DC block configuration
The DC block's four sides, how blocks stack into clusters, and what the 2×2 / 2×1 / 1×1 setting constrains.
The DC block is the containerized battery unit — a standard 20 ft container, 6.058 × 2.438 m, holding 5–6.25 MWh depending on the manufacturer. How blocks combine into clusters drives everything else in the layout: the PCS & MV skid each cluster needs, the aisles around it, and ultimately how much energy fits on the site.
The four sides of a DC block
A DC block is not symmetric. Each side has a role, and each role has a clearance:
Manufacturers offer A and B mirror variantsso two blocks can pair with their service sides facing outward. The cable end is on the same short end for both; the mirror swaps which long side carries the doors. The solver places A/B pairs automatically — that's why paired blocks in a layout always show their doors to the outside of the pair.
The supported design
DC blocks are not all built alike — manufacturers vary where the cooling sits, which sides carry doors, and which sides can stack. The tool currently models one archetype: the typical current-generation liquid-cooled 20 ft design— the cooling and connection section on one short end, service doors along the long side, and blind sides that stack into Solo, Twin, and Quad clusters. Jinko's SunTera G3 (6.25 MWh) is a representative example, and the tool's default dimensions and energy match that class exactly — see Supported equipment for the full reference set.
Designs with different cooling and access arrangements pack differently, so they need their own models — support for additional DC block designs, and for selecting specific manufacturer equipment, is on the roadmap. Until then, if your project uses a block of this archetype with slightly different clearances, adjust the service distances in Site details.
Clusters: how blocks stack
Blocks stack only on their blind sides, with the 0.2 m internal gap between them. Service clearance then applies around the cluster's outer boundary, not between the blocks inside it — this is what makes stacking denser than placing singles.
Two rules are never violated: an S2S stack always joins blind short sides — the cable ends face outward, toward the PCS; and an L2L stack always joins blind long sides — the doors face outward. If a layout ever hid a door or a cable connection inside a stack, the equipment would be unserviceable, so the solver treats these as hard constraints, not preferences.
What the DC config setting does
The DC config picker constrains which cluster sizes the solver may use:
- Auto — the solver chooses per site. Larger clusters pack denser, so it prefers 2×2 where space allows and steps down to 2×1 or 1×1 where a zone is too small for a quad.
- 2×2 / 2×1 / 1×1 — restricts the solver to that cluster size and smaller. A forced choice decomposes downward (a 2×2 site may still use twins in a leftover strip) but never escalates — forcing 2×1 guarantees no quad appears anywhere.
The 2×2 is available at both durations, but the skid count changes. A PCS & MV skid serves a fixed power, so the DC blocks it can carry scale with duration: at 2h a quad's four blocks draw more power than a single 40 ft skid can convert. At 4h the 2×2 uses one40 ft skid (the one-skid Quad); at 2h it builds as Quad Split — two skids, one on each DC row, two blocks per skid — so the 2×2 footprint stays available with correct PCS sizing.
Forcing a cluster size is the right move when your project standard, supplier, or O&M concept is built around one configuration. Otherwise leave it on Auto: the per-zone fallback is what lets an irregular boundary use all of its space.
Why this matters for density
Every cluster needs its own PCS & MV skid and its own service perimeter. Fewer, larger clusters mean fewer skids and less clearance area per MWh — which is why the same boundary can hold meaningfully more energy at 2×2 than at 1×1. The skid pairing per cluster size — and the defaults Auto applies — is covered in BESS section configurations.