Is it possible to upgrade?

Hello,
I have recently got installed a solar+battery system, with 4.8kw of solar panels and 4.8kwh of batteries
My inverter is a Growatt SPH5000https://www.growatt.co.nl/upload/file/contents/2019/06/5d0359c10d35d.pdf
The batteries are 2x Growatt US 2000 48V 2400Whhttps://solarclarity.nl/wp-content/uploads/2019/07/Growatt-US2000.pdf

I asked the installers how much it would cost to double the battery capacity so that i could use everything i produce (here selling to the grid get's you about 6c / kwh and buying is about 22c / kwh ) and they quoted me 4000euro for 4.8kwh

Since that is a bit too much i was thinking about expanding it on my own, would it be possible to build a 48Vbattery with 13S 40P 18650and just connect it in parallel with the existing two?

Thanks

gokalex said:

Hello,
I have recently got installed a solar+battery system, with 4.8kw of solar panels and 4.8kwh of batteries
My inverter is a Growatt SPH5000https://www.growatt.co.nl/upload/file/contents/2019/06/5d0359c10d35d.pdf
The batteries are 2x Growatt US 2000 48V 2400Whhttps://solarclarity.nl/wp-content/uploads/2019/07/Growatt-US2000.pdf

I asked the installers how much it would cost to double the battery capacity so that i could use everything i produce (here selling to the grid get's you about 6c / kwh and buying is about 22c / kwh ) and they quoted me 4000euro for 4.8kwh

Since that is a bit too much i was thinking about expanding it on my own, would it be possible to build a 48Vbattery with 13S 40P 18650and just connect it in parallel with the existing two?

Thanks

Click to expand...

Yes, you can add additional battery capability in parallel BUT.there are things to consider/figure out....
- Voltage Range compatibility
The Growatt shows "Discharge Voltage 45-53.5v" A 13s 18650 (lithium-ion) would be pretty close - e.g. 53.5v/13s = 4.11v/cell top which is not bad. 45v/13s = 3.46v low which again is not bad... and since both top and bottom are well within the max for 18650 it would trend toward longer life of the 18650s as you wouldn't be pushing them to extremes (4.2v top and 3.0v low).

- Charge/Discharge Current
The Growatt shows <25a recommended and 50a Max. A0.5a/18650-cell to 1.0a/18650-cell charge/discharge would be roughly 50a/0.5a = 100p to 50a/1a = 50. So you might think in terms of13s50p to 13s100p as a guide - e.g. 650 to 1300 cells - to aproximate a Growatt. A 13s50p @ 48v with average 2200mah cells would have around 4,000wh of power with ~ 2400wh useable (60% DOD) due to the less than max voltage range.

- Other - you'll need BMS, fusing, wiring, independint monitoring. I'm not sure how integrated the Growatt SPH5000 and Growatt batteries are but you'll need to think in terms of monitoring the 18650 battery independently.

Please take the above as thoughtsof what you should investigate/learn about - not as a completed design

- Voltage Range compatibility
I don't mind getting less capacity but more life out of the cells so this is great!


- Charge/Discharge Current

The inverter is rated for 5kw output, and by putting the DIY battery in parallel with the 2x Growatt US2000 batteries i would assume it would draw around 2.5kw from it when running at max load, so 52A at 48V meaning with a 40p pack it would be above 1A per cell... is that too much? looking at the datasheet of the cells i have around they are usually rated for 0.5A normal and 3-5A max
Is the max usuallyonly for very short bursts?

to keep it around 0.5A per cell at max load i would need a 100P pack

(usually my house doesn't run above2kw, only when cooking it goes to 4kw and almost never above 4kw)

- Other
I was planning for a BMS, Fusing on the positiveconnection between the other batteries and the DIY
But i am a bit concerned on how the inverter works with the batteries, since they are connected with a data cable, i wonder if adding capacity will "confuse" the inverter

gokalex said:

- Charge/Discharge Current

The inverter is rated for 5kw output, and by putting the DIY battery in parallel with the 2x Growatt US2000 batteries i would assume it would draw around 2.5kw from it when running at max load, so 52A at 48V meaning with a 40p pack it would be above 1A per cell... is that too much? looking at the datasheet of the cells i have around they are usually rated for 0.5A normal and 3-5A max
Is the max usuallyonly for very short bursts?

to keep it around 0.5A per cell at max load i would need a 100P pack

(usually my house doesn't run above2kw, only when cooking it goes to 4kw and almost never above 4kw)

Click to expand...

Assuming you build a 13s50p (in range of same power as each Growatts) and hook the 18650 battery in parallel with the Growatts - the inverter will draw (equally) from all batteries hooked in parallel. 5000w/48v = 104a. 104a / 3 batteries (2 x original and 1 x 18650) = 35a/battery. This yields35a/50cells= 0.7a/cell. If the 18650 is 13s100p'ish - e.g. twice the capacity of a growatt - then it would supply 1/2 the power - and again your still in the 0.7a/cell range. You can affect this by doing 40p or 60p or 120p etc... and by whether each cell is 2000mah or 2600mah etc. That would all be part of building your 18650 battery - e.g. you can build it to have specific power (watt hours).

