diy solar

diy solar

20kw with 40kwh battery but Battery draw and house consumption don't balance

Mendoza333

New Member
Joined
Dec 3, 2022
Messages
4
Hi all,

New here on the forum. But keen to contribute.

I've recently had my setup and do have some anomalies.

First setup details:
31 x 660watt panels facing East and West over four mttps (20kw)
31 optimisers on each panel
2 x deye 8k hybrid inverter
8 x 5Kwh Li battery in parralel (40kwh)
Installed on single phase with 100 amp fuse

So first of all. Everything is working fine. However I am struggling to understand the following

1. CT Clamp on the deye inverter connected to the mains cable just after the main meter. Now I am noticing that the battery is always draining more than the consumption for the house. For example, the house consumption is showing 0.5kwh, battery shows 0.69 kwh drain. Its always around 100-190 watt more than the drain. However if I force the system to draw from the grid. The Consumption of the house and the draw from the grid is always equal


2- is it normal for the system to randomly pull 0.1-0.2kwh watt from the grid even if the battery / solar is supplying the full consumption needed. It goes on and off. Not a big deal just a bit weired.

3- one time the system started drawing from the grid while battery showing soc of 59%. When I accessed the bms details, found that the bms shoes a message "force reachrge request" but there was no error codes. Noting the voltage was 51v so it hasn't dropped under 50v .I've deactivated the battery and activated it again, and it continued drawing from the battery as normal.

FInal query:
My plan is to charge the cars and battery at night. And I am wondering whats the best way to get the cars to charge and the batteries aswell at night, while not forcing the inverter to drain the battery to the cars. I do have two car chargers. A zappi and a tesla charger. Zappi ct clamp connected to the mains before the consumer unit, and after the isolator of the Inverter so I am able to monitor the inverter output and grid from the app. Tesla there is no ct clamp feature so it's not connected to anything.

At first I thought I would just set the inverter to charge the battery the same time as the cars and rely on the zappi to reduce its draw so it doesn't max the network (since it has a ct clamp on the grid and setup to max 80 amp draw). But I am unsure if there is a better way!

Hope this make sense. Appreciate your thoughts!

Some settings attached
20221204_012218.jpg20221204_012227.jpg20221204_002749.jpg20221204_012248.jpg20221204_012308.jpg20221204_012303.jpg20221204_012244.jpg20221204_012238.jpg
 
Last edited:
If you want to limit the charge from the grid, in the batterie setting you can click the down arrow and a new screen will let you adjust the amps.
and keep the system work mode screen for that time period batterie percentage high then there is no drain from your batterie until the next period if that percentage is lower then will start draining again.
 
Utility and house match
Inverter and house mismatch by 100-190 W - Sounds like the inverter consumption and system losses (heat)
Sorry correction. Utility and house also don't match. Same % difference. What's strange is myenergi is reporting correct figures.

Screenshot_20221204_023051_Deye Cloud.png

20221204_002553.jpg

If you want to limit the charge from the grid, in the batterie setting you can click the down arrow and a new screen will let you adjust the amps.
and keep the system work mode screen for that time period batterie percentage high then there is no drain from your batterie until the next period if that percentage is lower then will start draining again.
That's a good idea to avoid Evs draining it. Thanks! Is there anywhere in deye to set the max amp of the grid to 100 amps so it also doesn't draw over that figure if both Evs and the inverters all trying to charge at the same time during the off peak tarrif?
 
Is there anywhere in deye to set the max amp of the grid to 100 amps so it also doesn't draw over that figure if both Evs and the inverters all trying to charge at the same time during the off peak tarrif?
I believe in the 8kW growatt setting 11 in the programing allow you to select max utility amps 10-70A
This may need to be per inverter since you have two, better check if you need to set as 50 each, or start with a lower figure and let it run to see what the two inverters do.
 
the house consumption is showing 0.5kwh, battery shows 0.69 kwh drain. Its always around 100-190 watt more than the drain.

2- is it normal for the system to randomly pull 0.1-0.2kwh watt from the grid even if the battery / solar is supplying the full consumption needed. I

My plan is to charge the cars and battery at night. And I am wondering whats the best way to get the cars to charge and the batteries aswell at night, while not forcing the inverter to drain the battery to the cars.
First you have Lithium, un-check Activate Battery, and since you Zero Export to CT put max sell power at 0. I also recommend that your Energy pattern be Batt First not load. Deye is an hybrid inverter that work in 2 directions it or charge or discharge not both a same time load first will stress the battery when you need more power that the panels provide, charge the battery first and than all solar will go to the load.

