BIM Modelling for Reliable Existing Conditions

A refurbishment model is only as reliable as the evidence behind it. When a design team starts with outdated floor plans, partial record drawings or assumptions made from a site visit, coordination problems are often built into the programme before design has properly begun. BIM modelling provides a structured way to turn measured existing conditions into usable digital geometry - but its value depends on how the building is captured, interpreted and checked.

For architects and consultants working on existing buildings, the question is not simply whether a Revit model is required. It is what the model needs to represent, how accurate it must be and whether the output will support the decisions that follow.

What BIM modelling means for existing buildings

BIM modelling is the process of creating an intelligent digital representation of a building. In an existing-condition workflow, it commonly begins with a survey-grade point cloud created through 3D laser scanning, supported where necessary by measured survey observations and photography. The model is then produced in Revit or another agreed BIM environment, with walls, floors, roofs, openings, stairs and key structural or architectural elements positioned to reflect the surveyed fabric.

Unlike a collection of CAD drawings, a BIM model connects geometry with object-based information. A wall is not merely a pair of lines: it is a defined element with dimensions, levels and a relationship to the surrounding building. This gives design teams a coordinated base for plans, sections, elevations, schedules, design options and consultant coordination.

That does not mean every model must contain every visible detail. A dependable model is one that is fit for its intended use. Modelling a warehouse for early feasibility requires a different level of detail from documenting a listed townhouse where retained cornices, irregular roof lines and non-orthogonal walls influence the design.

The difference between a point cloud and a Revit model

A point cloud is a dense record of measured points captured across the site. It is highly valuable evidence, particularly where geometry is irregular or difficult to access. It can show floor level changes, distorted walls, exposed structure and roof form with a degree of spatial context that photographs and manual dimensions cannot provide alone.

A point cloud is not, however, a design-ready model. It requires specialist software to view effectively, can be heavy to manage and does not automatically provide the object structure needed for a typical Revit workflow. BIM modelling translates the relevant surveyed information into clear, organised building elements that the project team can work with.

There is a necessary judgement involved. Existing buildings are rarely perfectly straight, level or square. If every minor deviation is modelled literally, the model may become cumbersome and less useful for design. If too much is simplified, it may conceal constraints that matter. The right approach is to agree tolerances, model purpose and areas of particular sensitivity before production begins.

Choosing the right level of development

Level of Development, usually referred to as LOD, helps define what the model will contain. It should be treated as a project instruction rather than a marketing label. An LOD300 model, for example, can be highly effective for design coordination when its geometry is based on accurate capture and its scope is clear.

For existing-condition BIM models, the requirement often sits somewhere between LOD100 and LOD400:

• LOD100 supports massing, broad feasibility and early area studies. Geometry is indicative rather than detailed.

• LOD200 provides general systems, approximate quantities and clearer spatial relationships for concept development.

• LOD300 represents elements with reliable size, location and orientation for coordinated design work.

• LOD400 is used where fabrication-level or installation-specific information is needed, and is only appropriate where the intended use justifies the additional effort.

The most useful instruction is rarely just an LOD number. It should state which elements are to be modelled, the anticipated accuracy, what is excluded and how the model will be used. For example, a team may need an LOD300 shell and core model, detailed roof geometry, floor-level data and accurate window openings, while internal joinery and loose furniture are outside scope.

Why capture quality determines model quality

The phrase “modelled from survey” can hide significant differences in process. A model produced from limited dimensions and reference photographs may be sufficient for a simple, regular space. It is a higher-risk basis for a heritage property, a multi-level conversion or a building with complex junctions and accumulated alterations.

Precision-first BIM modelling starts with an appropriate capture strategy. Laser scanning establishes dependable spatial data across the building, while targeted survey control, photographic records and site notes help resolve material changes, concealed conditions and elements that need interpretation. The modeller should work from that evidence rather than redraw assumptions from old plans.

This is particularly important in buildings where design risk sits at the edges: a roof that falls in more than one direction, a bowed external wall, unequal window bays, a historic stair, or a structural opening that does not align with the levels shown in archive drawings. These conditions do not make BIM unsuitable. They make a measured and carefully scoped BIM process more valuable.

What a design-ready BIM deliverable should include

A useful handover is more than an RVT file. It needs to allow a project team to understand the model quickly and use it without spending days repairing its structure.

At a minimum, the model should be correctly located and levelled, with logical naming, sensible worksets where relevant, and clearly defined existing elements. Plans, sections and elevations should derive coherently from the same model rather than being treated as disconnected outputs. Where a point cloud is supplied or retained for reference, its relationship to the model and coordinate system should be clear.

Good documentation also makes exclusions visible. If services, inaccessible roof areas, concealed structure or detailed finishes have not been modelled, that should be stated plainly. Honest scope definition is not a limitation of the service. It protects the design team from relying on information that has not been verified.

For complex projects, it can also be helpful to identify key deviations or uncertainty zones. An area behind fixed joinery, for instance, may be represented from accessible evidence but require opening-up before detailed design. Recording that distinction gives the team a better basis for managing risk.

Common BIM modelling problems and how to avoid them

The most expensive model is often the one that appears inexpensive at the start but cannot be trusted later. Typical warning signs include walls that do not align in section, generic roof geometry, missing level changes, openings placed from old drawings rather than measurement, and files with inconsistent naming or excessive imported geometry.

These problems are usually caused by one of three issues: insufficient site data, an unclear modelling brief or weak quality assurance. Each can be addressed early.

Before commissioning, establish the design stage, required software version, survey coverage, expected outputs and critical building areas. Confirm whether the model is for planning, feasibility, design development, coordination, heritage consent support or another defined purpose. A provider should be able to explain how they will capture the building, what their model will include and how checks will be carried out before issue.

Quality assurance should compare the completed model against the source data in plan, section and elevation, not simply check whether the file opens. In a complex building, sectional checks are especially revealing. They expose incorrect floor build-ups, roof forms, level relationships and junctions that can remain hidden in plan view.

Where specialist modelling adds the most value

Standard spaces can often be documented efficiently with a straightforward scan-to-model workflow. Specialist input becomes more valuable where the building resists standard assumptions: listed properties, converted industrial spaces, historic interiors, irregular roofscapes, curved façades and buildings altered in several phases.

In these cases, the aim is not to create an unnecessarily elaborate digital replica. It is to capture the geometry that affects design, approvals, coordination and cost. A well-resolved model can reduce repeat visits, limit avoidable redesign and help consultants work from the same dependable record of the existing building.

Space Captures approaches this work as a documentation service first. The Revit model is supported by measured capture, clear scope and structured outputs, so the information can move into design with fewer unanswered questions.

A good existing-condition model should leave the project team with a clear starting point: geometry they can rely on, limitations they can see, and enough confidence to make the next design decision without guessing.