LiFePO4 for the off-grid campervan

intro

This is not a rah-rah session for LFP; it’s about how in my particular case lithium (Li) is a good fit. And how I use and charge the battery.

my use case

I boondock full-time in the campervan. I previously had a FLA GC2 bank that provided excellent service at a good $/kAh cost. The bank started to decline after 1000+ cycles and my needs slowly evolved.

practical effects on my boondocking

For my present uses lithium offers these benefits over lead:

ability to camp in areas with limited sun. No need to get 6+ hours of continuous sun (to complete Absorption), no need for strong sun at particular times (Bulk charging in the morning to reverse soft sulfation) or keep good sun to sundown.
ability to run loads when I want. Since Li doesn’t care about sitting at partial state of charge you can “borrow against future sun” in a way that lead chemistries do not appreciate.
at my sub-C uses Li should last a very long time, and may be less $/kWh than Pb. [Update: I did the math and the breakeven appears to be 2,462 cycles, about 6.75 years. Check back on Saturday, July 8, 2028. :-) ]

my bank

My original bank was 220Ah of 6v GC flooded in series for 12v. When that died I switched to lithium.

v1.0 - 1x 100Ah LFP battery from Rebel Batteries. By spec my usable Ah decreased from 110Ah (FLA @ 50% DoD) to 80Ah (LFP @ 80% DoD). But in reality the 100AH LiFePO4 was outpeforming my old FLA bank.
v1.1 - In Feb 2023 I was given a 50Ah Chins LFP when the previous owner upgraded to the 100Ah version.¹ I paralleled it with the original 100Ah and it worked surprisingly well.
v2.0 - replaced the hodge-podge bank with a single SFK-275SE “275Ah”² battery for the reasons described in this blog post. At $905 delivered it wasn’t as cheap as the value brands like LiTime, Chins, etc, but it is maintainable, has an active balancer, a BT enabled JBD-adjacent BMS, and I got to pick the cells I wanted (REPT 280Ah).

heating

I use a 25w mat to keep the bank at ~50° F.

Theory:

there is solid evidence that LFP’s charging rate should effectively stop at freezing (“cold shut-off”)
there is also solid (but varying) evidence that LFP’s charging rate should decrease at lower temps
there is some evidence that LFP is happiest when protected from thermal shock (big current at low temps causing rapid internal temp rise)

… so I set it on the warm side.

Discussions on the DIY Solar forum suggest people are using minimum temps between 35F and 50F.

power use patterns

overnight

In summer tend to average -40Ah by morning

In winter longer nights mean more hours of lighting and colder temps mean running contact heat like the electric mattress pad. so maybe -70Ah in the mornings.

In freezing weather another ~10Ah will be needed for battery warming (see below).

In practice, it is a rare day when I do not hit my charging targets for the day.

daytime power use

I have copious solar so most of my daytime loads run off the panel

how I charge

Because of my camping patterns solar is the dominant charging source:

solar (90-95% of total harvest)
relay (5-10% of total harvest)
shore power - very rare

My overall approach is to charge at ~0.2C (max of 0.4C) at moderate voltages. Keep the bank warmed to 50F.

Note: for purposes of accuracy and low-temp cutoff, Voltages are taken from the battery with Victron’s Smart Battery Sense. It’s annoying that it’s a $40 BT device rather than a $5 wire, but that’s how Victron does things. Accurate voltage-sensing is a major upgrade in my system.

solar charging

normal use

In this scenario there is bank capacity well in surplus to needs. We can run the bank at middling SoC without worry of runnnig low.

Absorption voltage: 13.6v (3.4Vpc) Absorption duration: until charge acceptance falls to 0.05C.³. pseudo-Float: 13.4v (3.35Vpc)

Observed result: 13.6v until 0.05C gets the bank to little less than 80%. The 13.4v float very gently pushes up SoC as long as the sun shines. Depending on use patterns I end the day in the lower 80s.

full power, Scotty!

In this scenario our comfort margin, capacity-wise, is reduced due to reduced harvest, forecasts of reduced harvest in the near future, or any other case where power use threatens to run the bank low.

Absorption voltage: 13.8v (3.45Vpc) Absorption duration: until charge acceptance falls to 0.05. pseudo-Float: 13.3v (3.325Vpc)

Observed result: Mid-to-high 90% SoC, held stably at that point by the float.

while on shore power

Absorption voltage: ~13.2v (3.3Vpc) Absorption duration: n/a pseudo-Float: 13.2v

Observed result: stays somewhere in middling SoC, although the measurement drifts badly because the charge/discharge currents are often too small for the shunts to measure accurately.

resetting the BMS and shunt to 100%

pseudo-Equalization voltage: 14.0v (3.5Vpc)

Observed result: bank charged to ~100%, shunt[s] reset to 100%.

misc setpoints

temperature adjustment (lead-style) - off
low temp cutoff - 34F, although the battery warmer will likely mean this won’t happen.

alternator charging

I camp in place most of the time so the Battery Doctor isolator (VSR type) doesn’t see much action (5-10% of the total harvest). But when it does it works fine. Yes, isolators can charge lithium.

The Promaster’s OEM 180A alternator puts out ~14.2v, which is acceptable for LiFePO4 charging. My reading on the Promaster forum suggested the PM alternator charges 100Ah of LFP at 20-30A. That sounded about where I wanted the current to be so I gave it a shot before investing in a DC-DC charger like the Renogy 20A.

I was pleased to see normal current acceptance ~0.2C, with occasional forays to 0.32C. No need for DC-DC at this point. The isolator was already was on an ON/OFF switch from the original FLA bank so I can disable it at will.⁴ It is easily accessible from the driving position without looking. I turn if off at ≥13.7v or when combined solar+alternator charging is >0.4C.

shore power charging

13.2v at 15A (DIY converter limited to 15A).