Plan the building in the 3D editor

The 3D editor (Plan3D) opens in Building mode when the scene is empty. This is where you give Reonic a faithful 3D model of the roof — outline, tilt(s), obstacles. Most projects start from an auto-seed (Building AI, or a photogrammetry drone model), but review every auto-seed before you move on to module placement.

This guide covers Building mode end-to-end. For module placement, stringing, and the simulation, see Plan PV modules in the 3D editor.

Before you start

• The offer must be created and unsigned. After signature the editor is read-only — to redo the geometry, fork to a new variant.
• Have the building's address right on the project. The editor centres on the project coordinates; if the address is wrong the satellite imagery is off-target.
• Know which roof shape you're modelling: pitched (gable, hip, mansard, mixed) → Smart Roof; flat or single-tilt → Flat Roof.
• Know your obstacle inventory — chimneys, skylights, dormers, satellite dishes, vents — before you start. You can add them iteratively, but drawing the outline and obstacles in one pass is faster than going back.
• Start the 3D scene from the single 3D with Google Data card at Edit Solar Setup — it's one card, not a separate map-source choice. The card automatically picks the best available 3D imagery for the address. You don't pick which one — the card decides at open time. This choice is baked into the saved scene and shapes how the AI obstacle detection runs later (see Things to know → Start mode is a one-shot decision).

Open the 3D editor and switch to Building mode

1. Open the offer in the Portal.
2. Go to the Planning tab.
3. Click Start planning on the PV system.
4. Pick a planning mode that uses the 3D editor (any mode other than Quick planning). The classic 3D planning lives in the same place — pick the standard PV planning mode and the editor opens.
5. Click Begin planning.
6. The editor opens. The top toolbar shows three phases — Building planning, Module planning, String planning. Empty scenes default to Building.

If the scene already has roofs (auto-seeded or from a previous session) but you want to revisit geometry, click Building planning in the top toolbar — Modules and Strings modes hide the outline edit handles, so you need to be in Building mode to edit shapes.

Pro tip: The top-left Layers button switches the background imagery (Orthophoto, Google Maps, Google Solar 3D, Google Earth 3D, Building AI, drone model, custom image). Switch to whatever makes the roof easiest to read while you draw — the imagery is visual reference only and doesn't change the simulation.

Switch an existing project to 3D planning

If a project was started in Quick planning (2D) and you want to move it to the 3D editor, you can't flip a single switch — Quick planning and the 3D editor are different planning modes. Create a new variant on the offer, pick a 3D-capable planning mode at variant setup, and re-plan in 3D. The Quick-planning variant stays intact for comparison.

Draw a roof from scratch

When Building AI doesn't seed a roof (no coverage, roof not detected, or you've deleted the auto-seed), draw the outline by hand.

1. Pick the roof type from the sidebar:
   • Smart Roof — pitched roofs with one or more slopes. Per-edge tilts. Use this for gable, hip, mansard, L-shape, and mixed roofs.
   • Flat Roof — single-plane roofs. One tilt edge + one tilt angle. Use this for genuinely flat roofs and shed / mono-pitch (Pultdach) layouts.
2. Click vertices in order around the roof outline on the background imagery. Each click drops a vertex.
3. Close the polygon by clicking back near the first vertex — the editor snaps the outline closed when the last point lands on the first. There's no double-click or Enter to close.
4. The roof appears on the scene as a Smart Roof with default tilts. Set tilts next.

A few drawing rules:

• Polygons must not self-intersect. If you cross your own line, the editor refuses to close the polygon.
• Vertex order matters for tilt direction, especially on Flat Roofs. The tilt edge is one of the edges you drew — the side that's higher.
• You can move corners after closing by dragging the vertex handles, but you can't split an edge to add a vertex, and Delete removes the whole roof (not one corner). If the corner count is wrong, re-draw the outline.

Draw a Pultdach (shed / mono-pitch roof)

A Pultdach is a single-plane sloped roof — common on garages, modern annexes, sheds.

