Behind every direct participation in a grid-scale battery stands a physical real asset: a battery system of thousands of cells, inverters, cooling and control. Its technical properties (cell chemistry, cycle life, degradation, warranties) are not spec-sheet trivia but the levers that determine the cash flow of the next 15 to 20 years. An investor does not buy a return, but a machine that has to earn that return over two decades. For an overview of the routes into such an investment, see Investing in battery storage: the options at a glance; how the overall plant is built, from container to grid connection, is explained in How does a grid-scale battery storage system work? Design, operation and revenue logic explained.
This article translates the technology into what matters for investors: what does each metric mean for the return, what for the risk, and which question should you ask any provider about it? It describes orders of magnitude that hold for the market, not a specific project. The technical risks themselves (degradation, fire, outage) are placed in the wider risk framework by Risks in BESS direct investments, and how they are structurally addressed; here we look at the physical substance behind them.
LFP or NMC: Which battery cell is inside a grid-scale storage system?
German grid-scale batteries today almost always use LFP, lithium iron phosphate. The vast majority of newly built grid-scale storage systems worldwide use this chemistry; NMC (nickel manganese cobalt) is the cell of the electric car, where high energy density per kilogram matters. For a stationary system that sits on open land and moves no weight, energy density is secondary; what counts is lifespan, safety and cost. That is exactly where LFP plays to its strengths: more full cycles, markedly better thermal stability and a lower price per kilowatt-hour, because no expensive cobalt or nickel is needed. For investors this means: the dominant chemistry is also the longer-lived and safer one. A system that opts for NMC instead needs a good reason to do so.
| Attribute | LFP (lithium iron phosphate) | NMC (nickel manganese cobalt) |
|---|---|---|
| Energy density | lower, uncritical for stationary storage | higher, an advantage mainly in vehicles |
| Cycle life | high: around 6,000 to over 10,000 full cycles | lower: around 3,000 to 5,000 full cycles |
| Thermal safety | very stable, high threshold for thermal runaway | more sensitive, higher risk of thermal runaway |
| Cost per kWh | lower, no cobalt/nickel required | higher, cobalt and nickel content |
| Raw-material dependency | low (iron, phosphate) | high (cobalt, nickel) |
| Suitability for grid-scale | very high, today's standard | limited, niche |
How long does a grid-scale battery last, and what are cycles?
A grid-scale battery has two lifespans that should not be confused. The calendar life describes how long the system ages regardless of use; for modern LFP systems 15 to 20 years are assumed. The cycle life describes how often it can be charged and discharged until the guaranteed remaining capacity is reached. A full cycle is one complete charge and discharge; two half cycles count as one. A system that cycles once or twice a day in arbitrage trading reaches roughly 7,000 to 14,000 full cycles over 20 years, exactly the order of magnitude LFP cells are designed for. For investors this is the true useful life: not the calendar alone, but the cycles and the trading strategy determine how quickly the system exhausts its design. How these cycles become revenue is shown by Battery storage returns: where the revenue comes from, and what is realistic.
What is degradation and how fast does the system lose capacity?
Degradation is the planned loss of capacity of a battery over time. It has two causes: calendar ageing (the cell ages even at rest) and cyclic ageing (every charge and discharge costs a little capacity). The curve is not linear but steeper at first and then flattens; the largest single step happens in the first year of operation. Under good operating conditions, the degradation of an LFP grid-scale system is around 2 to 3 percent per year, so that after ten years roughly 75 to 85 percent of the initial capacity is still usable. Under demanding conditions it can be considerably higher:
- Favourable: operation at around 20–30 °C, moderate charge rates (below 1 C), active cooling, no permanent 100-percent cycles, around 2–3 % loss per year.
- Demanding: sustained above 35 °C, high charge rates, constant full cycles without buffer, 5–7 % per year, leaving only 50–60 % remaining capacity after ten years.
For investors, the key point is that degradation is not a surprise risk but a calculable quantity: in a sound economic model, the declining capacity curve is factored in from the outset. The risk lies not in degradation itself, but in a model that glosses over it; anyone who assumes constant capacity over 20 years systematically overstates the returns of the later years. Which operating conditions a project creates is, moreover, not a matter of chance but of location, cooling concept and operations management, and thus a mark of quality. The assumed degradation curve therefore belongs among the assumptions a provider should disclose. How to check that assumption in the yield or revenue assessment is shown in Reading a yield assessment: what P50 and P90 mean for a solar park direct investment.
What does augmentation mean?
Augmentation refers to the targeted addition of battery capacity during the life of the system in order to keep the usable storage size constant despite degradation. Because capacity declines as planned, a system without augmentation would deliver less energy after a few years than at the start; augmentation adds further cell modules and compensates for the loss. For investors this is no side detail but a real cost item: augmentation causes additional investment later on. The decisive questions are whether these costs are accounted for in the economic model, when they arise and who bears them. A model that promises constant capacity but conceals the augmentation costs is incomplete. How a system is operated and monitored over the years is described by What happens after closing: reporting, asset management and why a partner is not a broker.
What do capacity and throughput warranties secure?
Warranties convert technical risk into a contractual floor; that is why they are the most important substance document of a participation. Two types matter. The capacity warranty guarantees that after a given time or cycle count the battery still holds a minimum share of its initial capacity; typical orders of magnitude are around 10 years or 10,000 cycles at at least 70 to 80 percent remaining capacity (state of health, SoH); the concrete values are negotiated per project. The throughput warranty guarantees an amount of energy or a cycle count the cell must deliver over its life. If the battery falls below the promised values, the manufacturer steps in.
For investors it comes down to the fine print, not the headline. A warranty is only as strong as its conditions: temperature windows, maximum charge rates, annual cycle caps and exclusions can hollow it out when it counts. And it is only as good as the manufacturer standing behind it; a 15-year warranty from a manufacturer whose survival over 15 years is uncertain carries a counterparty risk. The warranty term, the SoH threshold and the creditworthiness of the guarantor therefore belong examined together.