LifePO4 GBS Amp Hour Testing 2.5v to 3.6v per cell

[createthis]

Why?

You are not even close to pushing thermal limits. Only time to ever be concerned is if the batteries are so hot, you cannot touch them and keep your hand on them. They perform best when at 85 to 100 degrees. You will never see them get that warm in room temps @ 2C operation. Heat does not become an issue until the batteries are in an enclosed box with no ventilation, in 100 degree plus weather, in an EV pulling 10 C accelerating and cruising at extended periods of 1C.

In short your batteries were made to be in an EV. Your mode of operation is not putting any stress on the batteries.

Interesting. This is why I think about heat: http://www.technomadia.com/2015/02/l...attery-update/

When our 500AH batteries were a year old (August 2012), I was able to drain 442Ah from them before having the EMS trigger a low-battery alarm cutoff.
When they were nearing 2.5 years old (November 2013) I was able to drain 409Ah before the system shut down.
Last month (December 2014), during my most recent drain test, I was only able to drain 378Ah before cutting out.
In other words – our batteries seem to be aging way faster than they should – now giving us just 75% of their original rated capacity.

And this a bit down the page:

But while brainstorming potential causes for our diminished capacity recently, Elite let me know that in their experience heat has a huge impact – even within that operating range.
They have observed that a 10C (18F) temperature increase over a baseline room temperature of ~23C (74F) results in the number of lifetime cycles being cut in half.
This means at 33C (91F) usable battery life will be cut in half, and presumably to a quarter at 43C (109F). This is VERY substantial.

No clue how accurate this theory/claim is. However, it seems the LifePO4 lifecycle claims tend to not pan out in practice for people due, perhaps, in part, to temperature. He also talks about float voltage, which we've covered pretty thoroughly elsewhere.

[Sunking]

Jesse all Battery Cycle Life claims are garbage.

To the point yes heat does destroy all batteries. Lithium is no exception. Manufactures claim lithium ion batteries are not a chemical reaction which is false. If it were a pure Ion Exchange they would last forever. There is no documented Lithium battery of any kind that has made it past 1000 cycles. Lithium batteries use organic electrolyte which breaks down with heating and cooling. You will notice it in time when your Ri starts rising up.

But the kind of heat required will be such you cannot touch the batteries and hold your hand on them. Your batteries are in open air at room temps. It would be very difficult for you to even to be able to tell the batteries are warm to touch which is around 96 to 98 degrees before it feels warm to you.

Will Lithium last longer when cold. Heck yeah every battery does that. But lithium batteries perform poorly when cold. In fact most you cannot even charge when freezing or lower, They remain sluggish below 60 F. Put those batteries in the fridge overnight, pull them out and do a Ri test. Then put a load on them and voltage will sag from the Ri. Warm them back up and they are fine.

All I am saying is in your application, heat is not an issue. Put them in an an EV in TX summer when it it 100 degree outside and go racing. Then you will have a heat problem to worry about. DIY EV buildrs that live where they have temps below 40, use electric blankets to warm their batteries while charing and just before the take off. They want their batteries at 80 to 100 degrees.

[createthis]

It was 98 degrees today.

[karrak]

No clue how accurate this theory/claim is. However, it seems the LifePO4 lifecycle claims tend to not pan out in practice for people due, perhaps, in part, to temperature. He also talks about float voltage, which we've covered pretty thoroughly elsewhere.

Thanks for the link, very interesting article.
This graph shows the effect of temperature on lifespan.
CapacityFadeTemp.jpg
It comes from this research paper researchgate.net/publication/251588109_Cycle-life_model_for_graphite-LiFePO_4_cells. This paper has lots of good information on factors that age LFP batteries. You might not understand all the detail in this paper, you can skip the maths and concentrate on the graphs and the text.

From all the research I have done as far as I can see the main factors that degrade LFP batteries are, Temperature, Battery voltage and charge and discharge rates. It looks like the article writer might have started out charging his battery at voltages that would have aged it, he has dropped his charge voltage to 3.55V/cell and floats at 3.38V/cell.

I have set my friend's system to charge to 3.45V/cell and float at 3.35V/cell. It has been running like this for most of the three years it has been operating. I recently thought that one of the cells in the battery had become defective and lost about 30% of its capacity. I replaced it and have been running some tests on the "defective cell". My first capacity test yielded a capacity of around 91Ah from a 90Ah cell, so obviously the cell is not defective and the cell has lost little if any capacity since it was installed. Unfortunately I didn't do a capacity check when the battery was first installed so can't say what the change in capacity has been.