1. Pick Flat Roof (not Smart Roof — Flat Roof is the single-tilt-plane primitive).
2. Draw the rectangular outline.
3. In the roof detail overlay, set the tilt angle (e.g. 15° for a low-pitch shed roof).
4. Pick the tilt edge — the edge that sits at the high side of the slope.

The Flat Roof primitive handles shed roofs the same way it handles east-west commercial ballast layouts — single plane, one tilt direction.

Draw a roof when satellite imagery is missing or unusable

When the building is too new to be on satellite imagery, the imagery is corrupted, or you're planning against an architect's PDF:

1. Open the Layers panel.
2. Use Own image to upload a custom orthophoto, drone still, architect's site plan, or a screenshot from a different mapping service.
3. Position and scale the uploaded image to match the project coordinates (see Position and size a custom-uploaded background image below).
4. Draw the roof against the custom image the same way you would against satellite imagery.

If you have specific surveyed dimensions (length, width, ridge height) and no usable imagery at all, draw a starter rectangle of roughly the right size on the orthophoto layer, then use the measure tool (ruler icon, bottom toolbar) to compare your polygon's edges to the surveyed dimensions and adjust vertex positions until the measurements match. There's no direct "type in width = 12.4 m" affordance — geometry is built by clicking on the imagery layer.

Position and size a custom-uploaded background image

When you've uploaded a custom image as the planning background, you need to make sure it's anchored at the right scale and position over the real-world coordinates before drawing on it.

1. Upload the image via the Own image tile in the Layers panel.
2. Drag the image to translate it across the canvas — align a feature in the image (e.g. a known building corner) to the real-world position you can see on the orthophoto layer.
3. Use the on-canvas scale handles to resize the image until known distances match the orthophoto scale.
4. To rotate the image, use the rotation handle on the image overlay until the image's north matches the canvas north.
5. Switch back and forth between the custom image and the orthophoto layer to verify the alignment.

Once the image is anchored, draw the roof outline on top of it — the polygon snaps to the canvas at real-world scale.

Reshape and correct an auto-seeded roof

Reonic seeds many roofs automatically from Building AI or from a photogrammetry drone model. The auto-seed is a starting point — review and correct it before placing modules.

1. Switch to Building mode if you aren't already there.
2. Click the roof. Vertex handles appear at each corner.
3. Drag a vertex to move that corner. The two adjacent edges stretch to match.
4. Confirm tilts in the roof detail overlay — auto-seeded tilts are approximations, not surveyed values.

You move corners by dragging the vertex handles. There's no edge-split gesture to add a vertex partway along an edge, and Delete removes the whole roof rather than a single corner. When the auto-roof has the wrong corner count (an L-shape captured as a rectangle, a gable end captured as a single line instead of two edges meeting at the ridge), re-draw the outline rather than trying to add corners to the seed.

Auto-seeded roofs are marked so you can tell auto from hand-drawn, but the mark isn't a lock. You can edit, reshape, or delete them like any other roof.

Pro tip: When in doubt about the auto-seed, switch the Layers background to Google Solar 3D or the drone model if one exists. Reading the roof shape on a 3D mesh is easier than guessing from a 2D orthophoto.

Set the roof tilt

Tilt is what turns a flat polygon into a 3D roof slope. The simulation reads tilts to compute irradiance, shadow casting, and yield — wrong tilts mean wrong KPIs.

Smart Roof — per-edge tilts

Smart Roof exposes a tilt entry for each edge of the polygon. A typical gable roof:

1. Click the roof to open the roof detail overlay.
2. The overlay lists every edge with its current tilt.
3. For the two long edges (eaves), set the roof pitch — common residential pitches in Germany: 30–45°.
4. For the two short edges (gable ends, the triangular sides), leave at 0° — they're vertical walls under the ridge, not sloped surfaces that take modules.
5. For more complex shapes (L-roofs, hip roofs), set each edge to its surveyed or estimated slope.

