Overheating in London schemes and the cooling hierarchy

London Plan Policy SI 4 does not ask whether a scheme overheats. It asks what you did about it, in a fixed order, and it treats active cooling as the last of six steps rather than as a design choice. Major development has to demonstrate through its energy strategy how overheating and reliance on air conditioning have been reduced, and the demonstration has to follow the cooling hierarchy step by step.

That order is the whole policy. A scheme that arrives with comfort cooling in every flat and a modelling report showing compliance has satisfied the numbers and failed the policy, and on a referable application the Greater London Authority will say so in its stage 1 report.

The cooling hierarchy and the order it imposes

The hierarchy runs as follows. Reduce the amount of heat entering the building in summer, through orientation, shading, albedo, the extent and specification of glazing, insulation and green roofs and walls. Minimise internal heat generation through energy efficient design. Manage the heat within the building, using exposed thermal mass and higher ceiling heights. Provide passive ventilation. Provide mechanical ventilation. Only then provide active cooling.

Each step has to be addressed in the submission, including the ones you did not use, with a reason. The common failure is to jump from step one to step six because the site is noisy, skipping thermal mass, ceiling height and ventilation strategy entirely. Those middle steps are usually where the least costly improvements sit.

• Step one is an architectural decision, taken on the first massing sketch
• Step two overlaps with the energy strategy and the choice of heating system
• Step three is a structural and ceiling height decision, lost once the section is fixed
• Step four depends on the plan form and the proportion of dual aspect units
• Step five is a services decision with plant space consequences
• Step six is accepted only where the preceding five have been exhausted

Reducing gains before adding cooling

The measures at step one that do real work in London are shading geometry, glazing proportion by orientation and the treatment of the roof.

External shading beats internal shading by a wide margin, because it stops the energy before it enters. Horizontal projections work on south elevations, vertical fins on west and east ones, and recessed reveals work everywhere at modest cost. Balconies provide shading as a by-product, which is one of the few cases where a planning requirement and a thermal requirement pull in the same direction.

Glazing proportion changes the result more than anything else, and it meets the most resistance from the design team. Reducing it affects the elevation and the daylight case, and the way BRE 209 daylight guidance applies in dense London means the daylight assessment is already working hard on a tight site. Resolve the two together rather than sequentially.

Roof treatment matters more than its area suggests. A dark membrane roof over the top floor flats raises their cooling load noticeably, and a green or reflective roof reduces it while contributing to the urban greening requirement. The top floor units are almost always the worst performers in the modelling, so this is targeted spending.

The modelling the Mayor's guidance expects

For residential schemes the assessment uses dynamic thermal modelling to the CIBSE residential overheating methodology, with the occupancy profiles and internal gains that methodology specifies. The Mayor's energy assessment guidance sets out what to submit, and it asks for more than a pass or fail statement.

Expect to provide results for a sample of units selected to represent the worst cases by orientation, floor level, aspect and exposure, with the selection justified. Expect to model communal corridors rather than only dwellings. And expect to be asked for sensitivity testing against hotter and future weather years alongside the baseline design summer year, because a scheme that passes on current weather and fails badly on a 2050s file invites a question about resilience.

State the assumptions that drive the results, because these are what reviewers interrogate: the window opening profile and the free area assumed, whether blinds or curtains are modelled and on what schedule, the ventilation rates with windows closed, internal gains from any communal heating pipework, and the surface absorptance of the external envelope. An Overheating Risk Assessment that reports results without its assumptions is unreviewable.

Window opening, noise, air quality and security

Passive ventilation at step four depends on windows being openable, and in London three constraints commonly remove that assumption.

The evidence has to be specific to the elevation rather than to the site. A perimeter block typically has one elevation where windows cannot be relied on and three where they can, and modelling the whole scheme on the worst elevation's assumptions produces an unnecessarily mechanical building.

Where the noise and air quality assessments were prepared for other purposes, use their data rather than commissioning assumptions. The facade noise levels in the acoustic report and the pollutant concentrations in the air quality assessment are exactly the inputs the overheating model needs, and inconsistency between those documents is the easiest thing for a reviewer to spot.

Corridors, risers and heat networks

Communal corridors are the London-specific problem that the modelling methodology addresses and that design teams still miss. A corridor containing poorly insulated heat network pipework can run well above the dwellings it serves, and every door opening transfers that heat inwards. On a scheme with internal corridors and no openable windows in them, the corridor becomes a heat store.

Three design responses carry weight. Insulate the distribution pipework to a standard set by calculation rather than by convention, and state the standard. Locate heat interface units with their losses in mind, and avoid clustering them in unventilated cupboards within dwellings. And provide corridor ventilation, mechanical where necessary, sized for heat removal rather than only for smoke control.

Where the scheme connects to a district heat network, the network's flow temperature drives the pipework losses, and a lower temperature network is better for overheating as well as for carbon. That is a point worth making in the Energy Statement where the connection is being negotiated, because it affects the Be Clean case and the heat risk case at the same time.

Where active cooling is accepted

The policy does not prohibit active cooling. It requires the preceding steps to have been taken, and it expects the need for cooling to be the residual after they have. In practice cooling is accepted on elevations with unavoidable noise or air quality constraints, in single aspect units where the plan cannot be changed, and in non-residential uses with high internal gains.

What is not accepted is cooling as a substitute for the first three steps. A scheme with an unshaded highly glazed south elevation and cooling throughout will be asked to revisit the facade, and the GLA will usually ask for the modelling to be rerun with passive measures before the cooling case is considered.

Cooling also has consequences the energy strategy has to own. Plant takes roof space that is already contested, condensers make noise that affects the amenity case, and the electrical load affects the capacity discussion with the distribution network operator. On a scheme where the roof was fully allocated at planning stage, adding cooling later is a planning problem rather than a services one.

Testing the layout before the design freezes

The least costly overheating mitigation is a plan that needs very little of it. Four decisions, all taken early, determine most of the outcome.

1. The proportion of dual aspect dwellings, which London housing design policy expects to be maximised and which the heat risk policy rewards
2. The orientation of single aspect units, with the unavoidable ones facing north or east rather than west
3. Floor to ceiling height, which gives a thermal buffer and cannot be recovered later
4. Balcony depth and position, which provide shading where they sit above the glazing

Those decisions interact with density. Pushing site capacity usually increases the proportion of single aspect units, and the relationship between optimising site capacity and heat risk is a real trade-off rather than a presentational one. Run an indicative overheating check on the worst unit type at the point the plan form is chosen, not after the planning drawings are issued.

One honest uncertainty. Weather files and the criteria applied to them continue to be revised, and a scheme assessed today against the current design summer year may be reviewed against a different file by the time it is built. Designing with a margin above the pass threshold is the only practical hedge.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.