Lithium Batteries (for the rest of us)

[goboatingnow]

There is a lot of data and research on small Li cells as regards life time and capacity and storage life etc. google is your friend.

Dave

[ebaugh]

At least one person was interested in this.... I measured the current taper for a cell during the constant voltage phase for terminating at 3 different voltages. The attached chart shows a profile of charging at 20A, or C/5 focusing on the current as it ramps down to C/100, or 1A on a 100Ah cell.

As we expected it drops off rapidly at 3.55V, and lasts almost an hour at 3.35V. The 3.35V cell after termination was the one charged to 3.55V. It took another 22Ah in to fully charge, so at the end of the taper for 3.35, it was about 80% charged.

This does not explain my observation from the installed bank in my boat being more like 95% after 3 days at 3.35, but then again I quit charging after only 1 hour of "floating". I will try a longer period one of these days.

[T1 Terry]

Originally Posted by goboatingnow

There is a lot of data and research on small Li cells as regards life time and capacity and storage life etc. google is your friend.

Dave

please don't take this the wrong way Dave.
There maybe lots of information about small capacity cells, but that has little relevance to house batteries, no one would consider building a house battery out of small capacity cells would they? Google is a great source of information and misinformation, so some filters need to be applied, is this information about the topic I'm researching? Do the voltages match up to the voltages for the cells I'm researching or does the person providing the information not know the difference between the cell types or chemistry.
It's like the apple and tomato thing, a smart man knows they are both a type of fruit and both have edible skins, a clever man knows which one goes in a fruit salad.

T1 Terry

[noelex 77]

Originally Posted by T1 Terry

please don't take this the wrong way Dave.
There maybe lots of information about small capacity cells, but that has little relevance to house batteries, no one would consider building a house battery out of small capacity cells would they?

There is still little information about lithium cells in a boat house bank and I think it is a mistake to dismiss the experience from other users.

Some RC modellers have used and killed dozens of LiFePO4 battery banks. They have used this sort of battery technology for a long time. The chargers they use are far more sophisticated than anything used on boats.

While there are some differences in how the cells are constructed and used , there are some valuable lessons to learned from their experience.

It says a lot about the relative maturity of the two industries when many boats are using a tiny lightweight cell logger that has been designed for the model industry.

[OceanSeaSpray]

Originally Posted by ebaugh

As we expected it drops off rapidly at 3.55V, and lasts almost an hour at 3.35V. The 3.35V cell after termination was the one charged to 3.55V. It took another 22Ah in to fully charge, so at the end of the taper for 3.35, it was about 80% charged.

This does not explain my observation from the installed bank in my boat being more like 95% after 3 days at 3.35, but then again I quit charging after only 1 hour of "floating". I will try a longer period one of these days.

Very interesting data. It needs much more time to fully absorb at lower voltages before an ultimate SOC can be matched to the voltage level.

This is not floating, it is CV charging. Floating would be disconnecting and waiting for the voltage to drop before resuming.

Thanks for taking the time to carry out the experiments!

[goboatingnow]

Originally Posted by OceanSeaSpray

Very interesting data. It needs much more time to fully absorb at lower voltages before an ultimate SOC can be matched to the voltage level.

This is not floating, it is CV charging. Floating would be disconnecting and waiting for the voltage to drop before resuming.

Thanks for taking the time to carry out the experiments!

So are we concluding here that very low charging voltages exhibit a long current tail ( which makes intuitive sense ) and this is a similar effect to high C charging

Dave

[goboatingnow]

Originally Posted by T1 Terry

please don't take this the wrong way Dave.
There maybe lots of information about small capacity cells, but that has little relevance to house batteries, no one would consider building a house battery out of small capacity cells would they? Google is a great source of information and misinformation, so some filters need to be applied, is this information about the topic I'm researching? Do the voltages match up to the voltages for the cells I'm researching or does the person providing the information not know the difference between the cell types or chemistry.
It's like the apple and tomato thing, a smart man knows they are both a type of fruit and both have edible skins, a clever man knows which one goes in a fruit salad.

T1 Terry

I would argue all we are talking about is varieties of tomato, but ultimately they are all the same fruit. What we have to determine is whether the construction of large prismatics or usage patterns show any differences over small cells

And in fairness as a practicing engineer with a half decent lab , I can sort through the google chaff. Furthermore I have several load sharing circuits using Li under my belt. The latest being LiFe. I've yet to see any appreciable difference in characteristics between large and small cells.

And yes the EV guys are building from small cells as is Tesla

The fact that for example you see 1000 cycles in itself isn't sufficient data. We need to see life time against capacity loss, a bit like ebaugh was attempting
Dave

[Maine Sail]

Originally Posted by OceanSeaSpray

Very interesting data. It needs much more time to fully absorb at lower voltages before an ultimate SOC can be matched to the voltage level.

This is not floating, it is CV charging. Floating would be disconnecting and waiting for the voltage to drop before resuming.

Thanks for taking the time to carry out the experiments!

This is what I was saying a few pages back. Dropping the charge voltages extends the absorption time at CV before full or what ever level of charge you have deemed full for your use.. Not a big deal, and it still happens much faster than LA, but it still happens at lower voltages..

[ebaugh]

Originally Posted by OceanSeaSpray

Very interesting data. It needs much more time to fully absorb at lower voltages before an ultimate SOC can be matched to the voltage level.