The 3D scene updates as you type. Confirm the geometry looks right against the Google Solar mesh or drone model if available.

Note: Each tilt you type binds to the edge you entered it on. Confirm the per-edge tilts land where you expect after you set them.

Flat Roof — single tilt + tilt edge

Flat Roof exposes one tilt angle and one tilt edge (which edge sits at the high side of the slope):

1. Click the roof.
2. Set the tilt angle — 0° for a genuinely flat roof, 10–15° for ballasted commercial east-west tilts, 15–30° for a shed roof (Pultdach).
3. Pick the tilt edge — the side that's elevated.

Estimating tilt from Google 3D / Google Solar imagery

When you don't have a surveyed roof tilt:

1. Switch the Layers background to Google Solar (3D) or Google Earth (3D tiles).
2. Click the measure tool (ruler icon, bottom toolbar).
3. Span the roof slope from eaves to ridge.
4. The measurement result shows the tilt angle (the editor derives it from the 3D mesh).
5. Enter that tilt into the roof detail overlay.

This is the standard fallback when you can't get surveyed values — Google's mesh is approximate but close enough for the simulation in most residential cases. For commercial projects with tight financial margins, push for surveyed tilts.

Set roof height alongside tilt

The roof's height (eave height + ridge derived from tilt) is implicit in the polygon + tilt combination — you don't enter a separate "ridge height in meters" value. If you need a specific ridge height (e.g. an unusual mansard), set the tilt that produces that ridge given the polygon span. For roofs that sit on a building taller than the surrounding terrain, the Plan3D scene shows the roof at the appropriate height once Building AI or photogrammetry has supplied the building shell; for hand-drawn standalone roofs the editor uses default building heights.

Note: Ridge height comes from the tilt and polygon together — set the tilt that produces the ridge you need for the polygon span. Rotation around true north sets cardinal alignment.

Align the roof model over the building

When you have an auto-seeded roof (Building AI or photogrammetry) plus a satellite background, the roof should sit precisely over the building outline. If it doesn't:

1. The address is wrong → fix the address on the project. The editor re-centres on the corrected coordinates.
2. The satellite imagery is offset → use the centring control to nudge the imagery into place (see the modules guide for the location-adjust control).
3. The seeded geometry is misaligned → in Building mode, click the roof and drag vertices to reposition. The polygon translates as you drag.
4. The photogrammetry drone model is misaligned → if the drone model lands clearly off-target, draw the roof against the orthophoto and disable the drone-model layer.

Add roof obstacles — Sperrzonen, chimneys, dormers, windows

Modules can't go where chimneys, skylights, satellite dishes, vents, or other obstructions sit. Mark them so auto-layout respects them and the shading simulation accounts for the obstruction.

Add a Sperrzone (exclusion zone) or chimney

1. In Building mode, find the Exclusion Zones (DE: Sperrzonen) panel in the left sidebar.
2. Click Add Exclusion Zone.
3. Click roof vertices around the obstacle's footprint. To close, click back near the first vertex — the editor snaps the polygon closed. There's no double-click or Enter to close.
4. Pick the Type — Chimney (Schornstein), Window, or Other.
5. A height field — labelled Chimney Height, in metres — appears only for the Chimney type (default 0.5 m, range 0.1–10 m). Window and Other zones have no height field; they block module placement as flush footprints.
6. Save.

Height matters for shading, and only the Chimney type carries it:

• Window / Other — block module placement in the footprint but cast no shadow (no height).
• ~0.5–1 m chimney — small chimneys, vents.
• ~1.5–2 m chimney — typical residential chimneys; casts a real shadow on neighbouring modules in the simulation.

For anything that protrudes and shades, use the Chimney type so the height field is available and the shading model handles the shape correctly.

Building AI seeds chimneys where the underlying map data is available. Review the seed — the Type, footprint, and height are estimates.

