Every off-grid solar system needs a battery, and the battery chemistry you choose determines everything: how long the system lasts, how safe it is, how it performs in the heat, and what it costs over its lifetime. At Ohm Network, we standardised on lithium iron phosphate (LiFePO₄, or LFP) after evaluating every viable option on the market. Here is why.
Safety first: no thermal runaway
The most important difference between LFP and other lithium-ion chemistries is thermal stability. LFP cathodes do not release oxygen when overheated, which means they cannot sustain a thermal runaway event — the self-accelerating reaction that can cause batteries to catch fire.
In a comprehensive safety review published in The Electrochemical Society Interface (Doughty & Roth, 2012), NMC cathodes were shown to begin exothermic decomposition at approximately 180 °C, releasing oxygen that can fuel a self-sustaining fire. LFP cathodes, by contrast, remain structurally stable past 270 °C. The strong covalent phosphorus–oxygen bonds in the phosphate polyanion prevent oxygen release even at elevated temperatures, depriving thermal runaway of its fuel source.
For a solar system installed in a home in Lagos, a shop in Accra, or a clinic in rural Tanzania — places where ambient temperatures regularly exceed 35 °C and professional fire suppression is not down the street — this safety margin is non-negotiable.
Cycle life: what the research actually says
Our LFP batteries are conservatively rated for 3,000 full charge-discharge cycles while retaining at least 80% of their original capacity. At one cycle per day, that is over eight years of daily use before you notice any meaningful degradation. But the real story is more interesting — and more favourable to LFP — than a single number suggests.
Laboratory data
A landmark 2020 study by Preger et al. in the Journal of The Electrochemical Society systematically compared degradation across LFP, NMC, and NCA cells under identical cycling conditions. The findings were striking: at 25 °C with 1C charge/discharge and cycling between 0% and 100% state of charge, LFP cells retained over 80% capacity after more than 4,000 equivalent full cycles. NMC cells under the same conditions fell below 80% after 600–1,200 cycles, depending on the specific nickel content and cell design.
Wang et al. (2014), reporting in the Journal of Power Sources, documented LFP cells exceeding 4,500 cycles at 100% depth of discharge before reaching the 80% capacity threshold. The degradation mechanism in LFP was found to be dominated by gradual loss of cyclable lithium inventory — a predictable, linear fade — rather than the cathode structural collapse and impedance growth that accelerates failure in NMC cells.
Real-world solar cycling
These laboratory numbers actually understate LFP's advantage in solar applications. Here is why: in a solar system, the battery rarely cycles from fully charged to fully discharged. Instead, it operates in a partial state of charge — typically discharging from near 100% down to perhaps 40% during the night, then recharging during the day.
Research by Saxena et al. (2016) in the Journal of Power Sources demonstrated that cycling depth dramatically affects longevity. When LFP cells were cycled between 30% and 70% state of charge (a 40% depth of discharge window), the equivalent full-cycle life tripled compared to full 0–100% cycling. In other words, if a cell is rated for 3,000 full cycles, it can deliver the equivalent of 9,000 full cycles — or 36,000 partial cycles — when operated conservatively.
This aligns with field data from our installations: we have LFP packs in daily use that show less than 5% capacity loss after four years of operation, which projects to well over 5,000 equivalent full cycles before reaching 80% retention.
Temperature effects on cycle life
Ambient temperature is the single largest external factor governing battery longevity. Preger et al. found that raising the cycling temperature from 25 °C to 45 °C roughly halved the cycle life of NMC cells, while LFP cells lost about 30% of their cycle life under the same temperature increase. This means that in the hot environments typical of our target markets, the cycle life gap between LFP and NMC widens further — LFP can deliver 2,500–3,000 cycles to 80% at 40 °C, while NMC may only manage 400–600.
For context: a lead-acid battery in the same conditions will typically last 200–400 cycles. The replacement cost and labour involved in swapping lead-acid batteries every one to two years makes them a false economy for any system expected to operate beyond the short term.
Sizing for discharge rate: LFP vs NMC
One area where NMC holds an advantage over LFP is continuous discharge rate (C-rate). NMC cells commonly support 3C to 5C continuous discharge — meaning a 5 kWh NMC pack can deliver 15–25 kW of power. High-performance NMC power cells, of the type used in power tools and some electric vehicles, can sustain 10C or higher.
LFP cells are typically rated for 0.5C to 1C continuous. This is a consequence of the chemistry: LFP has lower electronic conductivity than NMC, and lithium diffusion within the olivine crystal structure is slower, which limits how quickly ions can move in and out of the cathode.