I have always been worried about his battery because it was installed in a black tin shed where the temperature would be getting to greater than 50C (120F) in summer. Fortunately the battery has recently been moved into his house that he has been building over the past three years.

Simon

[karrak]

I actually intend to discharge at up to 2C or 200 amps. I have a Xantrex SW 2000 inverter, a 250 amp class-t fuse, and a Victron BP-220 for the LVD. I'll be using 4/0 MTW for the cable. This is just a toy system; an experiment, so I don't know how often I'll draw 2C, but it's definitely on the radar. Use cases include running power tools like table saws, planers, drills, and even small window air conditioners.

So yeah, I'm thinking about heat.

If you are only discharging at this rate for 1% of the time and it caused say a 10 times increase is cell aging it would only cause an overall drop in the life of 10% (10*1%) which you have to way up against the increased use you will get from the battery. Because of the high capital cost of LFP batteries you have to cycle as much power through them in the least possible time to get the best return on investment.

Considering installing the battery in a temperature controlled environment is probably only worthwhile if your battery costs more than $10,000, where the cost of the power and the equipment to maintain the temperature is less than the costs of the battery aging faster.

Simon

[karrak]

You idiot, you just confirmed what I have been saying all this time. No BMS is required, cells do not become unbalanced, do not fully charge or discharge the batteries. Only difference is you Top Balance and willing to risk over discharging your batteries and relying on a monitor to save your dumb arse. You are such a moron you do not even know you agreed with me.

I think some sort of BMS is needed even if it is just an SOC meter and voltage monitor to make sure that any cells do not go out of their proper operating range and to flag any faults or stuffups that may happen that could damage the battery. My BMS has saved my dumb arse. As for automatic balancing, no you don't need it if you are prepared to do it manually. I have found that LFP batteries do go out of balance. It happens quicker when the battery is new as one would expect. It is quite possible that as the technology becomes more mature that this will be less of a problem.

In the past three years my battery has never gone below 10%SOC, and averages around 75%SOC so I am in no danger of over discharging it. My BMS would warn me if I were to ever to take any cell below 2.8V. A bigger worry for me is overcharging it as I try to charge it to >99% whenever there is enough sunshine to give me as much reserve capacity as possible for cloudy weather. That is why I have my battery top balanced and again my battery monitor will warn me if any cell gets higher than 3.6V.

Simon

[karrak]

Lithium batteries use organic electrolyte which breaks down with heating and cooling.

The electrolyte reacts with the lithium ions. These reactions occur at a greater rate as the temperature and voltage rise. Lithium plating (which is bad) occurs at the anode if the charge rates are too high at low temperature because the lithium ions can't diffuse (move) into the carbon anode fast enough.

All I am saying is in your application, heat is not an issue. Put them in an an EV in TX summer when it it 100 degree outside and go racing. Then you will have a heat problem to worry about. DIY EV buildrs that live where they have temps below 40, use electric blankets to warm their batteries while charing and just before the take off. They want their batteries at 80 to 100 degrees.

This is probably why people that do this get such poor lifespans from their LFP batteries. They should be using batteries that are better suited for the application.

Simon

[createthis]

Thanks for the link, very interesting article.
This graph shows the effect of temperature on lifespan.
[ATTACH=CONFIG]n325434[/ATTACH]
It comes from this research paper http://www.researchgate.net/publication/251588109_Cycle-life_model_for_graphite-LiFePO_4_cells . This paper has lots of good information on factors that age LFP batteries. You might not understand all the detail in this paper, you can skip the maths and concentrate on the graphs and the text.

That's pretty interesting. I note two things:
1. It says they used 2.2ah A123 cylindrical cells. Probably prismatic GBS cells won't perform anywhere near as well.
2. That graph shows AH throughput on the X axis, not cycles. It's not clear whether 1000 ah throughput means 500 cycles, or 250 cycles.

This might actually be a better graph:


From that graph, it appears you can expect a max of 1000 cycles from a cylindrical cell at 90% DOD and 60C.

