AGM still makes more sense than Lithium

Re: AGM still makes more sense than Lithium

Skyko wrote: »

With an AGM, you buy the battery, install it, and check it again in a few years.

With LiFePO4 you buy the batteries, buy some rods and plates to compress them, buy battery monitoring circuitry, balance your cells, run a few cycles, check your cell balance...

I think a battery monitor is even more important with AGM than flooded lead acid. When you buy an AGM you should commission it, which involves cycling and checking cell balance.

Skyko wrote: »

yes I know AGM can outgas but I don't think it can get to dangerous levels if you even have just a little vent.

AGM batteries have been involved in some impressive fires and explosions. The venting requirements are the same as for flooded batteries. Normally they don't gas, and for that matter normally fuses don't blow. When things go wrong, fuses can blow and AGMs can gas and burn.

--vtMaps

Re: AGM still makes more sense than Lithium

Skyko wrote: »

Can you give a link of AGM fires and explosions due to outgassing? Any battery can cause a fire if you directly short the terminals or some such misuse.

Read this about the Lifeline AGM:

http://www.lifelinebatteries.com/safety_bull.php

"Under space available contract with the U.S. Government, each LIFELINE battery not only passed the gas emission test in MIL-SPEC B-8565J (Section 4.6.22), but no single unit generated over one percent hydrogen gas in any test section. That equates to over 99 percent recombinant gas efficiency even during a simulated-shorted cell overcharge."

I wasn't referring to shorting a battery. The problems I've read about are shorted cells, over voltages (due to human or charge controller error), and over current (usually an open cell in a one string causing all charging current into a parallel string). As far as providing links, I don't have a list handy, but I have read one account just a week or two ago on the RE-wrenches forum. That was caused by someone fiddling with the controls and setting the absorb up to 3 volts per cell. I don't recall what brand of AGM battery was involved.

I wasn't aware of MIL-SPEC B-8565J (Section 4.6.22), but this is from the lifeline manual:

Thermal runaway is a condition in which the battery temperature increases rapidly resulting in extreme overheating of the battery. Under rare conditions, the battery can melt, catch on fire, or even explode. Thermal runaway can only occur if the battery is at high ambient temperature and/or the charging voltage is set too high. As the battery accepts current, its internal temperature rises. The rise in temperature reduces the battery impedance, causing it to accept more current. The higher current further heats the battery, and so on, causing the battery temperature to “runaway”. An upper limit will eventually be reached when the electrolyte starts to boil, but once the electrolyte has boiled away, the temperature can climb even further to the point of plastic meltdown and possible fire.

As of this writing, Concorde does not know of any Lifeline® AGM batteries that have failed due to thermal runaway. To preclude the possibility of thermal runway, the charging instructions of Chapter 4 should be carefully followed, especially if the battery will be subjected to high ambient temperatures. Batteries should not be installed near heat sources or in direct sunlight that may artificially elevate their temperature. Also, there should be adequate air circulation around the batteries to prevent heat build-up.

--vtMaps

Re: AGM still makes more sense than Lithium

Skyko wrote: »

Here is what happens when you short a LiFePO4 cell

What does that have to do with AGMs? AGMs also burn or explode when shorted. In the quote I posted, Lifeline is referring to overcharging, not shorting. (although they mention a shorted cell as a cause of overcharging in the mil spec). They are NOT saying their batteries are safe if shorted out.

--vtMaps

Re: AGM still makes more sense than Lithium

Cariboocoot wrote: »

As it is, the alternate technologies such as LIFEPO4, NIFE, and lithium are not suited for the average Joe who just wants to get a system up and running. But if a few people don't take the chance on them these battery technologies will never advance to where they are viable for everyone.

Exactly, they aren't for everyone and one should always match the battery to their application. However the average joe that just wants to experiment with a simple 12v system can just put 4 of them in series (4S), or on a small scale deal with pre-built powersport batteries like Shorai, Antigravity, and others just to get your feet wet.

Unfortunately many look for drama rather than factual material. I could post a link to the guy whose face is burned off from that pesky battery acid hydrogen explosion underhood when he went to recharge his overgassing lead-acid in his truck.

The point however is just to look for the facts. And instead of repeating what one has merely read about, put a simple *quality* 4S Lifepo4 under your belt and get some hands on. You may be surprised.

Re: AGM still makes more sense than Lithium

Actually Lipo is a category without specifying the phosphate chemistry, although most associate that with the critical LiCo02 cobalt types.

For most, that would be a very bad idea, and is one of the reasons that LiFepo4 (IRON phosphate) is a much better deal in regards to safety for the common man.

