Ramblings of a Techno-Viking

Battery Price Comparison -- LiFePO4 vs AGM vs FLA

Many pople have told me that Lithium Feric Phospate (LiFePO4 or LFP) batteries are too expensive for them to consider. This is a price comparison of three posible systems for a van dweller with moderate usage, (30-40 Amp-hours per day) and the ability to go a couple of days with limited sunlight. For longer periods of no-sun, the ability to charge from a small generator is included. Prices of quality components from reputable dealers (listed below) have been gotten online without calling for better prices, finding out shipping costs, or using short-term sale prices. The generator and miscilanious wire, fuses, and crip rings are not included, but they will be about the same price for all three systems. Other than the batteries, the systems should last at least a decade with minimal upkeep. These prices can beet, especially for the AGM (Absorbtive Gas Mat variation of Lead-Acid) and FLA (flooded lead-acid) but I would not expect more than a 25% reduction without reducing system quality. In this article, where AGM and FLA have the same properties, I lump them together as "Lead-Acid".

# LiFePO4 AGM FLA
1 four CALB CA100Fl $500 two Lifeline GPL-24T $500 two Trojan T105-RE $348
1024Wh 976Wh 1372Wh
50A 8A 11A
2 Morningstar SS-MPPT-15L $213 Morningstar SS-MPPT-15L $213 Morningstar SS-MPPT-15L $213
3 - Morningstar RTS $26 Morningstar RTS $26
4 Morningstar UMC-1 $35 - -
5 Meanwell RSP-500-15 $89 Progressive Dynamics PD9245-14.8 $188 Progressive Dynamics PD9245-14.8 $188
6 two Renogy RNG-100P $290 Renogy RNG-100P RNG-150D $365 Renogy RNG-100P RNG-150D $365
7 $1127 $1292 $1140

Direct data download

Batteries (row 1)

GBS, CALB, and Winston are the most reputable LiFePO4 batteries. After the demise of Balqon (due to the electric truck side of the business) Winstons are harder to get in the US. Lifeline has the best reputation for AGM, and Trojan for FLA.

Usable Watt-Hours are used for comparison between chemestries. 80% discharge at 12.8v average is used for LiFePO4, and 50% discharge at 12.2v average is figured for Lead-Acid. The peak current (in amps) is shown for each, the available power will decrease minimally on LiFePO4, some on AGM and a lot on FLA if this is exceeded significantly. The batteries will still need the full charge however. (Amp-hours is rated at the 20 hour rate on Lead-Acid, the 2 hour rate on LiFePO4.)

Solar Charge Controller (row 2)

LiFePO4 require an MPPT controller as the voltage pulses from a PWM controller can damage the battery. The same controller was specified on the Lead-Acid for ease of comparison.

Battery Temperature Probe (row 3)

Lead-Acid require higher voltage (and charge less efficiently) at lower temperatures. The probe lets the controller compensate for this fact. LiFePO4 do not want charge voltage changed with temperature, but should not be charged at a high rate below freezing. All batteries have reduced life at high temperatures.

Controller programmer (row 4)

Since none of the controller profiles match LiFePO4, set the controller to Gel and use this with a compter to reprogram the settings on install. 13.8v absorbtion voltage for 1/2 hour and 13.3v float is reasonable for this system. (Choosing this could be the subject of an entire article.) The "float" is only to prevent discharge when solar power is available, it is not a continuation of the charge like on Lead-Acid. This can also be used to monitor the controller operation, and is not a bad idea to have for Lead-Acid to optimise settings and do monitoring.

Converter/charger (row 5)

This takes the output of the generator or shore power and charges the battery. None of the reasonably priced options here are good, and getting close to the price of the rest of the system for something used infrequently is rediculous.

The Meanwell chosen for LiFePO4 needs set to 14.0v (or 13.8v) at install, and needs to be turned off manually when full charge is reached. (After about 3 hours if the battery was discharged to 10%.) Leaving it on after full charge is reached will reduce the battery life.

The Progressive Dynamics charger chosen for FLA does not have remote voltage or temperature sensing. If used by itself on a battery discharged to 50%, it will take 6 or 8 hours to charge the battery.

Either option could be used with AGM, with the charge time of about 5 hours. (Setting the Meanwell at 15v or using the fob on the PD to get it to stay at 14.8v.) The PD has the advantage of not overcharging if left on.

The best way to determine charge state when charging is via the battery current, but that is not an option on any converter/charger I know of.

Solar Panels (row 6)

With an MPPT controller, house panels are cheapest if they can be obtained localy and you have the space for the larger panel. Freight costs can easily exceed any cost savings, so smaller panels are specified here.

Since the Lead-Acid batteries are less efficient, require more charge time, and should be kept as close to fully chaged as possible, 250W of panels compares well to 200W of panels on LiFePO4. 250W of panels can produce more power than the specified Solar Controller can use, but with non-tilted panels that will only be on days where the panels produce more power than the batteries can handle anyway.

Total Price (row 7)

Total system price being cheapest on LiFePO4 surprised me. AGM not being cost-effective did not, but Lifeline GPL-4CT are cheaper per Watt-hour than the GPL-24T. True deep-cycle FLA not being available in smaller sizes and specifying the RE varient of the Trojan T-105 (which may not be cost-effective) upped the price of the FLA. If the roof space is available, some cost saving by using 300W of panels and a PWM controler on the Lead-Acid. Shopping around can reduce the cost, more so on the Lead-Acid than the LiFePO4.

No Cell Ballence Boards or BMS

Cell Ballence Boards are frequenty recomended for LiFePO4, but anacdotal evidence shows they may cause as many failures as they prevent on low-voltage (less than 36 volt) low current (less than 50A for a 100Ah) system. Cell ballencing is only an issue near full charge or full discharge, conditions that should be avoided for long life of an LiFePO4 battery anyway. Lead-acid batteries have been shipped for over 100 years without even the ability to easily monitor cell ballence, and it is an issue with them as well.

A BMS (Battery Management System) with monitoring, high voltage cutoff (for failed charger), low voltage cutoff (for excessive use) and out of balance cutoff (for mismached cell charge levels) would be nice, but is not cost effective on a small system like this.

Issues

FLA require checking acid levels and refilling if needed. Monthly is recommended. FLA require a vented compartment with easy access to the top for this checking. An equalization charge does about the same thing as cell balancing for LiFePO4.

AGM recomend a vented compartment.

The Meanwell power supply is not designed as a battery charger, and needs manual intervention to shut off. Since this is only done when on a generator or shore power, rarely with this hypothetical situation, it may not be much of an issue. A "countdown timer" can be used to shut off the charger after a programmed time. It also does not come with a power cord, switch, or fuse.

LiFePO4 should be installed with pressure plates to avoid capcity loss due to cell distortion. This is mostly an issue high temperatures or high charge or discharge rates. Terminal up mouting is recomended on most brands, and large side down should be avoided.

LiFePO4 are about 1/2 the volume and 1/3 the weight of equivelent usable power Lead-Acid. Shape differences can be an advantage or disadvantage.

Conclusion

LiFePO4 systems are comparable in total price to an AGM or FLA system for a fulltime boondocker. When you add the longer expected life (2000-4000 cycles rather than 500-1000) the life-cycle cost winds up much cheaper. FLA can be cheaper to set up, but only significantly if lower cost options are used throughout reducing system reliability and costing more in the long run.

Part Sources used for prices

Posted Tuesday 13 December 2016 15:34 UTC
Last edited Sunday 18 December 2016 00:10 UTC
Tags: