Choosing a Battery Chemistry: LiFePO4, NMC, LTO and Lead-Acid in Practice

Nearly every power project we are asked to quote begins with the same question: which battery chemistry should we use? It is usually the wrong first question. Chemistry is a consequence of the duty cycle, the environment and the service life you need — not a starting preference. Below is the framework we apply when specifying a pack, and an honest account of where each of the four common chemistries earns its place.

Define the duty cycle first

Four numbers decide most of it: depth and rate of discharge (C-rate), cycles per year, the ambient temperature range the pack will actually see, and the calendar life you are committing to. A UPS bank that discharges fully twice a year is a completely different animal from a solar self-consumption pack cycling once a day for fifteen years, even if both are quoted at the same kWh. Get those four numbers on paper before anyone opens a catalogue — they eliminate most of the options by themselves.

Lead-acid (VRLA, AGM, gel): still the default for standby

Lead-acid remains genuinely competitive wherever the pack sits float-charged and rarely cycles: telecom cabinets, fire and security panels, switchgear tripping supplies, small DC-UPS. It is inexpensive per installed kWh, tolerant of simple constant-voltage charging, universally recyclable, and its failure modes are familiar to every service technician in the field. The weaknesses are equally well known — usable depth of discharge is realistically around 50 percent, cycle life falls off sharply with both depth and temperature, and sustained heat above roughly 25 °C shortens service life considerably. If the pack cycles daily, lead-acid rarely wins on total cost of ownership even though it wins on purchase price.

LiFePO4 (LFP): the workhorse for cycling applications

LFP has become the default for almost anything that cycles: battery energy storage systems, solar self-consumption, motive power, mobile equipment, and increasingly as a UPS replacement. It delivers thousands of cycles at deep discharge, holds a flat voltage plateau across most of its usable range, accepts high charge rates, and has the most benign thermal behaviour of the mainstream lithium chemistries. The trade-offs are real but manageable. Energy density is lower than NMC. Charge acceptance below 0 °C is poor — charging a cold LFP cell plates metallic lithium and causes permanent damage, so a low-temperature charge lockout is not optional. And that same flat voltage curve makes state-of-charge estimation from voltage alone unreliable, which is precisely why coulomb counting and a properly designed BMS are requirements rather than refinements.

NMC: when mass and volume dominate

NMC buys energy density — more watt-hours per kilogram and per litre than LFP — which matters in vehicles, portable and wearable medical devices, airborne equipment and anywhere the enclosure dimensions are fixed and non-negotiable. The cost is a narrower safety margin: a lower thermal runaway onset temperature and a more energetic failure mode. That shifts weight in the design review toward mechanical protection, thermal management and a BMS with genuine fault handling rather than mere reporting. Where mass and space are not binding constraints, we generally recommend LFP instead.

LTO: the specialist

Lithium titanate is expensive and energy-poor, and it is still the right answer in a narrow band of applications: very high cycle counts, fast charge and discharge at high C-rates, and operation at low temperatures where other lithium chemistries simply cannot be charged. Rail, grid frequency support, port and industrial equipment on short repetitive duty cycles are the typical homes. If a requirement reads "charge in minutes, cycle several times a day, for a decade", LTO deserves evaluation despite the price per kWh.

The decisions that matter more than chemistry

In our experience most pack failures are not chemistry failures. They come from undersized conductors and loose terminations, cells drawn from mixed batches with no matching, an enclosure designed with no thought given to heat, a charger profile that does not match the cells, or a monitoring system that reports state of charge but cannot actually protect anything. Specify the balancing strategy, the temperature sensing points, the communications interface and the maintenance access at the same time you specify the cells — every one of those is expensive to retrofit and cheap to get right at the design stage.

One last point that is easy to defer: plan the end of life at the beginning. Chemistry determines the recycling route, the transport classification and the second-life potential of the pack, and those are commercial questions as much as technical ones.

If you are weighing chemistries for a specific application and want a second opinion on the duty cycle before the specification is frozen, write to us at info@lmelectro.com. We respond within 24 hours.