My 5KW Deye consumes about 80W you have 2 so 100-190W is correct.

The value that the system pull in your configuration show 20W so 0,1 to 0,2 seems to much but it depend on the load you pull, mine sometimes pull more others export to the grid and I have checked Zero export to CT but it's normal and the way it works, Deye recommend 25-100W increase your value to 50W to see if is more stable.

For your plan Max Solar Power is the max power that the inverter can handle, check your model but an SG05LP1 can do 10400W but don't use this for more than a few minutes.

Grid peak shaving is where you define how much you draw from the grid, you have 8000W and since your circuit can handle 100A you won't trigger the breaker 8Kw are 72A at 110V, since you have 2 probably you should set 50 in each, analyse both and adjust for your load.

The way to charge battery and cars at night without empty battery is to charge like you do from 21:00-1:00 at max power and next charge the cars and during the hours they took to charge in system work mode put the power at 0.

Hope I could help.
 
Last edited:
First you have Lithium, un-check Activate Battery, and since you Zero Export to CT put max sell power at 0. I also recommend that your Energy pattern be Batt First not load. Deye is an hybrid inverter that work in 2 directions it or charge or discharge not both a same time load first will stress the battery when you need more power that the panels provide, charge the battery first and than all solar will go to the load.

My 5KW Deye consumes about 80W you have 2 so 100-190W is correct.

The value that the system pull in your configuration show 20W so 0,1 to 0,2 seems to much but it depend on the load you pull, mine sometimes pull more others export to the grid and I have checked Zero export to CT but it's normal and the way it works, Deye recommend 25-100W increase your value to 50W to see if is more stable.

For your plan Max Solar Power is the max power that the inverter can handle, check your model but an SG05LP1 can do 10400W but don't use this for more than a few minutes.

Grid peak shaving is where you define how much you draw from the grid, you have 8000W and since your circuit can handle 100A you won't trigger the breaker 8Kw are 72A at 110V, since you have 2 probably you should set 50 in each, analyse both and adjust for your load.

The way to charge battery and cars at night without empty battery is to charge like you do from 21:00-1:00 at max power and next charge the cars and during the hours they took to charge in system work mode put the power at 0.

Hope I could help.

Thanks a ton for this. Very helpful!. I did change now the configuration to 50w and also tried to reduce it to near zero - but didn't notice a difference. I'll keep monitoring.
I am in the UK, so the calculation 8KW on 230V will be 34 amps in my case. My issue with grid peak shaving is if the load consumes more, it just limits the draw from the network, and uses the battery - which is good but that means it will drain the battery when EV charges. I will use the timing setup so I charge the inverter at max power first then the evs to overcome this. Would have been ideal if the inverter can reduce its draw form the mains while charging if the load reaches the peak - which is a function the zappi charges have via the ct clamp.
In theory I should be able to charge the batteries at 450 amps - but max charge rate per inverter is 185amps (so total is 350amp).

Sorry if this sound like a stupid question, but how long would it take to charge all the batteries from 10-100% lets say on full power?

Done some calculation based on battery parameters. They are as following
1670159551402.png

I am unsure why, but when I set the batteries to charge from the grid currently, they are charging at max 2KW. I might check some of the settings again tonight..

you want to edit you first post, inverter say Growat while you have Deye.
thanks for pointing this out. done!
I believe in the 8kW growatt setting 11 in the programing allow you to select max utility amps 10-70A
This may need to be per inverter since you have two, better check if you need to set as 50 each, or start with a lower figure and let it run to see what the two inverters do.
Yes sorry - I meant deye not growatt. Been a long day..

It's the grid peak function

Do you mean peak shaving function? is my understanding correct that it will just load use the battery to supplement the grid - if the grid reaches the peak, rather than limit the battery charging if the grid Is reaching the peak?

Grid and load will always the same, both are AC
yes I understand. But even the grid and load (if battery and solar are on idle), they don't match the same. I've got another ct clamp that's measuring things, and its reporting 0.2kw constant draw to the inverter which is a bit high - but potentially that's whats skewing the data? it does not show that is inverter consumption/loss.
1670158558130.png

Example currently we are at end of day in the gloomy winter of the UK. 702-674+570-469 = 129 watt constant phantom load in losses + inverter usage
 
Last edited:
Sorry if this sound like a stupid question, but how long would it take to charge all the batteries from 10-100% lets say on full power?

yes I understand. But even the grid and load (if battery and solar are on idle), they don't match the same. I've got another ct clamp that's measuring things, and its reporting 0.2kw constant draw to the inverter which is a bit high - but potentially that's whats skewing the data? it does not show that is inverter consumption/loss.
View attachment 123058

Example currently we are at end of day in the gloomy winter of the UK. 702-674+570-469 = 129 watt constant phantom load in losses + inverter usage

Time is easy. You say have 40kWh of battery, 90% of that is 36kWh. If you charge at 2kW it should take 18 hours + about 20% due to typical losses in the charger.

