LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[deckofficer]

Originally Posted by Jd1

How are these less than $1 per ahr ????
They are 60 ahr cells and are being sold for $110!

You need to look again, depending on quantity they ranged from $52~$60.

They have 110 quantity of each model of 60 ahr cells, maybe that is where you got $110. Scroll down the page further.

[Jd1]

I sit corrected ... I only scrolled down to where it says

*110 each USA stock available (No Minimum Order)

I now realize that they mean to say that they have 11o UNITS in stock

Definitively 'old fart' syndrome

[deckofficer]

They have a total of 220 cells at that price, 110 for each model number.

[goboatingnow]

Originally Posted by T1 Terry

We have a pack still in operation with over 1200 cycles on it, still no capacity loss, it not only a matter of "it's all about the batteries" as Plasma Boys says, it's also all about how they are treated at the initial charge, then every recharge after that. I 100% agree with the "Battery Murder System" claim and have done from the very beginning, don't do it if you want a long cycle life, as simple as that.
The next thing is to understand that cells aren't all created equal and different construction types suit different purposes. Cylindrical cells are for full capacity cycling at their rated discharge, not for stand by use, light discharge rates or long life, prismatic cells respond well to low discharge rates, partial capacity use and low voltage float charging. Cylindrical cells are one single set of plates wound into a cyl, the inner cools a lot less than the outer layer but there is a very limited path for the heat to dissipate, so it's the heat that kills them. The prismatic cells, particularly the Winston and Sinopoly brands, have individual plates bolted to a lump of copper or aluminium, depending if it's the anode or cathode, and this acts as a heat sink for each individual plate. The separator is a single sheet woven back and forth between the plates, this rests against the side of the case and also helps to dissipate the heat, so as long as those 2 areas of cooling remain available and the cell voltage kept below the upper knee, the cell remains within it's operating temp and seems to happily cycle with very little degradation, as long as the cell voltage remains between the upper and lower knee. Now for the battery murdering system, that is designed to hold the cell in the upper knee voltage, then dissipate energy via a controlled short between the 2 electrodes, turns the electrical energy into heat energy, then dumps all that heat into the copper block all the negative plates are attached to..... so where can they dump their heat??? The only place left to dump all that heat is into the electrolyte and separator, the copper sheet temp rises and the copper starts to corrode, stuffing the cell..... murder by a thousand cuts.
If you want long cell life, do not store them for long periods above 3.2v, that self discharge of a stored fully charged cell is it slowly eating itself, exactly the same as storing them below 2.8v for long periods. I have both thundersky and Sinopoly cells here with 6 yrs calendar life on they, still in their packing crate, still at 3.2v, that kinda blows the constant 2% to 3% self discharge being a linear thing out the window eh :lol:


T1 Terry

hmm, the literature doesnt really support what you say. its not about heat per say. The results from pouch cells ( which as prismatic cells) doesnt suggest anything different from cylindric cells. Theres no evidence to suggest that low discharge use is any different in 18650 configurations then any other form factor. But perhaps you have access to literature I haven't seen.

If you want long cell life, do not store them for long periods above 3.2v, that self discharge of a stored fully charged cell is it slowly eating itself, exactly the same as storing them below 2.8v for long periods.

I dont understand what you mean by " eating itself". I understand the science as its explained in the literature, what do you mean.

I have yet to see a science orientated explanation

dave

[Grateful]

I've just read through about 60% of the thread and I hope that I am not asking a question that has already been addressed.

Does anyone here have experience with the 'all-in-one' solutions like Smart Battery? The batteries look expensive, but the Pollyanna in me wants to believe that the built in BMS is properly engineered to maximize longevity and performance.

I'm in the very early stage of designing a system to live on 12v for full time cruising in lower latitudes with the usual fridge, AP, lighting, & nav off of a 300Ah LiFePO4bank charged primarily using solar & wind-gen with engine alternator only when necessary. Very little time will be spent in marinas and on the rare occasions we would ever be away from the boat, I'd isolate the bank from the charging sources.

[Grateful]

Uh - I may have just answered my own question, "Is it too good to be true?" by checking out the BBB page for Smart Battery. Yikes!