However, when you look at the actual matrix of cells they used for testing:


It appears only two cells where tested at 60C and 90% DOD, and they only made it to 754 and 518 cycles, respectively. So, I'm not really sure what's up with the other graph. I'd personally look at the matrix. It appears to be the most telling. At 45C and 80% DOD and C/2 rate you might be able to expect 2000 cycles from a cylindrical cell. 2C appears fairly untested, but what they did test looks super weird. It appears to perform BETTER than C/2 at 60C and 90% DOD. Weird.

Prismatic will probably fair a little worse in every way.

[Sunking]

It was 98 degrees today.

Are your batteries outside, in direct sun, enclosed in a box, no or little ventilation, and discharging at 1 to 5C?

or

Setting in your cool basement in the open air doing light duty work?

It is highly unlikely you ever have to worry about over heating. Just guessing due to the fact you have a basement you should be more worried about Winter and freezing temps.

However it is easy to tell if your batteries are getting to warm. If you cannot touch them and hold your hand on them because they are to warm, time to worry and take action. Turn off whatever is turned on.

What I can tell you from experience of running a lot of lithium batteries ran at 50C continuously until discharged in the summer outside, the batteries do not get warm enough I cannot comfortable hold, with one exception, a damaged cell. In your low rate or range up to 2C operating in your conditions the batteries should never feel warm to touch unless you have a cell problem. If you feel a cell getting warm, replace it. Guarantee you the Ri shot up which is what generates the heat. As temps go up, Ri goes down.

FWIW do not compare A123 Cells to your cells. That would be quite an insult to A123 as they are far superior to any of the Chi-Com Prismatic Cells. That is why they cost 3 to 5 times more than you paid per wh.

If you were to compare say 50 of the A123 Cells in parallel which is equivalent to one of your cells what would you expect the Ri to be? We know yours are roughly 2.5 milli-ohms. What would A123 be? Well each A123 18650 cells are 5 milli-Ohms each, twice that of your single cell. If we connect 50 in parallel to equal your capacity the Ri drops from 5 milli-ohms to 100 micro-Ohms. That is 1/25th of what your cells are. That means you would have to use 25 of your cells in parallel to equal A123 Ri. Worth noting A123 does not own the lowest Ri award. They are in the middle of the pack.

So how does that effect you on say 2C discharge. Well your cell voltage sags 5 volts (2 volts on 4S 12 volt battery) and generates 100 watts as waste heat. Let's stop and think about that for a minute. Your pack voltage is 13.6 volts fully charged up connected to your Inverter with your LVD set to 12 volts. We hit the batteries with a 200 amp load, and your battery voltage sags to 11.6 volts. Whats happens to the LVD? Nothing wrong with your battery, it is fully charged. Why did the lights go out? Your LVD operated right?

If you used A123 of same capacity your voltage sag is .02 volts and you loose 4 watts. Your Inverter and my EV are Happy Happy with 13.58 volts.

What I am driving at here here is Ri explains and answers most all your questions. That is why I ran you through the drill.

[createthis]

Are your batteries outside, in direct sun, enclosed in a box, no or little ventilation, and discharging at 2 to 5C?

or

Setting in your basement doing light duty work?

They'll be outside, potentially in direct sunlight, but not if I can help it. They'll be inside a Pelican 0350 case. Whether or not I ventilate the battery compartment of the case is a design issue. That's why I'm thinking about it.

Long term, I think it would be really cool to utilize some PC water cooling tech on each component that sheds heat (Midnite Classic 150, BP-220 LVD, Xantrex 2000W Inverter, and LifePO4 battery). This would be awesome, because I could have the radiator on the outside of the case and have all of the other components inside the water tight case with the lid closed.

However, in the short term, my plan is just to open the case lid and place those components that shed the most heat above the case lid line so they'll have access to air flow. The battery will be below that line inside the guts of the case, but I can add a fan and a vent hole in the top plate of the case so the battery heat has somewhere to go.

[Sunking]

They'll be outside, potentially in direct sunlight, but not if I can help it. They'll be inside a Pelican 0350 case. Whether or not I ventilate the battery compartment of the case is a design issue. That's why I'm thinking about it.

Long term, I think it would be really cool to utilize some PC water cooling tech on each component that sheds heat (Midnite Classic 150, BP-220 LVD, Xantrex 2000W Inverter, and LifePO4 battery). This would be awesome, because I could have the radiator on the outside of the case and have all of the other components inside the water tight case with the lid closed.

