WHat we must be very careful in evaluating LI large prismatics i.e. the
" dude my beer is cold the system works" pseudo science. ( with apologies to whoever said that type of thing).
Firstly we have very little reliable data on the life cycle of large prismatic Li formats, hell we are only building up data on small format Li as it is. The EE industry is replete with Li issues, most of which resulted in unexpected premature cell death
heres a amatuer blog KA7OEI's blog: Problems with Lithium Iron Phosphate (LiFePO4) Batteries
One only has to follow the progression in IC Li charge controllers to see the way the thinking has changed in a relatively short time.
It will be a sorry person , whose "beer' isnt "cold" in two years time and down a $5000 investment.
So lets review what the industry ( small Li charging) is saying and doing for best practice (its an area I actively design in). This is current
best practice and of course you may not be able to implement that and the consequences may or may not be obvious .
Nor am I trying to frighten people off, I personally think Li is great . Ive just finished two designs with them in it.
So to summarise
1. Voltage stress is a key factor in determining reliable life and capacity , you pick you charge cut-off appropriately . HVC and LVC also are needed to avoid damage.
2. Float charging or maintenance
charging is not regarded as a good idea, The theory seems to suggest that maintaining any electrical
field , cause Ion movement and in effect "works the battery". Chargers should be disconnected after charge cutoff has been reached. Dont make the mistake of analysing Li in La(Pb) terms , very very different process at work.
3. The charger reengagement point ( ie the voltage where the charger reconnects) seems to be a hotly debated point. Some bring the charger in at about 80% of cuttoff, others try to ensure the battery does a full or near full discharge.
4. "mini-cycles" causes problems as they use up the expected life cycles to some or greater extent. Mini cycles are typically caused in load sharing environments or where , the battery is effectively under constant demand and recharging occurs very regularly ( like in solar charging) .
5. Temperature plays and important part in Li life. extremes cause lower life, below freezing is very problematic
6. Load sharing causes mini cycles, which reduces life times, especially where high peak , short duration currents are being supplied by the battery in a load sharing situation. The battery should if possible be isolated from the power source in that the power source should not cause charging to occur.
NOW, lets look at what we got in terms of existing boat charging sources and examine the issues
1. Smart LA orientated chargers, also used as a power source.
In my mind this is the biggest culprit as its often the source that is left active for long periods. Most people in marinas
have teh battery on 24/7 with it supplying the boats loads in parallel to the battery.
Without modifications, a 3 stage charger is quite unsuited to Li charging.
Most chargers when the boat is being used are load sharing with typically more then float voltages being applied, this is causing mini cycles, we can only wait and see what effect it has on life cycles.
Ebaugh , I believe saw 15% reduction in capacity after 1 year, That if continued could change the cost equations.
AGain this is a problematic method of charging Li, As my definition it tends to be connected for long periods, dropping in and out of the circuit. Theres no easy answer to solar and Li
3. Genny, ALternators etc,
Much less of an issue as they tend to used solely as a recharger and then shut down. good match for Li. ( assuming cutoffs appropriate etc).
Does that mean a system today with standard chargers, solars and Alernators can be made to work, yes, is it optimum , no , what will the effect be, Time will tell. ( but it might be an expensive lesson)
My own view , is that Li should be justified for its recharge,discharge , storage and use reasons, BUT not on life times. Justifying the cost because based on flimsy data that suggests you'll get 10 years or something , given our state of knowledge and the less then optimum usage environment is very silly
ie , dont justify Li on cost with the present state of knowledge.
After that YMMV.
enjoy the cold ones.