“What does a megawatt-hour of storage cost?” is the most frequently asked question in the BESS market — and the most frequently answered incorrectly. Because the price of the battery system is only one line in the investment calculation. This benchmark breaks down the CAPEX of a utility-scale project into its components and shows which factors drive the range.
The cost structure of a BESS project
For a turnkey 2-hour system in Germany, total investment costs in 2026 are indicatively in the range of 350–550 €/kWh. For 4-hour systems, the specific value drops to about 280–450 €/kWh, because grid and civil costs are spread over more capacity. All figures are indicative, site- and project-dependent and do not constitute a basis for offers.
| Cost block | Share of CAPEX (indicative) | What drives it |
|---|---|---|
| Battery system (DC blocks, containers) | 35–45 % | Cell prices, chemistry (LFP standard), warranty scope, origin |
| PCS & transformer (AC side) | 10–15 % | Power class, voltage level, delivery times |
| Grid connection | 10–20 % | Distance to connection point, grid operator, reinforcement measures |
| Civil works, foundations, cable routes | 8–12 % | Soil, plot layout, fire safety distances |
| Engineering, permitting, project management | 5–10 % | Complexity, zoning plan status, expert reports |
| EMS, SCADA, fire protection, auxiliary systems | 5–8 % | Functional scope, interfaces to the optimizer |
| Development costs (up to RTB) | separate | Site securing, grid application, permitting process |
What explains the range
Grid connection costs are the most underestimated item. Between a site with a connection point at the plot boundary and one requiring kilometres of cable route lie seven-figure amounts. They often only become reliable after the grid study — one reason why grid connection must be structured early.
Storage duration changes the €/kWh logic. More MWh at the same MW power spreads PCS, grid connection and civil works over more capacity. 4-hour systems therefore achieve significantly lower specific costs — but must refinance the additional cells through the revenue model.
Warranty and service cost — and pay off. Comprehensive warranty packages (capacity guarantee, availability guarantee, augmentation) increase the entry price but reduce risk over 15–20 years of operation and improve debt financeability.
Cell prices are volatile. The significant price reductions of 2023–2025 made projects economically viable in the first place. Those who calculate should work with ranges and sensitivities, not point values.
CAPEX is only half the equation
A low CAPEX with a weak grid connection or unsuitable sizing is no advantage. What matters is the ratio of investment, achievable revenues (see our article on revenue stacking) and lifetime. Levelized Cost of Storage (LCOES) and the debt service coverage ratios (DSCR) used by financiers are therefore more suitable for evaluation than €/kWh.
Three practical rules
First: always compare total CAPEX. Offers that only show the battery system are not comparable with turnkey offers.
Second: demand sensitivities. Serious calculations show what happens at ±10% CAPEX, changed cell prices or a shifted COD.
Third: calculate CAPEX against revenues, not against the market average. A more expensive system with better efficiency and longer warranty can be the cheaper one over its lifetime.
Conclusion
In 2026, CAPEX for utility-scale BESS in Germany is in a range that fundamentally makes projects economical — but the spread within a single project is larger than the market movement. Those who structure grid connection, sizing and warranty architecture early calculate more reliably and negotiate better.