Closing the gap between designed and as-built performance

A performance gap planning condition asks you to show that the building you built performs as the building you consented was modelled to perform. The gap between the two is rarely a mystery. It comes from airtightness, from construction quality at junctions, from commissioning that was rushed, from products substituted during the build, and from modelling assumptions that were optimistic to begin with.

What makes verification work is specificity. A condition requiring the development to achieve the carbon reduction in the approved energy statement is almost unenforceable. A condition requiring an as-built calculation, an air permeability test result and a commissioning schedule, submitted before occupation, is straightforward.

Where the gap comes from on site

Five causes account for most of it, and four are documented somewhere on site.

Airtightness is the largest single variable a design can control and the easiest to lose. A detail that works on a section drawing fails at the service penetration, the joist end and the window reveal, and the test result at the end reflects the worst of them.

Thermal bridging comes second. Modelled psi values assume a detail built as drawn. Compressed insulation at a wall plate, a missing cavity closer or a cold bridge at a balcony support all move the fabric performance without appearing in any drawing revision.

Commissioning is third and the most avoidable. Ventilation systems not balanced, heat pumps left at factory flow temperatures, controls with default schedules, and time clocks never set. This is a fortnight of work at the end of a two-year build, and it is the fortnight most often compressed.

Substitutions are fourth. A different heat pump model, a smaller array, a change of glazing specification, a switch of insulation product. Each is defensible individually and their combined effect is rarely re-modelled.

Fifth is modelling itself, including occupier loads the calculation never counted. That one cannot be fixed by verification, only by honest reporting.

Conditions that require as-built verification

Conditions on energy performance have become common, and they come in three broad shapes.

The first requires as-built compliance outputs: the completed calculation for the finished building, demonstrating the percentage reduction the approved strategy promised, submitted before first occupation. This is the most workable because building control already generates the evidence.

The second requires evidence of specific measures: the array capacity actually installed, the heat pump model and its seasonal efficiency, the air permeability result, the glazing specification. This works where the approved strategy stated those parameters as numbers rather than as intentions.

The third requires post-occupancy reporting of metered energy over a period. In London, the reporting stage of the energy policy asks for predicted and in-use performance for major development, and elsewhere similar conditions are appearing. This shape is the hardest to discharge, for reasons covered further down.

Read the draft conditions before determination rather than after. A condition drafted for a different building type is easier to amend at that point than at discharge stage. Our note on zero carbon conditions and how they are discharged sets out the discharge mechanics.

Airtightness testing and commissioning evidence

The 2021 edition of Part L tightened the evidence required for dwellings at completion, including photographic evidence of construction stages and a wider requirement for air permeability testing, with the detail in Approved Document L Volume 1. That change helps a planning discharge as much as it helps building control, because the evidence already exists.

Use it. The photographic record of insulation continuity, membrane laps and service penetrations is exactly what a planning officer asking about fabric performance wants to see, and it costs nothing extra to submit.

For air permeability, three practical points. Test early on the first unit of each type so that failures can be diagnosed while the details can still be changed. Record the result against the design value used in the calculation as well as against the regulatory limit. And where a result is worse than the design assumption, re-run the calculation rather than hoping the margin absorbs it.

Commissioning evidence means the ventilation airflow measurements, the heating system commissioning record with flow temperatures and control settings, the metering installation record, and the renewable system commissioning certificate. Ask for those as a named deliverable in the contract. Asking for them after practical completion is how they go missing.

Substitutions and change control

This is the part architects can control and often do not.

Set up a change log for anything that appears in the approved energy strategy. Every proposed substitution is checked against the parameter it affects: U-value, efficiency, capacity, area, airtightness assumption. Where the change is neutral or better, record it. Where it is worse, the calculation is re-run and the shortfall is either compensated elsewhere or flagged.

The planning consequence depends on how the strategy was secured. Where the energy statement is an approved document under a condition, a departure from it needs either agreement from the authority or a formal route: a non-material amendment under section 96A of the Town and Country Planning Act 1990 for minor changes, or an application under section 73 to vary the condition where the change is more than minor.

The common failure pattern is a sequence of small value-engineering decisions, each agreed verbally, which together take the scheme below the consented reduction. Nobody notices until the as-built calculation is run, by which time the array is not going back on the roof.

Keep the Energy Statement parameters in a single table on one page, and make that page the reference for every substitution decision. It is a five-minute job that saves a section 73 application.

What a monitoring condition can reasonably ask

Conditions have to meet six tests: necessary, relevant to planning, relevant to the development, enforceable, precise, and reasonable in all other respects. In-use monitoring conditions strain two of them.

Enforceability is the first problem. A condition requiring metered energy data for five years after occupation depends on occupiers who were not party to the permission, and on meters they control. For a single-let building with a facilities team this is workable. For 80 flats sold to individual owners it is not, and an authority insisting on it is drafting something it cannot enforce.

Precision is the second. A condition requiring the building to "perform in accordance with" the approved strategy gives no test. What performs, measured how, against what baseline, over what period, and what happens if the answer is no.

Where monitoring is really wanted, a planning obligation is usually the better instrument, because it can bind a management company, specify a reporting format and a recipient, and provide for what happens on a shortfall. Argue for that at the drafting stage rather than accepting an unenforceable condition and dealing with it later.

What a condition can sensibly require is the metering strategy itself: sub-metering by end use, landlord and tenant separation, and a data connection. That is a design deliverable, it is precise, and it is enforceable before occupation.

Handover information and the occupier

A building that is commissioned correctly and then operated on default settings closes none of the gap.

Building Regulations already require information to be provided to the occupier, in the form of a home user guide for dwellings or a logbook for other buildings. Those documents are often produced as a compliance formality. Written properly they are the only mechanism that reaches the person whose behaviour determines half the consumption.

Three things make them work. Written for the occupier rather than for the file, in plain language with the controls illustrated. Specific to the unit, with the actual system installed and its actual settings. And handed over with a walkthrough rather than left in a drawer.

For buildings pursuing certification, the post-construction stage of a BREEAM Pre-Assessment covers much of this ground, and a licensed assessor verifies it. Aligning the planning discharge evidence with the certification evidence avoids producing two packs.

Summer performance deserves its own handover note, because overheating mitigation often depends on the occupier opening or shading something. Our note on overheating assessment for a planning application covers how those assumptions are modelled, and an assumption about user behaviour in a model is a line in the user guide.

Writing a verifiable commitment

Write the commitment so that a stranger could verify it. That means parameters, evidence and dates.

1. State every performance parameter as a number: U-values by element, air permeability target, system efficiencies, renewable capacity, and the resulting emission and primary energy rates.
2. Name the calculation method, software and version, and the assessor.
3. List the as-built evidence that will be produced: photographic record, test certificates, commissioning records, final calculation.
4. Say who produces each item and at what stage.
5. Set out the metering strategy, including what is sub-metered and how data is collected.
6. Describe the change control process for substitutions during construction.
7. Acknowledge the unregulated loads the calculation excludes, so the reported figures are not misread.

Carry the same numbers into the Sustainability Statement, because two documents with different parameters give the authority two things to enforce and you one problem.

Enter your postcode on PlanScope to see which energy and sustainability documents your scheme is likely to need.

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