Design Decisions Happen in the First Three Weeks

By the time you draw your first wall, the biggest energy decisions are already half-made. Orientation, aspect, and the position of neighbouring buildings set hard limits on the passive solar gain, daylight, and shelter you can realistically achieve. CAD gives you a cheap way to test those limits before anyone orders a brick.

Start with a massing model of the site rather than the house. Include boundary walls, mature trees, and adjacent buildings, then run shadow studies across the year at hourly intervals. A garden that looks generous in July may leave a north-facing kitchen in gloom from October to March. If you are working with a constrained plot, knowing that in week one lets you adjust the plan rather than explain it away later.

As a rule of thumb, due south is ideal for habitable glazing, and anything within about 30 degrees of that still performs well. East and west elevations are harder work: they admit low-angle summer sun that drives overheating, yet contribute relatively little useful winter gain.

Getting Glazing Ratios Right

Glazing is the weakest part of most thermal envelopes, so its area deserves deliberate control rather than aesthetic accident. For a typical UK dwelling, a glazing-to-floor-area ratio of around 20 to 30 per cent per elevation is a sensible working range, with south elevations at the upper end and north elevations kept closer to 15 per cent.

The practical workflow is straightforward:

• Set up wall, floor, and roof build-ups in the model with real U-values, not placeholders. Around 0.15 to 0.18 W/m²K is achievable for walls in a well-detailed new build.
• Schedule glazing by orientation so you can see, at a glance, how much glass faces each direction. Most BIM and CAD packages will do this automatically if windows are placed as objects rather than drawn as lines.
• Check frame U-values separately. A superb triple-glazed unit in a poor frame can perform little better than a decent double-glazed one. Targets of 0.8 W/m²K for the whole window are realistic today.
• Remember that openable sections cost more, leak more, and gain you nothing thermally. Fix what does not need to open.

Insulation, Thermal Bridging and Airtightness

Fabric first remains the most reliable route to a low-energy home, and CAD is where the details either work or fall apart. Aim for a continuous insulation layer that wraps the building without being interrupted by structure. Every steel beam, balcony, or lintel that punches through that layer creates a thermal bridge.

Junction detailing matters more than most designers expect. Default psi values in SAP calculations are penalising — often around 0.15 W/mK — while a well-considered junction can achieve 0.05 to 0.08 W/mK. On a moderately sized house the difference can be several hundred kilowatt hours a year, and it comes almost entirely from geometry rather than extra material.

Draw your junctions at 1:5 in the model, and model them in three dimensions where the sequence is unclear. This is also the moment to check buildability: an insulation layer that cannot be installed in the right order will not be installed at all. Set an airtightness target early — 3 m³/(h·m²) at 50 Pa is a reasonable aim for a good new build — and mark the airtight line on drawings so the team on site knows exactly where it runs. Window reveals, service penetrations, and sole plates are the usual weak points.

Daylight and Overheating Are a Package Deal

More glass means more light and more heat, and in a warming UK climate the second of those is becoming the bigger problem. Analysis in CAD lets you balance them before the design is fixed.

For daylight, an average daylight factor of 2 per cent with a minimum of around 0.5 per cent in kitchens and living spaces is a useful benchmark. Run the analysis under a standard overcast sky, since that represents the worst realistic case rather than the best.

For overheating, the TM59 methodology is now widely used for residential schemes, with criteria such as bedrooms not exceeding 26°C for more than 1 per cent of annual occupied hours. If your model suggests a room will fail, the fixes sit in the design: reduce west glazing, add external shading, size overhangs to the latitude, deepen reveals, or specify glazing with a lower solar gain factor. Internal blinds rarely help much, because the heat is already inside by the time they are drawn.

Keep the Model Live and Useful

A model built purely to satisfy a planning application is a wasted asset. Keep it coordinated as the design develops, with consistent room names, accurate build-ups, and window objects that carry their full specification. When the SAP assessor or energy consultant asks for information, you should be able to hand over a coordinated model rather than a bundle of PDFs and a spreadsheet.

Update the model for as-built changes too. It costs an hour during construction and saves a great deal of guesswork if the client later wants to add a conservatory, extend into the loft, or install a heat pump sized on real fabric performance rather than assumed figures.

Where the Value Actually Sits

The economics are simple. Moving a window in CAD takes minutes; moving it on site takes thousands of pounds and a difficult conversation. Testing orientation, glazing ratios, insulation continuity, and overheating risk during the concept stage is the cheapest energy-efficiency measure available to any project.

You do not need specialist software for all of it. A well-built model, a shadow study, a glazing schedule, and a handful of junction details will answer most of the important questions. Get those right early, and the heating demand, daylight, and comfort tend to follow — quietly, and without anyone having to bolt on a fix later.