DIY lithium for novices!

[Poey50]

That is comforting - I'd be careful of your final statement

I know you are joking but it is the case that no-one knows how long a well-installed LFP system will last for. They have far-exceeded expectations so far, and Rod Collins 13 years with no loss of capacity is a case in point. For those restricting charge usage between 15 - 85%, they could probably leave them in their will.

 

[geem]

I know you are joking but it is the case that no-one knows how long a well-installed LFP system will last for. They have far-exceeded expectations so far, and Rod Collins 13 years with no loss of capacity is a case in point. For those restricting charge usage between 15 - 85%, they could probably leave them in their will.

There has been some interesting stuff published on calendar aging. It is likely that they will 'age' before you hit the design cycles. We have had our lifepo4 installed for 56 days and done 12 full cycles. I don't think we will get to 6000 with them
The actually amout of power used is way more than the accumulation of cycles. We run a lot of our loads when the solar is at its peak so the 25A or so at 24v being provided by the solar reduces the power needed by the battery.
At our current cycle rate we would get 76years out of the battery assuming no calendar aging. In addition, the predicted cycles are at 25degC not the +30degC we have. Interestingly, calendar aging is less at higher temperatures but then the heat starts to kill them

 

[Neeves]

There has been some interesting stuff published on calendar aging. It is likely that they will 'age' before you hit the design cycles. We have had our lifepo4 installed for 56 days and done 12 full cycles. I don't think we will get to 6000 with them
The actually amout of power used is way more than the accumulation of cycles. We run a lot of our loads when the solar is at its peak so the 25A or so at 24v being provided by the solar reduces the power needed by the battery.
At our current cycle rate we would get 76years out of the battery assuming no calendar aging. In addition, the predicted cycles are at 25degC not the +30degC we have. Interestingly, calendar aging is less at higher temperatures but then the heat starts to kill them

You and Geem represent the pioneers (I may have missed someone - apologies) for LFP for those of us with an interest. I am expecting developments that either of you might discuss in the future - its too 'new' to be definitive (and may be replaced by 'something else').

My comment was tongue in cheek - but tinged with some reality.

Its too early for me to comment on uptake here, Oz, in the marine industry but it has been embraced by suppliers to (I struggle to define it) the 'caravan, RV and off road market'. The suppliers promote Lithium, I don't know what the market thinks. Lead, in its various forms, is an option, Lithium (I think LFP) is the standard.

Jonathan

 

[vas]

In my experience, voltage is useless at judging SOC at the upper end, as the voltage rises so quickly with lots of solar connected. You would have to be sat there watching it or you would miss where the SOC went from constant voltage to fully charged. It may be far slower with less solar panel capacity.

+1
that's why you do really need an intelligent system that will get all subsystems communicate efficiently one another. So when cells voltage start climbing towards 3.43-3.45V, it will instruct the MPPT controller to drop it's output and when it reaches your set balancing voltage (in my case 3.46V) it will drop to 2A that my passive balancer can follow for half an hour or so and then just drop to practically nothing. Cycle starts when voltage drops due to consumptions.

 

[pmagowan]

Anyone seen this;

Nigel Calders retirement gift to cruisers! Looks like an interesting system for an all electric boat with an aux diesel as the main generator. They appear to have a full lithium system running at 48V

 

[geem]

+1
that's why you do really need an intelligent system that will get all subsystems communicate efficiently one another. So when cells voltage start climbing towards 3.43-3.45V, it will instruct the MPPT controller to drop it's output and when it reaches your set balancing voltage (in my case 3.46V) it will drop to 2A that my passive balancer can follow for half an hour or so and then just drop to practically nothing. Cycle starts when voltage drops due to consumptions.

We have the same set up. Balance on at 3.45v. 30 mins absorption.

 

[Kelpie]

+1
that's why you do really need an intelligent system that will get all subsystems communicate efficiently one another. So when cells voltage start climbing towards 3.43-3.45V, it will instruct the MPPT controller to drop it's output and when it reaches your set balancing voltage (in my case 3.46V) it will drop to 2A that my passive balancer can follow for half an hour or so and then just drop to practically nothing. Cycle starts when voltage drops due to consumptions.

What kind of BMS and MPPT would you need to achieve this?
One of the things I'm not 100% happy with in my system is the interplay between the BMS and MPPT. If I set the MPPT too conservatively, it will never get the batteries full. If I let it go all the way, the BMS will stop the charge and that risks damaging the MPPTs by disconnecting the load.

 

[vas]

VenusOS (running on a raspberry pi 3B+) is doing all the comms controlling.
Victron MPPT 30/100 controller is "instructed" by VenusOS on what to do with the 2X300W panels
diyBMS BMS is communicating with VenusOS via CANBUS telling it how much current it wants at any given point.
diyBMS has it's own custom 200 or 500A shunt doing the in out measurements on a daily or overall level.
Victron Multiplus II is doing the charging and inverting (in my case 99% inverting) again instructed by VenusOS on charging profile when I'm running the generator - last time I hooked up on shore power was 3yrs ago iirc...

Haven't checked extensively, as I quickly decided to work with diyBMS, but there must be other BMSs that communicate in a similar maner.
I understand that many think that my setup is way too expensive and I shouldn't be using Victron kit, fair enough, guess it's my money I decide what to use them for.

For the record the reason I was attracted to this system is that they are both OpenSource/supported by the builder - both the Victron ppl and from the designer of the diyBMS. I wouldn't invest my time nor money on a hard sealed system but then I like to tinker with things s/w and electronics...

V.

PS. edit to add that a Tclass Fuse and a 500A bluesea contactor are part of the system which is running together with the old 4x T105RE trojans.

 

[geem]

What kind of BMS and MPPT would you need to achieve this?
One of the things I'm not 100% happy with in my system is the interplay between the BMS and MPPT. If I set the MPPT too conservatively, it will never get the batteries full. If I let it go all the way, the BMS will stop the charge and that risks damaging the MPPTs by disconnecting the load.

What are your settings?