One reason for banding individual cells together is that you don't want to place mechanical stress on the connector terminals and risk breakage, which are usually affixed with short rigid bus bars. Some have adopted flexible braid connectors to accomplish this. Even though a house-bank is not mobile, this is a very good idea. You'd want to do this lead-based cells too, so no surprise there.

I can say that these would make the perfect test bed for Mike's solar-monolith!

But yes, I'm not throwing my AGM's away either if they don't fit my application.

Re: AGM still makes more sense than Lithium

Skyko wrote: »

I went to West Marine and asked them if I could buy some LiFePO4 and got a blank stare.

It is not ready for the masses. Early adopters, yes.

Well, if you walked into West Marine and asked them for a wind diversion controller you'd probably get the same blank look. But that doesn't speak to the readiness of diversion charge controller technology.

(They are, however, now selling them in at least one local auto parts store, under the name "Super B." Haven't used any yet.)

Re: AGM still makes more sense than Lithium

Skyko wrote: »

Everything I have read seems to indicate people band the cells together to prevent swelling, which just seems insane. Either the design is flawed or the manufacturer needs to construct a better case to withstand the pressures without swelling.

The primary market for these cells is electric vehicles and other applications where size and weight are an issue. To make a prismatic case that would stop the bulging would require extra weight and size. You can pack the cells into a cylinder, but cylinders don't pack very efficiently.

As far as i can see the cases will only bulge if the cells are abused or charged at high current rates. My LiFePO4 batteries have no strapping on them and have been in use for nearly a year. To date I haven't noticed any bulging.

Really makes me think the price of these things in China is about $10 per cell and they are trying to save $0.25 on case materials...then the exporters mark them up to $130 per cell.

I calculate that $130 would buy a 100Ah cell in Australia, tell me where in China I can pick up 100Ah cells for $10 and i will pack my suitcase and get them in person. I thought the LiFePO4 market was quite competitive.

Re: AGM still makes more sense than Lithium

Skyko wrote: »

2) Nobody has documented (that I have found) a LiFePO4 installation with anywhere near 2000 cycles, much less 5000

This is true but there has been lots of testing which suggests these figures are attainable, only time will tell though ...

3) LiFePO4 suffers from age loss of capacity issues even if unused, so I question if 5000 cycles is even possible before the battery has degraded. 5000 cycles at one per day is over 13 years.

This is the elephant in the room. At elevated temperatures and high SOC, capacity loss is a severe problem but things are far less clear at normal temperatures and if we keep the SOC less than 90%.

Next is the issue of availability. I can find 20 local places and 200 online places where I can buy 10 different brands of AGM. With LiFePO4 I am stuck with a couple of Chinese manufacturers. How well are warranty issues handled?

If there are two LiFePO4 battery suppliers in Perth, Western Australia, there must be hundreds of companies in the US that are selling LiFePO4 batteries. Balqon is the distributor of Winston batteries in the US.

Then there is complexity. With an AGM, you buy the battery, install it, and check it again in a few years. With LiFePO4 you buy the batteries, buy some rods and plates to compress them, buy battery monitoring circuitry, balance your cells, run a few cycles, check your cell balance...

I am no expert on AGMs but i would think they need the right charge profiles to self balance and would have to be balanced at the start, if you get that wrong they won't last as long. If you start out with a balanced set of LiFePO4 cells and don't charge above 90%SOC or discharge below 10%SOC you could probably get by without any balancing or monitoring. Not something i would recommend. I think you would be crazy with any battery type not to check them on a regular basis.

Final is price. With the extra parts and battery monitors, I am seeing Lithium cells at 2x the price of AGM even if you equalize the amp-hour rating to 0.8 Lithium 0.5 AGM. With 1000 cycles of AGM and 2000 cycles of LiFePO4, price is a wash, although you have lost opportunity cost in the extra $$$ you spent on Lithium

I have calculated that the cost of storing power with LiFePO4 cells at between 15-30 cents per kWh which i think makes them cost competitive. For details see http://forum.solar-electric.com/showthread.php?19079-Alternative-to-Lead-Acid-Batteries&p=178256#post178256

One of the big advantages of LiFePO4 batteries over Lead Acid is that you don't need to keep them as full as possible to extend their life. During our last winter the batteries were less than full for around two months, and below 60% full for several weeks at a time. Good for the LiFePO4s, not good for Lead Acid. We don't need a generator to make sure the cells are kept topped up when the weather is bad or to give the cells a good charge to stir up the electrolyte etc.

If my assumptions about LiFePO4 battery life are correct, the cost of energy storage goes down with increased usage of the battery. This means we can use more power during summer when it is available without a penalty and make better utilisation of the solar panels and inverter.