For a solar home system powering lights, a television, a refrigerator, phone chargers and fans or even a small AC unit, the continuous load is rarely above 2 kW, and a 0.5C rate on a 5 kWh LFP battery provides 2.5 kW — more than sufficient. This is why LFP is an excellent match for residential and small commercial solar storage.
However, some use cases demand higher instantaneous power. A borehole pump, a welding machine, or a powerful motor can require 5–10 kW of startup surge current. In these scenarios, an NMC pack could meet the demand with a smaller (and cheaper) battery bank, while an LFP pack would need to be oversized — sometimes by a factor of two or three — to deliver the same peak power without exceeding its rated C-rate.
The practical implication: if you are sizing an LFP-based solar system, you must calculate the peak power demand in addition to the total energy requirement, and size for whichever constraint is binding. In most residential solar contexts, energy (kWh) is the binding constraint. In systems with large motors, power (kW) may dominate, and oversizing an LFP bank is the correct engineering decision. We help each customer run both calculations during system design.
No cobalt, no conflict
LFP cathodes contain no cobalt. This matters for two reasons.
First, cobalt mining — particularly in the Democratic Republic of the Congo, which supplies approximately 70% of the world's cobalt — has been linked to serious human rights concerns, including child labour and unsafe working conditions documented by Amnesty International and other observers. By choosing LFP, we avoid contributing to that supply chain entirely.
Second, cobalt is expensive and its price is volatile. In 2018, cobalt traded above $90,000 per metric tonne; by 2019 it had fallen below $30,000, and it has continued to fluctuate sharply. Removing it from the chemistry lowers the raw material cost and insulates battery pricing from geopolitical shocks. That cost advantage flows directly to our customers.
Built for hot climates
LFP performs better at high temperatures than other lithium chemistries. Where NMC cells lose capacity and degrade faster above 40 °C, LFP holds its own. The degradation acceleration factor with temperature — approximately 2× for every 10 °C rise for most lithium chemistries — is measurably lower for LFP, as documented in the Preger et al. (2020) degradation study.
Combined with the inherent thermal stability mentioned above, this makes LFP the obvious choice for Africa, South America, and Southeast Asia — precisely the markets we serve.
We do not need active cooling systems that add cost, complexity, and a parasitic load on the very battery they are meant to protect. A well-designed passive enclosure is sufficient for LFP in all but the most extreme environments.
Voltage behaviour: flat vs steep
LFP cells have a remarkably flat discharge curve — they hold roughly 3.20–3.30 V per cell from 80% state of charge down to 20%, with a voltage change of less than 0.1 V across that entire range. This makes them easy to integrate with inverters and charge controllers because the voltage stays predictable. It also means your appliances see stable power throughout the discharge cycle, unlike lead-acid batteries whose voltage sags significantly as they empty.
NMC cells, by contrast, have a steeper and more sloped discharge curve. A typical NMC cell starts near 4.2 V fully charged and drops continuously to around 3.4–3.5 V by 20% state of charge — a voltage swing of roughly 0.7–0.8 V across the usable range. The slope in the middle SOC region is approximately 3–4 mV per percentage point of SOC, compared to LFP's near-zero slope of roughly 0.3–0.5 mV per percentage point in its flat region.
Each voltage profile has engineering implications. The steep NMC curve makes state-of-charge estimation straightforward — a simple voltage measurement gives you a reasonably accurate SOC reading, which is why consumer electronics often rely on voltage-based fuel gauges. The flat LFP curve makes voltage-based SOC estimation nearly impossible in the middle range; a 5 mV measurement error translates to a 10% SOC error with LFP, versus roughly 1–2% with NMC. This is why LFP systems require coulomb counting (tracking current in and out over time) rather than voltage-based gauging.
For the end user, however, the flat LFP curve is a benefit: your inverter delivers consistent power at a predictable voltage from sunset to sunrise. With NMC, the same inverter would need to handle a wider input voltage range, and your lights might dim noticeably as the battery discharges unless the inverter compensates.
The bottom line
We chose LFP because it is the safest, longest-lasting, and most ethically sourced lithium chemistry available at scale. The research is unambiguous: LFP delivers 3–6× the cycle life of NMC under real-world solar cycling conditions, widens its advantage in hot climates, and provides inherently safer chemistry with no thermal runaway risk.
For the discharge-rate limitations — real but narrowly applicable — we size systems accordingly during design. For the flat voltage curve that complicates SOC estimation, we use coulomb counting with periodic recalibration at full charge. These are solvable engineering problems, well worth accepting for the safety and longevity gains.
For customers who depend on their solar system every day — to keep the lights on, the refrigerator running, and the business open — those qualities are not optional. They are the whole point.