Skip to content
zorowatt
Začni zdarma

Usable capacity, depth of discharge and round-trip efficiency: reading a C&I battery datasheet

Three things on a datasheet decide most of a battery study: the usable energy (not the nominal), the round-trip efficiency and where it was measured, and the warranty terms. They are rarely comparable between brands as printed. sonnen states 245 of 261 kWh usable, Huawei states 215 kWh rated with no usable figure, and efficiencies range from SMA's "up to 98%" for the battery alone to Huawei's 91.3% at the AC side including auxiliaries. Put every product on the same basis before you compare prices per kWh or size anything.

7 min read · Updated 2026-10-30

1. Nominal vs usable energy

Nominal (or rated) energy is what the cells hold; usable energy is what the battery management system lets you take out. The depth of discharge (DoD) is the ratio. Size and price on usable energy.

Illustrative: the same €100,000 is €100,000 ÷ 261 ≈ €383 per nominal kWh but €100,000 ÷ 245 ≈ €408 per usable kWh. Comparing one vendor's nominal price with another's usable price hides a difference of about 6.5%.

2. Round-trip efficiency, and where it was measured

Round-trip efficiency (RTE) is the energy you get out divided by the energy you put in. The number depends on the measuring point: at the battery's DC terminals, at the inverter's AC side, or at the grid connection, with or without the battery's own consumption.

  • SMA: "efficiency (battery) up to 98%", for the battery alone and not explicitly labelled round-trip (SMA datasheet).
  • sonnen: 93.75% total battery efficiency, with the inverter's maximum efficiency of 98.5% stated separately (sonnen datasheet).
  • Huawei: a maximum cycle efficiency of 91.3% at rated conditions (25 °C, 0.5 C, AC output); the Spanish version of the datasheet labels it including auxiliaries. It is a maximum, not a typical or lifetime value (Huawei datasheet).
  • Fox ESS: round-trip efficiency above 89%, without saying where it was measured (Fox ESS datasheet). Sungrow's product page gives a system RTE of 90% for the ST255CS (Sungrow).

A neutral default when a datasheet is unclear: PNNL's 2022 assessment uses 89.55% DC to DC for LFP, and applies an inverter efficiency of 0.98 in each direction, which gives 0.8955 × 0.98 × 0.98 ≈ 86.00% AC to AC at the inverter; with a transformer at 0.98 each way, about 82.6% (PNNL 2022). For a fleet sanity check, US utility-scale batteries averaged 82% monthly round-trip efficiency in 2019 (EIA, 2021).

3. The battery's own consumption

Cooling, the battery management system and the controller draw power all the time. In the MESA-1 field test of a 1 MWh system the auxiliary load was 30 kW, and measured round-trip efficiency ranged from 66% to 91% depending on power, rest periods and whether that load was counted; the report concludes that no single RTE represents a battery's performance (PNNL, MESA-1). A study of a 192 kWh LFP container found conversion efficiency of 70–80%, and 8–13 percentage points lower once system consumption was included (Schimpe et al. 2018).

No published auxiliary-load figure for C&I cabinets was found in Zorowatt's research. Ask the vendor for it in kW, and whether the stated efficiency includes it.

4. Power: the inverter and the cabinet

  • The usable kW is the lower of the inverter's and the battery's limit. SMA's battery allows up to 1 C, but the package's 50 kW inverter sets 50 kW (SMA datasheet).
  • SolarEdge's CSS-OD 197 comes with two 50 kW inverters, but the cabinet's maximum C-rate of 0.5 caps the pair at 98.5 kW (SolarEdge datasheet).
  • Check whether the kW is AC or DC, and whether a short peak rating (Fox ESS: 100 kW rated, 110 kW peak) is being quoted as the continuous one.

5. Cycle life vs warranted cycles

A cycle-life figure on a datasheet is a test result, often without its conditions; the warranty is the promise. They can be far apart.

  • SMA: battery cells "for up to 12000 charging cycles", with no test conditions; the warranty covers 8,500 cycles or 6.5 MWh per kWh of throughput (SMA datasheet). 6.5 MWh per kWh is 6,500 full cycles — fewer than the 8,500 also named, so ask which limit applies first.
  • Fox ESS: a cycle life of at least 8,000 cycles to 70% on the datasheet; the EU warranty is 10 years or 1,492 MWh of discharge throughput to 70% (Fox ESS warranty). 1,492 MWh ÷ 215 kWh ≈ 6,940 full cycles.
  • SolarEdge: 8,000 cycles or 10 years to 70% state of health, whichever comes first (SolarEdge datasheet).
  • sonnen: 8,000 cycles at 60% state of health (sonnen datasheet).

How a warranty becomes a degradation and replacement assumption is in Degradation, cycles and replacement: reading the warranty.

A checklist for every datasheet

  1. Usable kWh at the start of life, and the depth of discharge it assumes.
  2. Charge and discharge kW, continuous, AC or DC, and which part sets the limit.
  3. Round-trip efficiency: the measuring point, the test conditions, and whether auxiliaries are included.
  4. Auxiliary consumption in kW, idle and in operation, in summer and in winter.
  5. Warranty: years, cycles or throughput, end-of-life capacity, and the conditions (temperature, cloud connection, maintenance).
  6. Which documents the numbers come from, and their date.

Zorowatt's battery database records each of these values with the document it came from and the date it was read; a value no source published is shown as missing, never guessed.

Frequently asked questions

Which efficiency should a study use?
The AC-side round-trip efficiency including the battery's own consumption, if the vendor can give it. If not, state the assumption: PNNL's 2022 default stack gives about 86% AC to AC at the inverter for LFP, and a separate auxiliary load should be added.
Is a depth of discharge of 100% a warning sign?
Not by itself: it says the usable window equals the rated energy. What matters is how the warranty counts cycles and at what capacity it ends. PNNL's 2022 table for LFP gives 2,400 cycles at 80% average depth of discharge and 8,000 at 60%, to 60% remaining energy.
Why does the battery database show some values as missing?
Because no datasheet, manufacturer page or other source published them when the database was compiled. Zorowatt does not fill specs with estimates; ask the vendor and note the answer in the study.
Are datasheet values guaranteed?
No. A datasheet describes the product under test conditions; the warranty is the commitment. Use the warranty's terms for the financial case and the datasheet for the physics.

Related guides

Sources

Last checked: 2026-10-30. Every figure in this guide comes from the source linked next to it, with the retrieval date Zorowatt's research recorded (28–30 October 2026), or from arithmetic shown step by step. Product values as recorded in Zorowatt's battery database (retrieved 28 October 2026); the per-kWh price example is illustrative.

Stuck on something else? Ask the community.

Zeptat se

Nové návody přímo do tvé schránky

Zhruba jeden e-mail měsíčně: nové návody a videa, řešení problémů, na které se lidé ptají, a novinky ze Zorowattu.

Zhruba jeden e-mail měsíčně s novými návody, videi a novinkami ze Zorowattu. Nejdřív ti e-mailem pošleme odkaz pro potvrzení a odhlásit se můžeš kdykoli. Zásady ochrany osobních údajů

Size it right with Zorowatt

Upload a site's meter data and bill. Zorowatt checks the data, rebuilds the bill, sizes a battery and writes the report. Your first site analysis is free.

More guides