Add a dormer (Dachgaube)

A dormer (Gauben) is a roof feature with its own slopes — a small extension projecting out of the main roof, often with a window. Dormers can carry modules on their own surfaces.

1. In Building mode, find Add Dormer in the sidebar.
2. Click a point on the roof to place the dormer. A single click commits it with default dimensions — there's no drag-to-size gesture; set the exact width (across the roof) and extension length (how far it sticks out) numerically in the next step.
3. Configure the dormer in the dormer detail overlay:
   • Side tilt — angle of the side faces (often steep, e.g. 60–75°).
   • Front tilt — angle of the front face (often steep or vertical).
   • For a right-angled dormer (vertical front face): set the front tilt to exactly 90°. The front then draws as a flat vertical plane.
   • For flat-topped dormers: enable the flat-roof option on the dormer (Flat Dormer Height Delta), which gives the top surface a flat planning area.
4. Save.

The dormer surface becomes available for module placement during the Modules phase, with its own per-surface tilt. Auto-layout respects dormer footprints — the rest of the roof avoids the dormer outline.

Pro tip: After a Building AI auto-detect run, the Building AI overlay shows the count of dormers it found, alongside the roofs and obstacles it seeded. Use it as a quick sanity check — if the overlay says it placed two dormers but the real roof has three, add the missing one manually before you move on to module placement.

Sperrzone validation messages

If the editor refuses to save your exclusion zone, you'll see one of:

• "Sperrzone kann nicht teilweise mit dem Dachrand überlappen" — the polygon crosses the roof edge. Either keep it fully inside the roof outline, or push it cleanly outside.
• "Sperrzone kann sich nicht selbst schneiden" — the polygon crosses itself. Re-draw cleanly.
• "Ungültige Platzierung der Sperrzone" — placement is otherwise invalid (e.g. degenerate polygon, zero area). Re-draw.

Fix the polygon and re-save.

Add an edge inset around the roof

PV regulations and aesthetic preference often demand a setback from the roof edge — typically 30–50 cm for residential, more for commercial fire-code compliance. The roof carries an inset value (DE: Randabstand) you set in the roof detail overlay, and you can set it per edge rather than one value for the whole roof. That lets you keep a wider eave-side setback for fire-code access and a tighter ridge-side setback on the same roof.

1. Click the roof in Building mode.
2. In the roof detail overlay, find the Inset / Randabstand controls.
3. Pick the edge you want to set, then enter that edge's inset in centimetres. On a Smart Roof the lower edge keeps the base inset and each upper edge takes its own override. On a Flat Roof every edge carries its own value.
4. The 3D scene shows the available module area shrink in from the edge you set.

Each edge's inset clamps to a per-roof maximum, and the auto-fill area (Cover roof side) is carved back independently per edge, so modules respect whichever setback you set on that side. Set the inset on each roof as you draw it.

Trees and natural obstructions

Building AI auto-populates trees where the underlying map data covers them. To model a tree's shading impact, leave it visible in the 3D scene — the simulation casts its shadow on neighbouring modules automatically.

• To exclude a tree that's clearly not there — delete it from the scene.
• To approximate a tree when the map data is wrong — delete the misplaced tree and drop a Sperrzone in its real footprint as a workaround. Give the zone the Chimney type so it carries a height, and set that height to model the tree's shading. Trees are the more faithful obstruction for a big canopy, so where a tree is roughly right, prefer keeping it over substituting a Sperrzone.

Multi-roof and multi-building projects

A project can carry as many roofs and buildings as the customer's site has. Draw each roof or building separately:

• Draw each roof as its own Smart Roof or Flat Roof.
• Add dormers, chimneys, and exclusion zones on each.
• Module placement happens per-roof in the Modules phase.
• The simulation aggregates across all roofs into one set of KPIs for the variant.

For a building with four separate roof surfaces at different tilts (e.g. a mansard or a complex hip roof), draw a single Smart Roof and set per-edge tilts so each surface gets the correct angle. For a residential house plus a flat-roof annex, draw two separate roofs — one Smart Roof, one Flat Roof.

