
LCOS is not a property of a chemistry. It is capital cost divided by the energy a system moves over its life, adjusted for round-trip losses and replacements.
The formula without ceremony
Total capital cost, plus the present value of replacements and O&M, divided by lifetime discharged energy. That energy is annual throughput times years, times efficiency, times the capacity still available in the final year.
A cheaper pack with a short life often loses to a dearer pack that cycles more. That is the whole argument for sodium in daily-cycling duty.
What moves the number
Cycles per year: one versus two cycles changes LCOS by nearly half. Depth of discharge: deeper uses more capacity but shortens life, and the optimum is usually below 100%.
Capex: the pack is 40–60% of a cabinet, so a 20% cell price difference is about a 10% LCOS difference. The tariff spread does not change LCOS — it decides whether the project exists.
Indicative 2026 ranges
| Sodium (NFPP) | LFP | Lead-acid | |
|---|---|---|---|
| Capex per kWh | 150–220 USD | 140–200 USD | 90–130 USD |
| Cycles at 80% DoD | 4,000–8,000 | 3,000–6,000 | 400–800 |
| Round-trip efficiency | 88–92% | 92–95% | 75–85% |
These move quarterly. The interesting column is the last one: lead-acid is cheap to buy and expensive per cycle.
How to use it
Build the model with your own tariff and cycles, then test sensitivity. If a 20% change in cycle life flips the decision, chemistry did not decide it.
If the project only pays back on optimistic assumptions, it does not pay back.
Cycle numbers to model with
Use the measured set rather than a brochure figure: 1,000 cycles at 1C on a 0.9 Ah high-rate cell, 2,000 at 0.5C on a 7 Ah pouch, 3,000 on an 18 Ah cylindrical, 4,000 at 0.5C on a 55 Ah prismatic. Put the one that matches your duty into the model — the spread between them is a factor of four in lifetime throughput.
Where to go next
Questions we get asked about this
What LCOS should I expect in 2026?
For daily-cycling C&I duty, published modelling commonly lands around 0.10–0.18 USD/kWh depending on tariff, cycles and capex.
Which assumption moves LCOS most?
Cycles per year, then capex, then efficiency. Doubling daily cycles nearly halves LCOS.

