Ok , my bank is slowly getting to its end of life. I might still get a bit out of them.
I have noticed that Lithium prices are “dropping” and wondering if I should move over to to lithium during November madness,
I see you can get a hubble lithium 100ah for R 26k and then some interesting diy ideas from
I don’t mind the diy approach but for the capital layout I would prefer to do a bit more home work.
What is the best bang for buck Lithium Batteries out there and is it worth going the DIY route. (I have noticed that there are some DIY bms kits but I would prefer not going that route)
These cells from LBSA are good quality and many on the forum use them. I have a few and @Gman has built many banks using them.
I would suggest that if you do go the DIY route that you use a BMS that does have communications capabilities. There are a few nice BMS options to choose from.
@justinschoeman created a Canbus module for the Smart Daly BMS (not the version LBSA sell - that one has no comms) that can talk to any Inverter that handle Pylontech batteries.
I created a driver for the Victron ESS system so that the LTT/JDB Smart BMS can communicate that works very well. It should work with any of their Smart BMS range.
And then there are the ANT-BMS which has bluetooth and RS485 capabilities (which I am planning to add to the Victron driver)
So depending on what Inverter you use, you have a few options to look at.
They have Bluetooth already built-in with the software available to program them.
And I have asked for my one to have a galvanically isolated CANbus interface added.
I will see when it arrives and then IF @justinschoeman and @Louisvdw wants, I can send it to them to see if they can interface with their software with it as a test. Idea is that maybe we find a BMS out there that meets all our needs and then see where that can go.
Am also of the view that although 120ah lithium cells are 100% ok, I would prefer to have like 200ah or bigger cells.
Why? Less connections.
Ideally, on a 5kva, one needs min 10kWh bank. I have seen 280ah and larger cells available from China. I’m watching them.
Its all about risk - are you prepared to risk the most expensive part of your build to save 40% on Pylontechs (or the like)?
If you’re keen to experiment, troubleshoot and fine tune, then there DIY route is less expensive upfront than the traditional Pylontech route.
Pylontechs give you a 10 year, approximately 4000+ cycle warranty at 80% DoD (this has been tested and is realistic). That’s 10 years of peace of mind.
With the DIY route you have zero warranty, especially if you are now starting to add your own BMS, custom charging rates etc. etc…A overcharge or 2 mishap?
With the DIY route re you getting new, class A batteries or used batteries? Probably used if they’re cheap. Have used batteries been tested to warrant 4000 cycles of use?
Does your BMS/Inverter talk seamlessly to each other?
I expect that battery technology will change in the near future, so in 10yrs time there will be better than Li options available, maybe buy cheap now, wait for 5 years and see what happens…
At the end of 10 yrs, approximately 4000 cycles your state of health should be 80%, so based on your experience, your state of health should be > 80%, which is what it says on the box when you buy them. Good to know…
I agree that you will nee to be happy with the risk involved. Personally I always aim for 40 to 50% less if possible. Reason being that if you were to supply a warrenty you need to make provision to replace a unit here or there so you will price that into the system. So if you are willing to carry the risk you should also be able to have some reward…
The part that I am not keen on is the BMS development. As this do carry a very high risk – Remember the Samsung note. How important is it for the Victron to be able to communicate with the bms.
@Paul the BMS will handle the safety of the cells, so should be fine there. In the Note’s case that one cell’s safety was not handled correctly. Now this can be the case with another BMS as well. Even perhaps a Pylontech (although I would guess that would be low)
If the BMS and the Victron system does not talk, then the BMS might disconnect leaving you without power. The battery will be safe. It is just not nice to sit in the dark when you are suppose to have battery lights If you set the settings correct the inverter will stop before this happens.
The Victron system will calculate/guess the SOC from the battery voltage, but this is not very accurate in a lithium system as the voltage drop is not the much. So more often the SOC value is out and the gap might grow a bit until it recalibrate at some stage. All this means is less optimum usage available for the inverter. With comms the inverter can use every drop you allow it too.
If you are using a Victron inverter then you can add my serial driver to the GX and hook up your smart battery. The GX will then use the driver to get all the details from the battery that is available. It works very well.
I do agree that the DIY does have a bit of risk as you do not have the same warrantee as an commercial system. However if you take the money you do save as your own warrantee kitty, the nice thing is you can use that money for anything you need to potentially fix/replace even if it would not have been covered by a commercial warrantee. There is no one that will reject your claim but you
Agreed , so I guess the obvious question would be what would a 10kw system go for? That is what I am currently drawing from the grid on average per day.
In my view, it is not important at all IF you have the right sized bank for the inverter with a accurate SOC.
BMS or lead-acid bank, if the batts are depleted, the inverter will shut down.
Now with lead-acid, you can draw much more out of the bank, as nothing protects it, where the BMS will shut down … and that is where we get back to size the bank correctly or limit the inverter as per the bank’s potential.
The example being if one takes a 2.5kwh bank on a 5kva inverter, then we all know that you HAVE to make adjustments. Now with a BMS and integration that is done for you by “force” from some of the suppliers who integrate with the system.
Adding to my post above, the BMS ideally should have a shunt built in to get an accurate SOC. Some of the brand names, I’m told, don’t have shunts, so it is not that accurate in any event.
Worst case one adds a BMV into the mix, as Revov insists on. Because if you have a very accurate SOC, it becomes easy and simple to program the rest to the bank’s limitations, just like with lead-acid.
If you set the settings correct the inverter will stop before this happens.