[mcarling]

^^
Their 8V and 16V LVE and HVE cutoffs are useless. The battery would have already suffered serious damage well before reaching either of those voltage extremes. 10V and 14V LVE and HVE cutoffs would be more appropriate. My sense is that the people who engineered this "Smart battery" know just enough to be dangerous.

[Jd1]

In case anybody is interested, I have attached a block diagram of my lithium system.
The SOC meter is a BMV-700
the BMS is a HousePower BMS
the power relays are Tyco units found for cheap on ebay
the relay inverter is no-name ebay stuff
Lots of time was spent glassing in the shelf for the batteries as well as mounting pads on the hull for wiring and determining cable layout as well as cable making and pulling etc but the actual battery system was really straight forward. It really isn't rocket science!

The diagram was drawn to go into my maintenance folder so it may not contain everything just 'so' but will give the general idea.

Side note: the power relays are voltage direction sensitive but that is not explicitly shown on the diagram. They will work with the control lines hooked up one way but not the other way.

[hellosailor]

The HiPower batteries certainly are priced right. Question being, did they go out of business because of ineptness? Or because there were problems with the product? Which will have no warranty now.

[klaus53123]

Hi Folks,

this a really extraordinary threat. I have tried to read through it starting around page 180, but didn't really manage to get till the end, so I skipped a few pages....

I am about to install 4*400Ah Winston LiFePo4 on my Catamaran. I load the batteries using those methods:

1. approx 600 Ah solar panel with a dedicated LiFePo4 charger
2. two alternators (one per motor) wiht a Sterling high effinciency regulator mounted to each
3. Victron Phonix MultiPlus Compact which can be adjusted to LFP

Here are my questions:

1. I want a BMS which can provide Over and Undervoltage protection via SSRelais. Also I would like to have decent monitoring for each cell built in. Can you advice which brand/model to buy

2. I intend to put all chargers on the same High Voltage Protection Relais. Is this ok?

3. Do I have to make any changes to my alternator? I am reading controversial statements. Will I destroy the alternator as it will be charging more or less always "full power" due to the low internal resistance of the LFPs?

I hope these stupid questions are not answered hundred time in this threat. I couldn't find it at least.

Many thanks for your help.

Klaus

[rfadler]

Originally Posted by klaus53123

Hi Folks,

this a really extraordinary threat. I have tried to read through it starting around page 180, but didn't really manage to get till the end, so I skipped a few pages....

I am about to install 4*400Ah Winston LiFePo4 on my Catamaran. I load the batteries using those methods:

1. approx 600 Ah solar panel with a dedicated LiFePo4 charger
2. two alternators (one per motor) wiht a Sterling high effinciency regulator mounted to each
3. Victron Phonix MultiPlus Compact which can be adjusted to LFP

Here are my questions:

1. I want a BMS which can provide Over and Undervoltage protection via SSRelais. Also I would like to have decent monitoring for each cell built in. Can you advice which brand/model to buy

2. I intend to put all chargers on the same High Voltage Protection Relais. Is this ok?

3. Do I have to make any changes to my alternator? I am reading controversial statements. Will I destroy the alternator as it will be charging more or less always "full power" due to the low internal resistance of the LFPs?

I hope these stupid questions are not answered hundred time in this threat. I couldn't find it at least.

Many thanks for your help.

Klaus

I am looking to install a similar setup. To answer your questions:

1. I was going to use the House Power BMS manufactured by CleanPowerAuto LLC. It's been referenced multiple times on this thread and should do what you need.

2. Good question, I *think* so.

3. Yes, my impression is you do need to make changes to the voltage regulator. I don't know this for a fact as I have not yet installed my system but my plan is to use a Balmar external regulator that is configured as described in Main Sail's how-to article.

As far as reading the whole thread, I'd suggest reading Main Sail's how-to article (link to it in his signature) for a concise description of how to achieve your goal. He has a ton of experience and communicates it well.

[rfadler]

Here's some information that some of you may find interesting.

Through my job I was able to meet with some researchers at the University of Washington in the chemical engineering department to talk about their work on Battery Management Systems (BMS).

They were focused very narrowly on developing an algorithm to optimally charge lithium-based batteries. They claimed that using a voltage-based charging profile like we use today is very inefficient and believe that by coming up with an alternative solution they will be able to charge batteries more efficiently.

