The chemistry sets the starting point for battery life. For stationary storage, the market has made its choice.
Lithium iron phosphate (LFP) and lithium nickel manganese cobalt oxide (NMC) are the two lithium-ion chemistries most buyers compare. They age differently, respond to heat differently, and fail differently, and those differences decide which one makes sense for a battery that has to last a decade or more. This guide compares them on what matters for battery life, explains why LFP now dominates new stationary storage, and looks briefly at LTO and sodium-ion.
The clearest public comparison comes from Sandia National Laboratories, which cycled commercial LFP, NMC and NCA cells for several years under identical conditions. The LFP cells reached 80% capacity after 2,500 to 9,000 equivalent full cycles, against 200 to 2,500 for NMC and 250 to 1,500 for NCA. At the same number of cycles, LFP cells had retained on average 9% more capacity than NMC.
Two caveats matter. First, these were small consumer-format cells; large cells built for energy storage carry much higher cycle ratings today. Second, Sandia notes the gap narrows when cells are compared by total energy delivered rather than cycles, because each LFP cell stores less energy. Even so, the direction is consistent across the literature: LFP is the longer-lived chemistry for daily cycling.
The chemistries also respond differently to the stress factors that age a battery, which is why one set of operating rules does not fit both.
Sandia's handbook explains the difference simply: in LFP, the phosphorus-oxygen bond is strong and the cathode does not release oxygen as it degrades. Metal-oxide cathodes such as NMC can, which means a cell in thermal runaway contains its own oxygen supply.
Laboratory tests on commercial cells show the effect. In a 2014 study published in RSC Advances, an LFP cell entered thermal runaway at around 195 degrees against 168 for an NMC cell, peaked at around 404 degrees against 678, and released roughly a third as much gas. LFP is not immune to thermal runaway, but its runaway is milder, and that matters for a container of thousands of cells sitting in Indian heat.
The International Energy Agency reported that LFP took 80% of new battery storage in 2023, and around 90% of deployments by 2025, up from well below half five years earlier. Its reasons: LFP is cheaper, better suited to frequent cycling, less flammable and longer-lived, and contains no nickel or cobalt. In a February 2026 commentary, the IEA put LFP batteries at more than 40% cheaper on average than NMC.
The one big advantage NMC keeps is energy density: more energy in less space and weight. That is decisive in a car and much less so in a container on a concrete pad. NREL's Annual Technology Baseline likewise treats LFP as the primary chemistry for stationary storage from 2022.
An LFP cell's voltage barely changes across most of its charge range. That makes it good at delivering steady power, but it makes state of charge harder to read from voltage, and Sandia notes it makes cell balancing harder during long periods at partial charge. LFP systems need periodic recalibration, typically by reaching full charge, and a battery management system tuned for the chemistry.
Two other chemistries come up in storage conversations.
Lithium iron phosphate
Long cycle life, a phosphate cathode that does not release oxygen as it breaks down, and no nickel or cobalt. It stores less energy per kilogram, which matters far less in a container than in a car. Like NMC, its graphite anode must not be charged fast in the cold.
The default choice for new stationary storage.
Datasheet cycle life figures are only comparable when their conditions match. One manufacturer's 280 Ah LFP storage cell, for example, is rated for at least 6,000 cycles at 25 degrees but at least 2,500 cycles at 45 degrees, both at 0.5C and to 80% of initial capacity. Change the temperature, the C-rate, the depth of discharge or the end-of-life threshold, and the headline number moves a long way.
Ingro EMS and Ingro Brain do not treat every battery the same. Operating limits, temperature targets and state-of-charge windows are set for the chemistry and the cell on site, and Ingro Brain learns separately how each chemistry ages, from the site's own data and from batteries of the same chemistry across every site Ingro runs.
That matters when you are choosing a battery, too: Ingro's engineers can tell you how a chemistry behaves under Indian heat and duty cycles, not just what its datasheet says at 25 degrees.
Battery health analytics and AI forecasting: state of health, early warnings before failure, and dispatch tuned to make the battery last.
4 questions
In most published comparisons, yes. Sandia's multi-year tests found LFP cells reached 80% capacity after 2,500 to 9,000 equivalent full cycles, against 200 to 2,500 for NMC, under the same conditions. The gap is smaller when measured by total energy delivered.
LFP is more thermally stable because its cathode does not release oxygen as it degrades. In laboratory tests it enters thermal runaway at a higher temperature, reaches a much lower peak temperature and releases less gas than NMC. It is not immune, so safety still depends on design, monitoring and fire protection.
LFP is cheaper, longer-lived, better suited to frequent cycling and more thermally stable than NMC, and contains no nickel or cobalt. Its lower energy density matters little in a stationary container. The IEA reported it took around 90% of new storage deployments by 2025.
Not yet for most uses. Sodium-ion's main advantage is performance in very cold climates, and the IEA reports it is not yet cheaper than LFP in most applications at current lithium prices. It is worth watching as field data and manufacturing scale grow.
Ingro Brain watches every cell and Ingro EMS runs the battery with its health in mind, on new projects and on batteries already in service.