How accurate are your measurements? You could easily get that sort of discrepancy with accumulated errors in measurements. CT based measuring devices are not particularly accurate unless calibrated.
 
Would have been ideal if the inverter can reduce its draw form the mains while charging if the load reaches the peak - which is a function the zappi charges have via the ct clamp.

Do you mean peak shaving function? is my understanding correct that it will just load use the battery to supplement the grid - if the grid reaches the peak, rather than limit the battery charging if the grid Is reaching the peak?


yes I understand. But even the grid and load (if battery and solar are on idle), they don't match the same. I've got another ct clamp that's measuring things, and its reporting 0.2kw constant draw to the inverter which is a bit high - but potentially that's whats skewing the data? it does not show that is inverter consu

Example currently we are at end of day in the gloomy winter of the UK. 702-674+570-469 = 129 watt constant phantom load in losses + inverter usage

I don't have an EV so may be writing incorrect info, to not draw from your power bank with charging the cars I only see 2 options first time of use and put the power to 0 or you can try to use the smart load to charge the cars when power bank is full and there is extra solar energy, Deye has lot's of functions compared to a normal inverter but not what you want, probably in the future inverters will came with a port to charge cars.

You can also use a contactor wired before and after CT, in before CT it will not draw energy from power bank and will charge EV at full grid power, when having extra solar activate the contactor for after CT, to activate the contactor you can use smart load or an app like home assistant.

Peak shaving do what you describe I assumed that you didn't want to trigger grid breaker.

Regarding "phantom loss" test and know your system and ignore small differences, a friend that also has an Deye wrote me this "seem that the inverter is less efficient when the production is low, if sun provides 150w and load is 100w then it will eat 150w from battery anyway, if the panel production is too low then most of it will get wasted." I didn't test and assume that is related with electronics efficiency and that for a 5KW the solar Start-up Voltage (V) is 125 so lower than 125V you will see solar production but won't do nothing.
 
Grid and load will always the same, both are AC
Battery is DC, Battery to load is about 94% efficiency.
The lower the load the less efficiency of the inverter.
Seems like a good thread to ask this Stand-by vs On consumption question. When there is zero load and placing the inverter from On to Stand-by the PV power increases over 400w.

This is with two EX6500 inverters, which Will tested at 70w. What would account for this, does this indicate consumption is over 270w when On?


PV On vs Standby.jpg
 
Seems like a good thread to ask this Stand-by vs On consumption question. When there is zero load and placing the inverter from On to Stand-by the PV power increases over 400w.

This is with two EX6500 inverters, which Will tested at 70w. What would account for this, does this indicate consumption is over 270w when On?


View attachment 124806
I only see pv power? If you want to find out turn off the ac input and output of the inverter and use a clamp meter… If your system is always on you shouldn’t be worry about the inverter consumption, just add one more panel to cover the losses….
 
I only see pv power? If you want to find out turn off the ac input and output of the inverter and use a clamp meter… If your system is always on you shouldn’t be worry about the inverter consumption, just add one more panel to cover the losses….
Yes, only PV power shown because utility (AC input) is not in use. So much wrapped in this statement: " If your system is always on you shouldn’t be worry about the inverter consumption, just add one more panel to cover the losses…."

The batteries were being charged. The inverters were on, then put in stand-by because lake effect clouds, rain and snow this time of year make it practical to place the system in stand-by until the batteries are charged, which could take days even with 10 460 bi-facial and 10 445 panels.

I was simply wondering (not worried) if the >400w difference in PV to charging between "On" and "Stand-by" when under zero load was typical.
 
Grid peak shaving is what you want. Set it to the max you want to limit your grid usage to. I use 15k. When I’m charging I’m using 11-12k. My house load is another 1.5k and if my Geo kicks in it’s another 2.5k so I see my Sol Ark dial back my charging amps to stay under 15k. Attached is one example
 

Attachments

  • 0444C2DB-BA45-4AC4-B370-DE5F12D27684.png
    0444C2DB-BA45-4AC4-B370-DE5F12D27684.png
    390.9 KB · Views: 8
Back
Top