LiFePO4 battery capacity.

[Peanut]

I'm glad I asked, the discussion has been helpful.

Lithium batteries have come down in price, and it appears that it might be worthwhile replacing my three 85 Ah SLA's with a single 200 Ah LiFePO4 battery.  The downside being the extra complexity and additional spend needed.  I am pretty confident that I am running my 12 volt off the charger at the moment.  The alternative would be the much maligned, hybrid system installed at a lower cost.

I am not confident of constructing a battery or BMS, so would need to buy in what is needed and find out what other changes I need to make.  Thank you for all your replies.

[Tony1]

51 minutes ago, Col_T said:

I don’t actually have lithiums, and am probably unlikely to ever get them, but out of curiosity a question regarding Ah counting on lithium batteries, specifically LiFePO4 batteries.

 

Possibly / probably wrong, but I thought that the Ah counting inaccuracies on LA batteries was due to their internal resistance, which meant that significantly more charge current was needed to replace the current taken by the loads. This then accounted for the ‘drift’, SoC vs voltage, seen by Ah counting devices . Given that lithium batteries have much lower internal resistance than LA batteries, which is why they charge so much more quickly(?), wouldn’t this mean that Ah counting is a viable means of determining SoC on lithium batteries?

 

An expert will hopefully answer this in due course, but my very inexpert understanding is that the 'drift' effect would happen to a battery monitor whether it was measuring lithium or lead acid. 

If the lead acids are resisting the charge to a degree, the monitor will only measure what charge actually goes in there.

Certainly I've seen several lead acid users with victron battery monitors, and they wouldnt be bothering with a BMV700 if there was an inbuilt error specific to LAs surely?

If I remember Nicks explanation correctly, it is the method of counting the Ah in and out that leads to the drift in accuracy. This counting method means that a small inaccuracy in a current measurement will, over time, be repeated many times, and thus magnified-  and will eventually become a significant error. 

So the monitor will measure the current going into a lead acid just as accurately as it will the current into a lithium. 

 

Hopefully that will earn me a gold star from Nick as the Teacher's pet. 

 

 

[nicknorman]

Well, a green star anyway!

 

I think LiFePO4 is slightly more “predictable” than LA in terms of Charge Efficiency Factor. With LA the instantaneous CEF varies with SoC and charge rate. With Li I think it is much more uniform. Actual capacity of LA varies according to how it’s feeling and the state of the moon, whereas again Li capacity is rock steady. In other words, LA is shit and Li great. But we knew that!

 

But as you say, the measuring process has intrinsic cumulative error. A good analogy is wear a blindfold, then walk 100 paces along a wall (at the bottom of the wall you idiot!) then walk 100 paces back. You should be back to where you started. Oh but not quite. Repeat the process and each time you think you came back to where you started, there will be a bit of an error. Since the wind is blowing one way, the error is the same way each time and so after a while you are several paces from where you started.

 

But no matter, you just need to charge to 100% to reset the counter every few weeks.

Edited September 16, 2022 by nicknorman

[Tony1]

7 minutes ago, nicknorman said:

Well, a green star anyway!

 

 

I'll take it! 😂

[Col_T]

Ah, okay. 
 

Is the ‘drift’ slower with lithium than with lead acid, or about the same? Fr’instance, would one fully charge LA every month to reset the SoC whereas it would only need be once every three months with lithium - numbers plucked at random from thin air!

[nicknorman]

1 minute ago, Col_T said:

Ah, okay. 
 

Is the ‘drift’ slower with lithium than with lead acid, or about the same? Fr’instance, would one fully charge LA every month to reset the SoC whereas it would only need be once every three months with lithium - numbers plucked at random from thin air!

I would say yes. But your numbers are perhaps too long, more like weekly and monthly. But then again it depends on how much error you are prepared to tolerate.

[Tony1]

50 minutes ago, Peanut said:

I'm glad I asked, the discussion has been helpful.

Lithium batteries have come down in price, and it appears that it might be worthwhile replacing my three 85 Ah SLA's with a single 200 Ah LiFePO4 battery.  The downside being the extra complexity and additional spend needed.  I am pretty confident that I am running my 12 volt off the charger at the moment.  The alternative would be the much maligned, hybrid system installed at a lower cost.

I am not confident of constructing a battery or BMS, so would need to buy in what is needed and find out what other changes I need to make.  Thank you for all your replies.

 

Not to harp on or anything, but if you are a liveaboard then lithiums really do make a difference- and especially if you have some solar panels and a decent charging system. You have to get into the mindset of looking at it as a 10 year investment, and then it makes a bit more sense.

Today I'm moored with a lot of trees around, and this has been the first day since April that I've had to run the engine specifically for charging- and even then it was just half an hour. 

And if I moved 300 yards further up to a more open location, I wouldn't have had to even run the engine today. 

For a liveaboard, lithiums with solar panels are a real game changer. 

I think a good branded 100Ah lithium battery at £500 is the best value there is going to be for a long time to come. 

 

 

[MoominPapa]

21 hours ago, 1st ade said:

Rounding error somewhere? Charge rounded down (at the n'th decimal place) and discharge rounded up?

 

Apologies if you have already thought of this.

There are lots of possible explanations, but the most likely is just systematic error in the current measurement hardware, especially as it's Hall-effect based, which are known for small biases due to residual magnetism after heavy current flow.

 

MP.

[MoominPapa]

15 hours ago, Col_T said:

Ah, okay. 
 

Is the ‘drift’ slower with lithium than with lead acid, or about the same? Fr’instance, would one fully charge LA every month to reset the SoC whereas it would only need be once every three months with lithium - numbers plucked at random from thin air!

 

Some of the drift is down to measurement errors, but a lot is down to unknown and variable charge efficiency. LA batteries need to be regularly charged to 100% and as they approach 100%, the proportion of the electrical energy used in turning lead sulphate into lead oxide decreases, and the amount used in turning water into hydrogen and oxygen increases. The H2 and O2 are lost from the system. so any energy that goes that way is not stored in the battery, and the actual stored energy is lower than that determined by measuring what goes in. Since the proportion of the water splitting side reaction and therefore  charge efficiency depends quite strongly on the state of charge and physical state of the battery, how much it has aged and deteriorated, it's very difficult to compensate for this in the AH counter.

 

Apart from the energy loss, and some water loss which is easily replaced, these side reactions don't damage the LA cells, and they do some good: the bubbling fixes stratification and the because most-charged cells start to gas first, this allows cell balancing without any external balancing system.

 

On the other hand, the equivalent side reactions in Lithium cells are bad news, and must not be allowed to happen. Hence sophisticated BMS system with charge termination and external cell balancers.  So, because side reactions are avoided in Lithium cells, the charge efficiency is higher and more predictable, so a major source of error in Ah counting doesn't exist.

 

MP.