Their approach is to utilize the chemical make-up of the battery to arrive at the most efficient charging profile. Of course there are external variables that can be factored in such as max temperature allowed, as well as the relative importance of the speed of charging vs. battery longevity.

Apparently one of their biggest challenges is to obtain the actual chemical make-up of the batteries they are testing. The manufacturers don't like to share the composition information as it's their "secret sauce".

I found the whole approach to be an eye opener. I assume such a solution would be integrated into the battery itself or implemented in a fully integrated system such as an electric car or a laptop where the charging sources and loads are all controlled.

Sorry if this is off-topic but I figured some of you would find this as interesting as I did.

Cheers,

[klaus53123]

Originally Posted by rfadler

3. Yes, my impression is you do need to make changes to the voltage regulator. I don't know this for a fact as I have not yet installed my system but my plan is to use a Balmar external regulator that is configured as described in Main Sail's how-to article.

Thanks for the link. I have already cross read the article. This is the source of my question concerning the alternator. However: this is the first time I am hearing this kind of advice. Would be great if others could comment on this also. I am concerned :-( to spend another few hundred bucks unexpectedly.

Thanks
Klaus

[OceanSeaSpray]

Originally Posted by rfadler

...
They were focused very narrowly on developing an algorithm to optimally charge lithium-based batteries. They claimed that using a voltage-based charging profile like we use today is very inefficient and believe that by coming up with an alternative solution they will be able to charge batteries more efficiently.
...

I think the issue is that "optimally" doesn't mean the same for everybody and "optimal" solutions can be quite harmful to the batteries in the type of application considered here. Also we don't have access to huge charging currents to "optimise" charging.

Rolf recently posted a very interesting link to a university lecture about Li-Ion chemistry and to sum it up, it says: don't leave them fully charged any amount of time and don't expose them to elevated temperatures (40degC or over) or they will deteriorate rapidly as the electrolyte reacts with the plates.

The temperature issue is quite easy to address, the charging one less so. I have been running systematic charging tests by reprogramming my solar controller and I have found that voltages of 13.6V or above are enough to get full absorption and a current taper down to zero in less than a day (daylight) - i.e. it does get them near-full and holds them there until dark every day, just the wrong thing to do (light discharge cell voltage in the early evening is 3.36V). Genasun is marketing a small 10A "lithium" charge controller, it holds the batteries at 14.2V indefinitely... nice.
The higher the charge voltage target, the quicker they reach a given SOC, but the difference between say 14.2V and 13.6V is not huge. The absorption period is not all that long and the result is the same each time.

Around 13.50-13.55V, I get SOC in the 80-90% range at the end of the day, but charging is also clearly much less efficient and current doesn't stop trickling in. It is very inefficient at 13.4V and things would quickly collapse below this. These "low charging voltage" methods would make very poor use of two hours of sun in the middle of an overcast day in winter or a short engine run.

A good algorithm for low C-rate house bank application would probably charge as quickly as possible (which is still a slow charge) up to a target (to be defined in terms of time-voltage) and then drop the voltage down to maybe 13.35V, just making up for consumption basically (I haven't started testing 2-stage algorithms yet).
An ideal house bank algorithm would probably operate them in a band around say 55% SOC, just charging enough to get the required cycling capacity and adapt to consumption if reserve capacity is still ok and it would be completely unsuitable for many other applications.

It is not all about efficiency, but also longevity and lack of other "problems". Those performing systematic heavy cycling need capacity and can charge more and get away with it because they don't spend any amount of time near full, but for all other applications where it is about reserve capacity for the bad days, it is rather a matter of charging fast but not more than required and still have enough reserve. It is not about depth of discharge, but rather maximum level of charge.
Maximum charge voltage is one much discussed parameter here, but how long it is actually held up there appears to be at least as important and holding continuously doesn't seem to be an option unless voltage is really low, the bank is oversize and charging efficiency doesn't matter.

[AlamedaEyes]

Hi Klaus,

My project is not complete although I am planning to install my system this
coming week.

I am using Manzanita Micro components. I have heard that
products from EV Power - Australian Electric Vehicle Specialists — EV Power - Australian Electric Vehicles sells Electric Bikes and car conversions, Electric Bicycles, Electric Vehicles, Conversion Kits.
Are also good.

I have a kilovac disconnect relay. I will probably have a pdf documenting
the project in about 2 months.

Michael