Daylight modelling inputs that get challenged

Daylight model inputs challenged in review are rarely the test or the target. Both sides usually agree that the vertical sky component matters, or that the climate-based target illuminance applies. The argument is about reflectance values, what geometry was included, which rooms were assessed and how the windows were mapped to them, because those choices move results by more than the difference between a pass and a fail.

A reviewer who cannot reconstruct the model from the submitted document cannot verify it. So the practical standard is not whether the modelling is right. It is whether someone else could repeat it and get the same number.

Geometry and the level of detail modelled

Start the audit with what is in the model and what is not.

Window reveals, overhangs, balconies, projecting fins, external walkways and deep soffits all reduce the sky visible from a reference point, and omitting them inflates results. Balconies on the proposed building are the most frequent omission, and on a single-aspect flat they can be decisive, as our note on daylight to single aspect flats sets out.

Then look outwards. The extent of surrounding context included should be stated with a radius or a boundary, and should extend far enough that further buildings would not subtend meaningful angles at the assessed windows. Terrain matters on a sloping site, because a flat ground plane in a model on a 1 in 12 street produces systematically wrong results.

Ask three questions of any geometry schedule. Was the surveyed or the drawn building line used for existing neighbours. Were roof forms modelled or simplified to a flat plane at ridge height. And were trees included, excluded, or represented as a permeable obstruction with a stated transmission value.

Surface reflectance values and their justification

Reflectance is where the largest unjustified gains appear, particularly in internally reflected component and climate-based calculations.

Figures of this kind are conventions drawn from the guidance and from CIBSE practice, and they should be reported as assumptions rather than facts. Where a scheme relies on a value outside the usual range, the assessment should name the finish, its measured or published reflectance, and the condition that will secure it. A model using 0.8 internal wall reflectance to pass a target, with no specification behind it, is not an assessment of the building that will be built.

External reflectance deserves equal attention on courtyard and light well schemes, where a light-coloured facade opposite does change the result materially. The point is to evidence it, not to assume it.

Glazing transmittance, frame factor and maintenance

Three multipliers sit between the aperture and the light, and all three get quietly optimised.

Light transmittance depends on the glazing build-up. Clear double glazing sits in the region of 0.7, and triple glazing and solar control coatings reduce it further. The transmittance used should come from the specified unit, not from a generic default, and where the unit is not yet specified the assessment should say which value was assumed and what happens if a lower one is chosen.

Frame factor is the proportion of the aperture that is glass. Slim aluminium systems and chunky composite systems differ enough to matter, and a frame factor taken from a schematic elevation rather than a window schedule is a weak input.

A maintenance or dirt factor accounts for soiling of the glass. It is routinely omitted, and omission is conservative in the wrong direction.

Report all three in a table, with the source for each. Where the glazing specification is a condition rather than a commitment, say so.

Existing obstructions and the baseline

The baseline choice frames everything that follows, and it should be stated explicitly.

The existing condition is the normal starting point for neighbouring amenity tests. But where a consented but unbuilt scheme sits between the application site and the receptor, an assessment may present a consented baseline as well, and a reviewer should see both. Running only the baseline that produces the better answer, without saying so, is the single most common criticism in review.

Temporary structures, scaffolding, hoardings and trees of uncertain longevity should be identified and their treatment explained. Construction stage effects are a separate question with their own approach, discussed in temporary loss of daylight during construction.

Room use assumptions and window mapping

This is the input a reviewer can most easily test and most often finds wrong.

For neighbouring properties the internal layout is usually unknown, so assumptions have to be made and stated. The defensible method is to derive layouts from the external fenestration, the property type and whatever evidence exists, including sales particulars, historic applications and the authority's records, and then to say what was assumed for each window.

• Which rooms are treated as main habitable spaces, and which as kitchens, bathrooms, halls or circulation.
• How combined living, kitchen and dining rooms are treated, and whether a separate assessment point was used for the kitchen area.
• Which window serves which room, where a room has more than one, and whether rooms were assumed to be single aspect.
• Room depth and the reference plane height used for the no-sky line.

Where an assumption is uncertain, the honest response is a sensitivity test rather than a confident single figure. A reviewer will accept a range with the reasoning far more readily than a precise number with a hidden assumption.

Sky models and climate files

The sky model has to match the metric. Vertical sky component and average daylight factor rest on the standard overcast sky, which is a defined luminance distribution rather than a weather condition. Reporting them against a climate file is a category error.

Climate-based modelling needs a weather file, and the choice is itself an input. State which file, which location it stands for, and which reference year, because a file for a city 60 kilometres away with a different cloud regime is not interchangeable. For overheating-adjacent work a design summer year is used instead of a test reference year, and conflating the two in one document causes confusion.

Sunlight calculations depend on the sun path rather than on a sky model, so state the latitude and the dates used, and confirm whether the assessment covers the whole year or only the winter period alongside it.

Documenting inputs so a reviewer can audit them

An assessment that can be audited includes an input schedule. Put it in an appendix and keep it to the facts.

1. Software and version, and the calculation engine where these differ.
2. Geometry sources, with drawing numbers and revisions, survey references and the context extent.
3. A reflectance schedule by surface, with the source or specification for each value.
4. Glazing transmittance, frame factor and maintenance factor, with the window schedule reference.
5. Baseline cases run, and which is reported as the primary case.
6. Room use and window mapping table for every assessed property.
7. Sky model or climate file, dates and reference plane heights.
8. Sensitivity tests carried out and their results.

Then supply model views: a plan and two axonometric screenshots showing the modelled context. Those three images answer half the questions a reviewer would otherwise raise in writing.

Local policy determines which tests carry weight in the first place, which is covered in local plan amenity policies and how they use BRE 209. Where sun on ground rather than skylight is the issue, the analysis belongs in an Overshadowing Study, and a Daylight and Sunlight Assessment carries the window-based tests. PlanScope drafts both to a structure that exposes the assumptions rather than burying them, and flags each site-specific input it cannot supply.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. For specialist technical reports, qualified professional input may be required.