However, in the short term, my plan is just to open the case lid and place those components that shed the most heat above the case lid line so they'll have access to air flow. The battery will be below that line inside the guts of the case, but I can add a fan and a vent hole in the top plate of the case so the battery heat has somewhere to go.

Jesse you do understand your batteries are not capable of running a large Inverter right? You cannot pull 2C from those batteries without your LVD operating. The Largest Inverter you can run for any meaningful length of time of say 10 minutes is 1000 watts or a 100 amp load. Only way you can make it work is to lower the LVD down to 10 volts. At 12 volts you trip from under-voltage.

The lower voltage from the voltage sag is going to make the Inverter draw even more current.I,

It is the Ri biting your butt.

[createthis]

Jesse you do understand your batteries are not capable of running a large Inverter right? You cannot pull 2C from those batteries without your LVD operating. The Largest Inverter you can run for any meaningful length of time of say 10 minutes is 1000 watts or a 100 amp load. Only way you can make it work is to lower the LVD down to 10 volts. At 12 volts you trip from under-voltage.

The lower voltage from the voltage sag is going to make the Inverter draw even more current.I,

It is the Ri biting your butt.

Let's do some math. My cell internal resistances:
Cell 1: 0.00256 ohms
Cell 2: 0.00272 ohms
Cell 3: 0.00288 ohms
Cell 4: 0.00272 ohms

In series, these are added, so 0.01088 ohms total.

V = R * I

V = 0.01088 ohms * 200 amps

V = 2.176

That doesn't seem too bad. 13.2V - 2.176V = 11.024V

Would that generate 435.2 watts as heat, btw? That seems like a lot of heat.

Granted, there are other resistances in the system. Am I missing something?

[karrak]

V = 2.176

That doesn't seem too bad. 13.2V - 2.176V = 11.024V

Would that generate 435.2 watts as heat, btw? That seems like a lot of heat.

Granted, there are other resistances in the system. Am I missing something?

You are not missing anything, yes you would be dissipating round 400W of heat in the battery. If you halve the current you only get 50W to dissipate.

Your figure of ~ 2 volts matches what I get in reality. With my 24 volt system I get a drop of ~ 1 volt with a load of 0.5C-0.6C. We have to divide this by 2 for a 12V system so we get 0.5V. Multiply by 4 for 2C and we get 2 volts.

I believe Barba is running loads of 2C. He has been using Winson batteries for over 6 years. Unfortunately I am not sure if he is still lurking on the forum after Sunking was so rude to him.

Simon

[karrak]

They'll be outside, potentially in direct sunlight, but not if I can help it. They'll be inside a Pelican 0350 case. Whether or not I ventilate the battery compartment of the case is a design issue.

If it were me I would put the batteries and equipment in a ventilated steel cabinet with a hat on top and maybe some extra panels mounted on any sides of the box with an air gap that get any direct sunlight on them. If you put vents in the bottom and top of the box you will get extra convection cooling, make sure you put fine mesh screens on any vents to stop bugs getting in. The reason for steel is that it will dissipate heat from inside the box far better than plastic, aluminum would be better but is more expensive. You would have to modify this if temperatures got down below freezing

If you use sealed containers you have to be very careful not to get condensation occurring.

If you were going to run high discharge rates I would put an aluminum plate between each cell in your battery and thermally bond it to something on the side of the battery pack to dissipate heat.

Simon

[dax]

They'll be outside, potentially in direct sunlight, but not if I can help it. They'll be inside a Pelican 0350 case. Whether or not I ventilate the battery compartment of the case is a design issue. That's why I'm thinking about it.

Long term, I think it would be really cool to utilize some PC water cooling tech on each component that sheds heat (Midnite Classic 150, BP-220 LVD, Xantrex 2000W Inverter, and LifePO4 battery). This would be awesome, because I could have the radiator on the outside of the case and have all of the other components inside the water tight case with the lid closed.

However, in the short term, my plan is just to open the case lid and place those components that shed the most heat above the case lid line so they'll have access to air flow. The battery will be below that line inside the guts of the case, but I can add a fan and a vent hole in the top plate of the case so the battery heat has somewhere to go.

The simplest method of keeping your pack at the right temperature is to insulate the housing and use a peltier unit to control temps inside. All you need is a peltier and thermostatic controls, we use them where ever a system is going to be subject to varying temps and they are set a between 18-22degC. Peltiers use very little energy when set up properly and provide heat as well as cooling.