Flexa
German Home Battery Market Explorer
MaStR bulk export · 2026-10-14
0–50 kWh residential · status: in operation
Marktstammdatenregister · brand-signature model
Reach out to Flexa Business Development team for more information.

Distribution by capacity band (full 0–50 kWh)

Static split of the selected installation period across capacity bands — independent of the slider.

Brands in range

Avg PV kWp / kWh/kWp = paired rooftop PV per battery (matched by household owner + postcode; ~98% of home batteries pair with PV). kWh/kWp = battery capacity ÷ PV, capacity-weighted.

Capacity share

Capacity-weighted share of the selected installation period & capacity range.

By Bundesland

Units & capacity of the selected range/period per federal state; top brand by fractional capacity.

Methodology

Data & scope

German Marktstammdatenregister (MaStR) bulk export, 2026-10-14. Residential storage 0–50 kWh, status “in operation” — 2.81M systems, 24.9 GWh total.

Installation period

The dataset is the cumulative stock (registrations 2012–2026). “2024 / 2025 installs” filter to systems registered that year, comparable to the annual RWTH Aachen / P3 (pv-magazine) market reports.

Brand assignment

MaStR has no manufacturer field. Each system is assigned a brand from: (1) an explicit brand in its free-text name (~8% of systems); otherwise (2) its (power, capacity) signature, learned from that labeled subset — a market-unique signature maps to one brand, a shared signature is split proportionally; (3) capacities never seen labeled are “not identified”. Shares are capacity-weighted and calibrated against the RWTH/P3 reports (≈1.1 pt mean deviation on 2025, 0.9 on 2024).

Why “fractional” shares

Many systems can’t be pinned to one brand from their (power, capacity) alone — the same specs are sold by several manufacturers. When a unit “could be BYD or Sungrow,” there are two ways to count it:

  • Hard = winner takes all. Give the whole unit to its single most-likely brand.
  • Fractional = split it. Give each candidate a slice by how likely it is (e.g. 60% BYD, 40% Sungrow).

This explorer uses fractional throughout — the same method the RWTH/P3 reports use — so the figures are directly comparable.