Re: AGM still makes more sense than Lithium

Notes -

If you want to deal with CALB cells stateside, you can also get them from Calibpower in Pomona, Ca. Supposedly the gray cells have better performance than the older blue cells especially in cold climates. Don't forget the cell interconnects.

Quick qualification test:
If you are experimenting with a nominal 12v pack like a simple 4S, then once you charge the bank fully to 14.6v, you should see them drop to 14.2v ocv after 8-12 hours. Another 8-12 hours after that, the ocv will settle to 13.3v and stay there forever - well maybe a year. BUT you do not want to store them in a fully charged state. If you are not going to use it for more than a few days to perhaps a week, then discharge them down to about 80% SOC or more. Many thanks to CY at advrider!

In practice, there is no need to fully charge them to 100% SOC at 14.6v all the time. 90-95 percent will do just fine and extend cycle life. I normally only charge my lifepo4's to about 14.1v from my charge controller, although some might take it to 14.4v. The typical "gel" setting on a lead-based solar charge controller does fine since they top out at 14.0 - 14.1v absorb anyway. Ending-amps is a joke since it is so fast - get distracted with a phone call, and you'll notice you are sitting there with barely any current being drawn in the wink of an eye. Well, relatively that is.

The hardest thing to get used to is not doing any sort of float (unless you have a parasitic draw), not charging them to 100% all the time, not storing them fully charged for long periods of time, and the extremely flat charge discharge curve. Since that curve is so shallow, an accurate DC voltmeter is a must - something with the accuracy of a Fluke 87V or better is recommended. If your existing solar CC has temperature-compensation, you may want to take that into account and perhaps disable it for the lifepo4 so you don't have voltages wandering around, especially if you initially set it for 14.4v and a deep cold snap makes the temp comp shoot those voltages higher than 14.6v! Just something to be aware of. Or get a lithium-iron specific charge controller, however my lead-based Morningstar pwm's do nicely when set for the lower voltage.

Some manufacturers list a "recommended" 0.3C charge / discharge. However, they typically can handle 1C charge/discharge or more as long as you are aware of it. Something only my beloved Enersys AGM's would tolerate with ease. Be sure to consult manufacturer data sheets for the *current* model, as voltages listed in internet forums from 4 years ago or so are WRONG. And as always, be aware that well-meaning individuals often confuse LiCo02 (lithium-cobalt phosphate) voltages with LiFepo4 (Lithium-Iron phosphate) We want the lifepo4 specs!

One thing I use instead of a dedicated top-balancing system for small stuff on occasion is a Tecmate / Optimate Lithium-Iron TM-291 charger. I'll spare all the feature details, but one thing I do like is that it will help balance cells that are not extremely out of whack by using an oscillating absorb voltage charge between 14.0 - 14.4v. The trick as I understand it, is that this takes advantage of the quick individual rise-time window of the highest capacity cell to get the lower-capacity cells charged and balanced - rather than just sitting at a static absorb voltage. They do the same thing with their latest model LEAD-SPECIFIC versions for agm's. But so far, I've not really needed or noticed a major balance problem, since I'm usually discharging at 0.1C in my application.

Re: AGM still makes more sense than Lithium

PNjunction wrote: »

If you are not going to use it for more than a few days to perhaps a week, then discharge them down to about 80% SOC or more.

While we are on the subject of fine tuning:
Some would say 80% SOC or less or would say discharge them more than to 80% SOC (which is a really awkward though correct statement.)

Re: AGM still makes more sense than Lithium

How about this:

When I design my system with LiFePo4, I'm going to calculate my needed battery capacity at 70% of rated value instead of 80%, and run with no more than 14.v after an initial balance to 3.55v per cell, easing my needs for external balancing hardware. That gives me some headroom (barring out right cell failure) for balance skew at the bottom, which I will diligently try not to reach anyway. The limits at the bottom are 12v under load, and 12.8v when rested. I will routinely (but not obsessively) check that when at top of charge, my maximum cell deviation is no more than 0.050v.

Sounds like a motto for a diy lifepo4 club. No wonder - it's 115F in the computer room!

Re: AGM still makes more sense than Lithium

Interesting. Do you have a source for this material?

feedhorn wrote: »

My theory of AGM may be different from yours:

AGM batteries work best in UPS or Grid Backup situations where the batteries are charged to 100% and maintained there for the life of the batteries.

New AGM batteries need to float at proper voltage for about 6 months to reach initial 100% charge. Its important to reach a true 100% charge.

Any portion of their capacity less than 100% will eventually turn to hard sulphate and that capacity will be lost.

Daily cycling of AGM batteries makes it impossible to reach a 100% charge and their capacity deteriorates day by day. Total capacity levels off at about the daily cycling capacity.