A worked overshadowing test for a garden
A step by step sunlight on ground test for a rear garden, using the BRE 209 equinox criterion, with an illustrative calculation and the changes that recover a failing result.
An overshadowing garden test asks a narrow question. On 21 March, does at least half of the amenity area still receive two hours of direct sunlight, and if the proposal reduces the sunlit area, does it fall below 0.8 times what it was before. Those two numbers come from the Building Research Establishment guide Site Layout Planning for Daylight and Sunlight, usually called BRE 209, and they are what a council officer will reach for when a neighbour objects that a new extension will put the garden in shade. Everything else in the argument is detail hanging off them.
What the sunlight on ground test measures
The test looks at sunlight reaching the ground, not light reaching windows. Window based measures such as vertical sky component and annual probable sunlight hours answer a different question and use different thresholds.
Amenity area means the usable outdoor space attached to a dwelling. For a terraced house that is normally the rear garden, excluding a side passage, a bin store or a strip of planting nobody sits in. For a block of flats it is the communal garden. Defining the area honestly is the step most homemade studies get wrong, because shrinking the assessed area to the sunniest corner produces a flattering result that an officer will spot.
The criterion has two limbs. At least 50 per cent of the area should receive at least two hours of sunlight on 21 March. Separately, if the sunlit portion after development is less than 0.8 times the area before, the loss is likely to be noticeable. A garden can pass the first limb and still fail the second, and vice versa.
Choosing the date and the hours
21 March is the assessment date because the equinox sits midway between the extremes and gives a fair year round picture. Midsummer flatters every scheme. Midwinter condemns almost all of them.
The two hours do not need to be continuous. Sunlight received from 10am to 11am and again from 2pm to 3pm counts. That matters on sites where a chimney, a mature tree or an existing outrigger casts a moving band of shade across the middle of the day.
Sun position is taken from the site latitude, so a garden in Newcastle and a garden in Exeter behave differently on the same date. Any model that does not ask for a location is not doing the calculation properly.
An illustrative worked example
The following figures are illustrative and are meant to show the method rather than to describe any real property.
Take a rear garden 8 m wide and 10 m deep behind a south facing 1930s semi. Before development, shadows come from the existing rear wall and a 1.8 m boundary fence, and 62 m² of the 80 m² garden receives two hours or more of sun on 21 March. That is 78 per cent, comfortably over half.
The owner proposes a two storey rear extension 3 m deep across the full width. Re-running the model with the taller massing, the area receiving two hours falls to 44 m². The garden still passes the first limb, because 44 m² is 55 per cent of 80 m². But 44 divided by 62 is 0.71, below the 0.8 ratio, so the reduction is one the guide treats as noticeable.
Now change one variable. A single storey extension of the same depth leaves 57 m² sunlit. The ratio becomes 0.92 and both limbs pass. The difference between the two schemes is height, not depth, which is why the second storey is where these arguments are usually won or lost.
Reading the result against the criterion
BRE 209 is guidance, not policy, and it says so. The numbers are advisory and are intended to be applied flexibly, especially in dense urban areas where meeting them everywhere would prevent any development at all.
That flexibility cuts both ways. An officer can accept a result below 0.8 where the garden is unusually large, where the loss affects a part of the plot nobody uses, or where local policy expects town centre densities. An officer can also refuse a scheme that passes numerically if the shadow falls across the only sitting out area and the local plan carries a strong amenity policy. The calculation informs the judgement, and the judgement belongs to the council.
Check the adopted local plan and any householder design guide before assuming the BRE figures are the whole test. Some authorities set their own garden sunlight expectations, and some apply a 45 degree rule alongside them. Where local and national approaches differ, the adopted plan carries the weight under section 38(6) of the Planning and Compulsory Purchase Act 2004.
Fences, trees and what the model should include
Model the world as it is. That means the existing boundary fence at its real height, the neighbour's garage, and any building that already shades the plot.
Trees are the awkward case. A deciduous tree on 21 March is mostly bare, so most assessors model the trunk and main limbs rather than a solid summer canopy, and say which they did. An evergreen conifer hedge is modelled as solid. Either way, state the assumption, because a study that quietly deletes a 6 m leylandii to improve the before case is worthless.
Do not model a fence you intend to raise, or a tree you intend to fell, without showing both cases. Officers ask for the existing situation and the proposed situation, drawn the same way.
Design changes that recover the position
Height reduction helps more than depth reduction, as the worked figures show. Beyond that, the changes that tend to move the number are a hipped or pitched roof in place of a flat parapet, setting the extension in from the shaded boundary by a metre, dropping the eaves line, and pulling the first floor element back so the two storey part sits away from the boundary.
Relocating the usable amenity area is sometimes the honest answer. If the scheme creates a new patio on the sunnier side of the plot, say so and assess that area too. A Design and Access Statement is the natural place to explain why the massing settled where it did, and officers respond well to a design evolution narrative that shows an option tested and rejected on sunlight grounds.
Presenting the diagrams so an officer can check them
A set of shadow diagrams should be self explanatory. Include a plan for the existing case and the proposed case at hourly intervals across 21 March, a north point, a scale bar, the assumed heights of every modelled object, and the latitude used.
Add a short table of the numbers rather than leaving the reader to measure hatched areas with a ruler. Give the amenity area in m², the sunlit area before and after, the percentage and the ratio. Where the study is being produced for a larger scheme or a contested householder application, an Overshadowing Study or a fuller Daylight and Sunlight Assessment will also cover neighbouring windows, which the ground test ignores entirely.
This is general guidance, not legal advice. Check the position with your local planning authority or a qualified professional before relying on it, because amenity judgements turn on the adopted local plan and the facts of the site.
Enter your postcode on PlanScope to see which supporting documents your extension is likely to need before you commission any modelling.
PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.
Frequently asked questions
How much sunlight should a garden get for planning purposes?
The Building Research Establishment guide suggests at least half the amenity area should receive two or more hours of direct sunlight on 21 March. It adds a second test, that the sunlit area after development should not fall below 0.8 times the area before. Both are advisory figures rather than policy, and councils apply them alongside their own local plan amenity policies.
Why is 21 March used for overshadowing studies?
The spring equinox sits midway between the summer and winter extremes, so it gives a balanced picture of the year. Assessing on 21 June would flatter almost any scheme because the sun is high, and assessing on 21 December would condemn most of them. Some studies add June and December diagrams for context, but the criterion is tested in March.
Do the two hours of sunlight have to be continuous?
No. Sunlight received in separate periods counts towards the two hours, so an hour in the morning and an hour in the afternoon meets the test. This matters where a chimney, an existing outrigger or a mature tree casts a band of shade across the middle of the day but the garden is open at either end.
Does a boundary fence count in the shadow model?
Yes. The existing situation should be modelled as it really is, including boundary fences at their actual height, neighbouring garages and any building that already shades the plot. Deciduous trees are usually modelled as trunk and main limbs on 21 March because they are largely bare, and the report should state which assumption was used.
Can an extension be refused only because of garden overshadowing?
It can, where the local plan carries an amenity policy and the loss of sunlight to a neighbour's usable garden is judged unacceptable. Numbers inform that judgement rather than settle it, so a scheme that fails the ratio may still be approved in a dense urban setting, and one that passes may still be refused on wider amenity grounds.
Sources and further reading
This article is general guidance about the planning system in England and, where stated, the rest of the UK. It is not legal or professional advice. Check your local planning authority’s validation requirements and take professional advice for your specific site.