advanced settings on PV master says charge voltage is 55.8. In any event it is around 52A.

@Bobster.

Now yes !
Let's forget about the theoretical 90.xx amps and 4600 watt hours.
So we start from the Base that its batteries are charged at 2880 real watts / hour.
Let's go back. Its nominal energy stored in 48 volts is 200x48= 9.6 kWh. If the SOC at its lowest point is 50% (as you corrected), it means you need to put 4800 watts of power back into your batteries.
You already know that Goodwe will not charge your batteries at more than approximately 2.88 Kwh. That is, if you set a Charge Period in economy mode of 1 hour 45 minutes with a Nominal Percentage of 70% or more, you should be able to reach 100% of your SOC without any problem, at night as you propose.
With reference to the "stability" of the charge that I raised earlier, it must be perfectly stable during the entire Bulk charge time (Maximum Amperage), then it will have peaks and valleys when for a short period of time they are in absorption / floatingas they said previously other participants in this thread.

Edited October 26, 20223 yr by Cef

Hi @markus_m2 Thank you for your interest, attached 2 images of the PV Master. I have 6450 nominal Watts of PV available (let's say in Solar zenith about 5160 effective Watts on a slightly warm day), but that does not matter since the images at this moment correspond to a configuration in Economic Mode with 40% charge of the Grid ( in case of lack of PV due to clouds giving me about 1900 extra watts to charge).

It is clear from the images that the batteries are charged at around 2.5 kWh, far from the theoretical fixed 90 Amper indicated in the battery configuration tab (unchangeable since it is controlled by BMS) and not to mention the 148 amps indicated in the "Batteries Charge Limit". (Obviously Goodwe/BMS error since the theoretical limit is 100 amps).
This shows that one thing is the theory informed by proprietary software and another is Reality...

Edited October 26, 20223 yr by Cef

That is strange @Cef

Not sure why your batteries are only charging at 48A (and apologies, I see now you have four and not three US3000s, nice! 😉

My Goodwe definitely charges at correct spec in Eco mode - did you maybe set, within eco mode, rated power to 50%?

P.s. the 148A charge limit shown is the value read from the BMS. Correct, 37 *4. If you check in SEMS as opposed to PVMaster it correctly labels this reading "Charge Current Limit (BMS)". The inverter itself is obviously limited to 100A, but BMS saying its happy to accept anything up to 148A...

Edited October 26, 20223 yr by markus_m2

@markus_m2  Exactly, the Upload/Download Limit is only informative for the BMS. It arises from the product of the theoretical maximum continuous load x 4 (37x4)=148 a.
I think you have not read well. My charging periods are not at night, but within the maximum daily solar radiation.
Putting 50% into a charging period tells Goodwe to extract approx. 2300 watts but as photovoltaic energy support.
In other words, even though I set 99% to a nightly charge period, my inverter does not charge my pylontechs 4 more than 2500 watt hours.
@Bobster.. defined that its batteries charge 2.8 kWh in one hour in real terms. Read above.
Let's do something, you asked me for a key for the software that is in the download area. Have your system monitor the charge time of your batteries and then we'll see how much power was charged for exactly one hour.

Edited October 27, 20223 yr by Cef

See the example figures I posted earlier. Charging rates are all over the place. In the end I decided to use, as a basis, data from a charge performed at 1 in the morning - thus untainted by the availability of PV.

It seems increasingly clear that that 80% that I specify is not 80% of some current, but 80% of capacity. Or 80% of something. One can't say "X hours at Y amps = Z kWh,  thus SOC will be..."

Exactly it is 80% of 4600 Volts/ampere. If the Power factor is 1, it will be 80% of 4600 Watts, approximately for the GW5048-ES.

16 hours ago, Cef said:

Exactly it is 80% of 4600 Volts/ampere. If the Power factor is 1, it will be 80% of 4600 Watts, approximately for the GW5048-ES.

Thanks Cef. This is getting very close to what I observe in real life. 

The sample data I gave is from SEMS, and that only reports at 5 minute intervals. And 5 minutes can cause a significant difference in the calculated rate of charging. 
 

@Cef, as promised I've run Eco mode and monitored with your software...

I'm currently running one US3000 & one US2000, giving a max charge limit (BMS) of 62A.

Screenshot one - after setting Eco mode, and it taking around 5min to slowly build up to max charge current, it finally settled at 60.2A / 3028W.
Screenshot two - I switched the kettle on for additional inverter load, barely made a difference (60.8A / 3064W).

(I was particularly amazed at how constant it held the charge rate - I'm used to seeing numbers jump up or down continuously, even if only a few Watts, but the charge Amps and Watts stayed incredibly constant...)

Edited October 29, 20223 yr by markus_m2

Hi @markus_m2.

Thanks !!!. I see that it has a fairly old Firmware: 212113, that is the one I had with Lead Acid batteries, the current one is 242416. Right now I am making a Ticket to the Goodwe technical service, so that they can explain to me why my 4 Pylontechs did not can charge more than 50-52 Amper  (could it be due to some bug in the latest firmware???... we'll see...).

Edited October 30, 20223 yr by Cef