Things to know

• Tilt is the single most load-bearing input you give Building mode. Wrong tilts → wrong irradiance → wrong yield. Push for surveyed values on commercial projects; the Google Solar measure-tool fallback is fine for residential.
• Vertex placement is freeform. Right angles come from where you click and drag, not from a snap-to-90 affordance — drag carefully on rectangular roofs. There's no edge-split gesture to add a vertex, so if the corner count comes out wrong, re-draw the outline.
• Smart Roof vs Flat Roof is a primitive choice, not a roof-type catalogue. Pitched roof → Smart Roof with per-edge tilts (gable, hip, mansard, L-shape all live in this primitive). Single-plane sloped or genuinely flat → Flat Roof.
• Building AI is a starting point, not a final answer. The underlying map data can be wrong, especially for new buildings or hand-mapped regions. Always review the auto-seed before moving to Modules.
• Photogrammetry drone models load as the truest 3D reference. When a drone-imagery job exists for the building, switch the Layers background to Drone model and draw against it — the geometry is more reliable than satellite imagery.
• The drone-model layer needs the photogrammetry capability and a completed job. No job, no drone-model option in Layers.
• Start mode is a one-shot decision. You open the scene from a single 3D with Google Data card in Edit Solar Setup — there's no separate map-source button. The card automatically picks the best available 3D imagery. That start mode is baked into the saved scene — you can switch the active layer in the Layers panel later, but the original choice still shapes how AI obstacle detection runs (see next point). On buildings where the highest-detail imagery isn't available, start the scene against the best imagery you have and review the roof manually for obstacles.
• Detection quality depends on the start mode. Plans started against the highest-detail imagery catch more chimneys and vents automatically. Plans started against lower-detail imagery may surface fewer auto-detected obstacles, so review the roof manually and add Sperrzonen where the detector missed. Obstacle detection behaves consistently across every workspace.
• The one externally-supplied 3D model is the photogrammetry drone model, which Reonic auto-loads from a drone-imagery job. If you don't have photogrammetry, model the roof manually from satellite or a custom image. CAD formats (USDZ, IFC, DXF, SketchUp, Rhino, Revit, ArchiCAD) aren't imported into Plan3D today; contact your Reonic account manager if you'd like to see CAD import.
• Plan3D models roof-mounted PV. Place modules on roof surfaces and dormer faces. Vertical façade / wall-mounted PV isn't drawn in the editor today; contact your Reonic account manager if you need it.
• Carports and ground-mounted free-field arrays — see the modules guide. The short version: model carports as separate Flat Roofs at the right height; free-field PV is roof-anchored today and may not fit cleanly.
• Right-angled dormers. A dormer with a front tilt of exactly 90° (a vertical front face) draws as a flat vertical plane. Vertical fronts are safe to use.
• Reshape a drawn outline by moving its vertices. To adjust a roof after you've drawn it, drag individual vertices to reposition them. There's no single "rotate the whole polygon by N degrees" control — vertex-by-vertex repositioning is the workflow.
• The "blue square only" symptom. If a roof renders as a single blue square in 3D and the polygon controls don't show, the seed geometry didn't load properly. Refresh the editor; if it persists, contact your Reonic account manager with the offer URL.
• Validation strings are German-language even in the English UI. Sperrzone, Schornstein, Gaube — Reonic's planning vocabulary uses German across markets for the obstacle types.
• Undo is capped at 50 steps. Heavy outline editing can blow through history fast — save partway as a checkpoint. Saved states are preserved in the version history.

Need help?

• Step-by-step questions about Plan3D → contact your Reonic account manager.
• Feature requests (e.g. CAD import, façade PV, direct ridge-height input) → contact your Reonic account manager.
• Bug reports (blue square roof, vertex handles missing, polygon won't close, validation messages on a clearly-valid Sperrzone) → include the offer URL and what you expected versus what you see.

Last updated on August 4, 2026