The Growatt batteries show 50a(Max) which would ba analogous to 1a/cell in this discussion of 13s50p. 100a (Peak@15sec) would be 2a/cell in a 50p battery for 15secs. You can check the specs of the cells you use - but its normal for a regular 18650 to be able to do at least 2a discharge / cell especially for 15secs. Its when you get above 3-4a/cell that you are entering a different relm usually though of as 'hi drain' such as cells used for skateboards. DIY Powerwalls are typically - as disccused here - <1a/cell charge/discharge range and that's why old laptop batteries are so perfect.

gokalex said:

[size=small]But i am a bit concerned on how the inverter works with the batteries, since they are connected with a data cable, i wonder if adding capacity will "confuse" the inverter[/size]

Click to expand...

The 'inverter' will work against voltage (for sure). So when you add a 3rd (or more) parallel battery it just has the affect of making the voltage of the ones its monitoring go down more slowly during discharge and up more slowly during charge. Voltage is not an issue.

The inverter might trackcapacity - usually expressed in ah - and might cause inverter to report % chargeinaccurately. It 'might' (don't know) cause the inverter to think that it's reach 0% charge too soon and do something like beep or shut-off?. Can think of a couple of things for you to try....
- See if there's a spot where you can manually enter the ah of your overall battery bank - so you could adjust it to account for the unmonitored battery.
- Maybe disconnect the monitoring of one of your Growatt batteries but leave the power leads hooked up and see what happens - e.g. simulate having an unmonitored battery in your battery bank
- Ask the manufacturer (or maybe someone knows on this forum or other forum)

The current batteries use Lithium iron phosphate (that explains the smaller voltage range)
I was looking online for availability of Lithium iron phosphate cells and found out that they are a LOT more expensive, 300-500$/kwh (that also explains why they quoted me 4k for 4.8kwh)

Why is it so much more expensive?

Is there a good supply of used batteries? https://batteryhookup.com/collections/lifepo4-batteries batteryhookup only has this cells that are "power cells" and cost around 400$/kwh (+ shipping and import taxes it can go as high as 700$/kwh so its out of the question)

I also found an European reseller (so no import tax and better shipping)
https://www.wesellcells.eu/presta/index.php?id_product=113&controller=product --> 343/kwh + shipping
https://www.wesellcells.eu/presta/index.php?id_product=35&controller=product --> 202/kwh + shipping

is there nothing that goes below 100/kwh for LiFePo4?


On another note, for the inverter ill need to try it out if no one around here has ever worked with this specific model and can let me know how it works, worse case scenario, if i need to get an other inverter, are there some that can "sense" when my PV panels are producing excess power and so charge the batteries and when my house is getting power from the grid so to contribute the load from the batteries?

If they don't exist, would it be possible to use clamp meters to sense the current going in/out of the current growatt batteries and maybe one also at the main circuit panel to see if the grid is feeding me or reverse?

and maybe with an arduino calculate how much extra power is needed and turn on the secondary inverter for a set ammount of Watts (like half the ammount the primary inverter is putting out)?

gokalex said:

Why is it so much more expensive?

Is there a good supply of used batteries? https://batteryhookup.com/collections/lifepo4-batteries batteryhookup only has this cells that are "power cells" and cost around 400$/kwh (+ shipping and import taxes it can go as high as 700$/kwh so its out of the question)

I also found an European reseller (so no import tax and better shipping)
https://www.wesellcells.eu/presta/index.php?id_product=113&controller=product --> 343/kwh + shipping
https://www.wesellcells.eu/presta/index.php?id_product=35&controller=product --> 202/kwh + shipping

is there nothing that goes below 100/kwh for LiFePo4?

Click to expand...

LifePo4 have higherlifecycle counts(longer life) than 18650 lithium-ion. They are a bitbulkier for the same power storage. 18650 cells are widely used/manufactured making the supply of used ones much larger. All of these things (and others) combine to create a significant price differential. Some claim good luck with Alibaba type purchases... but I've not been lucky that way. So unless you have a supply somewhere, the price differential seems pretty real in my experience.

Batteryhookup is the real deal- but they have recycled batteries and they don't control their supply and they are in the US - so I can understand shipping costoverseas problem.

However, as discussed in the 1st post - you can combine 18650 lithium-ion and LifePo4 batteries into a single battery bank of parallel batteries(with different BMS/monitoring) as long as the voltage ranges are compatible.

gokalex said:

On another note, for the inverter ill need to try it out if no one around here has ever worked with this specific model and can let me know how it works, worse case scenario, if i need to get an other inverter, are there some that can "sense" when my PV panels are producing excess power and so charge the batteries and when my house is getting power from the grid so to contribute the load from the batteries?

and maybe with an arduino calculate how much extra power is needed and turn on the secondary inverter for a set ammount of Watts (like half the ammount the primary inverter isputting out)?

Click to expand...

There are 'simple' grid-tie inverters with current limiting such as this -https://www.amazon.com/ECO-WORTHY-S...miter+grid+tie+inverter&qid=1586612711&sr=8-2 This particular style can feedpower to grid but doesn't have PV monitoring - its on/off based on battery voltage. You could turn it on/off arduino (as you mention below) and there's also threads about controlling units via raspberry -https://secondlifestorage.com/showthread.php?tid=2328&highlight=rs232

There are integrated monitor PV power +battery + grid-tie hybrid systems butI'm 100% off-grid with individual components that I manage and its out of my wheelhouse. Part of the reason I went with Midnite Classic controllers and Batrium BMS is because they are internet with tools for Windows - so I can control everything from my desktop. Raspberry and Arduino are a DIY level lower than I wanted to go for now - maybe one of these days.