Renovation costs per m² by depth of intervention

Four stages can usefully be distinguished, though the boundaries are fluid. Which stage a project falls into decides the order of magnitude of the benchmark.

These stages are not defined terms. They describe how deeply the building is opened up, and that is precisely the quantity that governs cost.

Comparing the figures published in 2026 for a full refurbishment shows the extent of the dispersion. It is wider than in new build.

All the figures date from the same year and describe the same operation. The lower bound of one source sits at 600 euros, the lower bound of another at 1,500. Notably, one and the same provider carries two different ranges on two of its pages.

The reason for the contradiction is linguistic rather than arithmetic. There is no binding definition of Kernsanierung. In one source the term describes replacing roof, windows, heating, insulation and electrics while the interior fit-out remains; in another it describes strip-out to the bare shell.

Between those two readings sits the entire interior fit-out, the cost block with the widest range in the whole project. That the published figures differ by a factor of four is therefore not imprecision but the logical consequence of an imprecise term.

Rule for cost estimating: in existing buildings you never work from a term but always from a schedule of trades. The question is not what a full refurbishment costs, but which trades are touched and to what extent.

In new build, a benchmark describes a building that does not yet exist but can be fully defined. In an existing building it describes an intervention into a structure whose condition is only partly known. Four effects follow.

The starting condition varies more than the target. Two buildings of the same period and size can differ considerably in fabric, existing defects and previous interventions. The target condition, by contrast, is relatively uniformly describable.

Part of the scope only becomes visible once opened up. Service routes, structural condition, moisture damage and contamination emerge during the works, not before. In older stock that is the norm rather than the exception.

The effort per unit of work is higher. Working in existing buildings means constrained conditions, hand work instead of plant, protection measures for retained elements, and frequently working in occupied premises.

The sequence of trades is more fragile. Where services must be exposed, walls chased and then made good again, costs rise in steps rather than linearly.

In practice the spread of tenders is therefore considerably wider than in new build. Differences of 20 to 40 per cent between tenders for the same renovation scheme are reported. In existing buildings such a difference is not negotiating room but generally a signal of divergent assumptions about the starting condition.

As throughout this cocoon, published renovation figures relate almost universally to habitable area. In existing buildings, though, a second and more serious error is added: the reference area is not the same as the area being worked on.

A renovation never affects the whole building evenly. Roof works affect the roof area, facade insulation the envelope area, a heating replacement neither. Spreading the cost over habitable area anyway produces a figure reflecting neither the quantity nor the price of the work.

This shows most clearly in building geometry. A compact apartment building and a detached house of the same habitable area differ considerably in the envelope area to be insulated. A benchmark per square metre of habitable area for facade insulation is therefore not transferable between them, despite carrying the same unit.

Practical rule: for element-based measures, calculate per element area, that is per square metre of roof, facade or window, and convert back to habitable area only at the end where a comparison figure is needed. The system of reference quantities is set out in the article on reference areas.

A sequence follows that applies more strictly in existing buildings than in new build. Before any cost statement comes a condition survey of the specific building: structure, moisture, contamination, condition of services, element build-ups.

The effort appears in cost group 700 and at first looks like an additional item. It reduces uncertainty in cost groups 300 and 400 by an order that routinely exceeds the cost of the investigation. The system of cost groups is set out in the article on ancillary construction costs under KG 700.

Full refurbishment raises the questions of intervention depth, the influence of construction period, and comparison against a replacement new build. The article on full refurbishment costs covers these points.

Energy retrofit follows its own logic, because it is planned, funded and evidenced element by element. The article on energy retrofit costs covers costs by element.

Conversion and change of use differ from renovation in that the use itself changes, which triggers consent questions and step changes in requirements. The article on conversion and change of use costs covers those risks.

• Full refurbishment: costs by depth of intervention
• Energy retrofit: costs by element
• Conversion and change of use: cost risks in existing buildings
• Reading and updating German construction cost benchmarks
• Construction costs per m² in Germany