This is not floating, it is CV charging. Floating would be disconnecting and waiting for the voltage to drop before resuming.

Thanks for taking the time to carry out the experiments!

Floating was perhaps a poor choice of words here, but I was thinking ahead, because I want to leave a cell at 3.35 for 48 hours and see if it continues to accept charge. Any point beyond the chart posted is floating. It just wasn't charged to a higher voltage first.

The problem is the PowerLab firmware won't allow charging for more than 8 hours before it shuts off due to a "safety" limitation. It will log voltages, but not current if an external power source is used. But that log would be pretty boring.

Probably use a lab power supply set to 3.35 for the 48 hour period, and then measure the current required to attain full charge.

[T1 Terry]

Originally Posted by goboatingnow

I would argue all we are talking about is varieties of tomato, but ultimately they are all the same fruit. What we have to determine is whether the construction of large prismatics or usage patterns show any differences over small cells

And in fairness as a practicing engineer with a half decent lab , I can sort through the google chaff. Furthermore I have several load sharing circuits using Li under my belt. The latest being LiFe. I've yet to see any appreciable difference in characteristics between large and small cells.

And yes the EV guys are building from small cells as is Tesla

The fact that for example you see 1000 cycles in itself isn't sufficient data. We need to see life time against capacity loss, a bit like ebaugh was attempting
Dave

I thought you might take it the wrong way. You need to sort between the battery chemistries before you start doing cell size and use comparisons, Tesla use Panasonic lithium nickel manganese cobalt oxide batteries, not Lithium iron phosphate batteries, it's the apple and the tomato thing I'm afraid.
The construction of cylindrical cells and prismatic cells is completely different, the cylindrical cell is like one packet from inside a prismatic cell wound up in a tight roll, they have no way of dispersing the heat build up from the inner windings of the continuous single pr sheet cell, so heat effect anywhere within that single coil will cause a failure of the whole cell as that is the only energy storage component within that cell. As each cell mates up to the next cell the heat is passed from one cell to the next, this doesn’t happen with prismatic cells. There is no load sharing, no heat dissipation, no allowance for differences in make up where one packet may charge faster than another packet or have slightly more or less capacity than another packet, there is only the one packet and it is in series with all the other single packets, perform or die.
Prismatic cells are not like that, there are a lot of packets within each cell, they all share the work load yet they are never worked as hard as the single packet cylindrical cell.
The chemical make up may be similar, the coating thicknesses are very different, the construction method is very different, the way they are used is very different, what you learn from the performance of cylindrical cells has little bearing on the performance of prismatic cells I'm afraid.

T1 Terry

[goboatingnow]

construction method is very different, the way they are used is very different, what you learn from the performance of cylindrical cells has little bearing on the performance of prismatic cells I'm afraid.

T1 Terry

I don't agree reading the research does not suggest that there is any significant difference between the various form factors. There's not a lot of real life data from large prismatics. But nothing I see suggests any significant difference.

Dave

[T1 Terry]

Have it your own way, makes little difference to me Dave, but this is the reason I spent AU$10,000 buying prismatic cells I knew very little about, so I could sort the myth, misunderstanding and plain false "research facts" from what was reality. My findings, believe very little of what you read because most “information” is from someone with a commercial barrow to push.
There is one person I do trust that his information is accurate because he spent so many years researching and testing these cells for NASA and the military and many other organisations and his own personal goal, to find a method of storing electrical energy while it was cheap or abundant and releasing it when the need arouse. He has since moved on from LiFeP04 due to the costs and the readily available compounds required. He now heads a big corporation that has the backing of some very big names because they also believe he knows what he's talking about.
I know you have already been here, but a revisit may remind you of a lot of information you may have missed or not fully understood at that time. http://www.cruisersforum.com/forums/f14/lifepo4-batteries-discussion-thread-for-those-using-them-as-house-banks-65069.html
It is quite humorous to read some of the expert findings with the aid of 2yrs of hands on experience.

T1 Terry

[goboatingnow]

T1 Terry, This thread isn't pissing competition, what we are trying to do is build up a proper scientific basis for evaluating large prismatic cells. With due respects to your fine effort, we need documented capacity graphs from typical uses of LI on boats. The more we get the more we understand.

Were a long way from that at present.

dave

[hellosailor]

"He has since moved on from LiFeP04 due to the costs and the readily available compounds required. "
Can I assume you mean he moved on to other chemistry because other chemistries use more readily available compounds? As opposed to what I'm assuming was a typo, you're saying he moved on because LiFeP04 was too easily available?

Don't be shy, tell us what chemistry you think this unsung hero has really found to be superior. That's the rabbit we're all chasing.

[T1 Terry]

He has developed lithium sodium batteries

As far as pissing contests, not really my thing either, as far as life cycles in any house battery situation, far too many variables to build any sort of meaningful data.
Establishing a max charging voltage and min cell voltage, does discharging below 70% DOD or 80% effect cycle life, what does the reported cycle life really mean, like, 1000 cycles to 100% discharge, does this mean after that the battery is finished or does it mean that it no longer holds 100% of the advertised capacity? The same with 3,000 or 5,000 cycles to 80% DOD, what happens after that, is there no capacity or just reduced capacity.... information like that is what is really needed, trying to establish methods of measure is becoming a pissing contest on it's own.

T1 Terry