Historic Building Conference 3: Insulating Historic Buildings

This is a collection of notes from the Historic Building Conference organised by CPD Essentials and held at the 16th century Shaw House in Newbury, Oxfordshire on 3rd June 2026. I am collecting them here to build a record of my learning, creating a treasure trove of notes that I can refer back to in the future, but also make them available to others.

Marianne Suhr's second session was on insulation, and the headline is that building physics, not product choice, is what keeps you out of trouble. Get the physics wrong and a well-meant upgrade can create damp, mould and rot exactly where you least want it.

Keep the wall dry first

A damp wall conducts heat far faster than a dry one, the way a wet oven glove burns you, so simply drying a wall out can double or triple its insulation value before you add anything. That usually means fixing the cause, gutters and repointing, and letting the wall breathe by stripping cement render or gypsum plaster and going back to lime.

A wet wall dries by three mechanisms. External evaporation driven by wind and sun. Internal surface evaporation, if the inside surfaces are breathable. And the role of outward heat flow carrying moisture towards the outer face. Marianne recommended Robin Pender's Historic England Technical Tuesdays talks on drying out damp buildings.

Breathability itself is not one thing but three. Water-vapour permeability, which is vapour diffusion. Capillarity, the liquid-water, blotting-paper movement. And hygroscopicity, the buffering of moisture by storing and re-releasing vapour as humidity changes. All three matter once you start insulating.

Question whether to insulate the wall at all

Solid-wall U-values have probably been overestimated. A rubble stone wall, with its rough faces and rubbly, air-pocketed core, is more thermally efficient than the textbooks assume, and a typical 800 millimetre rubble wall is already reasonably insulated. A theoretical nine-inch solid brick wall is around 2.1, a 1965-era cavity wall around 1.7, while current regs target around 0.3 for an upgraded wall, which is a huge jump. So the first question is whether it is worth it. With rubble stone, often not. With nine-inch brick, possibly, but once you subtract internal walls, party walls and floors, the actual external area left to treat can be small. One small rubble house she worked on, just stripped, deep-repointed in non-hydraulic mortar and dried out, came out around 1.19 with no added insulation at all.

The thermal gradient is the key idea

Across a uniform wall the temperature falls in a straight line from about 20 degrees inside to about 5 outside. What you do to that line determines whether you create a condensation problem.

• Internal insulation with foil-faced foam blocks the heat flow at the inner face, so the masonry behind goes cold. You then rely on a perfect vapour-control layer, and the moment a trade cuts a socket into it, vapour reaches the cold masonry, hits dew point, and you get interstitial condensation inside the wall.

• The first mitigation is to use wood-fibre board instead of foam. It comes in 20 millimetre multiples, is hygroscopic, vapour-permeable and somewhat capillary-active, so it stores and blots moisture rather than dumping it at the cold interface, and can re-evaporate it back inward, finished with breathable lime plaster. You parge-coat the masonry flat, fix the board directly with no unventilated air gap, and plaster directly onto it. You need roughly twice the thickness of wood fibre to match foam. Crucially, get a hygrothermal analysis from the supplier, a dynamic dew-point risk assessment that models orientation, exposure and build-up, rather than the old static method.

• The second mitigation is less is more. A thinner layer of wood fibre lets a little heat flow through, keeping the masonry warmer and pushing the dew point further towards the outer face, out of the risk zone.

Exposure and driving rain

Exposure matters enormously. The west and south-west of the country get up to three times the rainfall of the east, and on a thin wall rain can reach the inner face before it can dry. Foam can trap that driving rain, whereas wood fibre at least lets some evaporate back. The traditional answers to exposure were cladding in tile, slate or weatherboard, and lime roughcast or render. Roughcast is clever because its textured surface has around nine times the surface area of a flat render, so it both absorbs and then evaporates far more, drying quickly once the rain stops. Insulating lime renders, some containing cork, are fully breathable and lightweight, and even on their own they help.

Internal versus external

External insulation is thermally far better, because the heat passes through the masonry and is stopped at the outer face, turning the wall into a storage heater that radiates warmth back in and using its thermal mass, while also protecting joist ends and timber lintels. But it is rarely appropriate on a historic or listed elevation, because every projecting feature, eaves, downpipe and reveal becomes a thermal bridge, and a date plaque she showed had left a cold square of mould. Done badly externally it is a disaster, and she quoted alarming figures on external wall insulation projects needing remedial work and large remediation costs. The takeaway was not never, but understand the risks, detail it carefully and supervise it, as on a successful project where careful carpentry, a stepped plinth, extended sills and lintels and a lime render gave roughly three times the insulation and still looked good ten years on.

Thermal bridging gets worse the more you insulate

Miss a patch and it becomes a major bridge and a mould spot. Window reveals and bay windows are tricky, and the one to watch is joist ends built into external walls, because internally insulating around them creates a cold, damp pocket exactly where structural timber sits, risking rot. You can pack around the joists, with some risk, or in a deep retrofit cut them back and re-hang them.

Lofts

Lofts want around 270 millimetres, with diminishing returns beyond. But adding loft insulation cools the loft, which raises relative humidity, since as temperature drops RH rises, and around 75 per cent is roughly the mould-growth threshold. So a previously dry loft can start to grow mould on the felt and on stored items simply because you insulated the floor below it. You counter this with good loft ventilation, airtightness at ceiling level including sealed loft hatches and recessed downlights, which leak warm moist air upward, and proper extract ducting taken right outside rather than left coiled in the loft pumping moisture into a cold space.

Two related warnings. Breathable roofing membranes rely on a vapour-pressure differential that often is not present, so they are not a reliable cure for condensation. And there is a serious bats problem, with bats becoming tangled and killed in the loose fibres of some breathable membranes, which she thinks may be the beginning of the end for them.

Use natural quilts such as wood fibre, sheep's wool, hemp or recycled denim rather than mineral wool or fibreglass, because the natural quilts are hygroscopic and buffer moisture, and because she objects to sending people into lofts to handle irritant mineral fibres in the first place.

Sloping and attic ceilings

These are best insulated from above during a re-roof. Never just stuff foam board between rafters, because of gaps, thermal bridges, no acoustic mass, and no decrement delay, the time it takes for summer heat to pass through, so a foam-lined room under a black slate roof bakes. The rule she gave is the wrinkled sandwich, where you can have as many breathable layers as you like as long as the top layer is actively ventilated, with a counter-batten creating a ventilated void above the breathable membrane. The common mistake is no counter-batten. The best build-up she showed was flexible wood fibre full-fill between the rafters plus a rigid tongue-and-groove wood-fibre board over the top, which gives excellent acoustic and thermal performance and decrement delay. As for spray foam, the advice was simply do not, and when surveying check whether it was sprayed onto a membrane, which is removable, or straight onto the tile undersides, which ruins the roof and tends to get it condemned.

On regulations, Part L allows relaxation for historic and listed buildings, there is good guidance from the Sustainable Traditional Buildings Alliance, and you do not have to meet the regs where doing so would unacceptably alter character or appearance, or risk long-term deterioration of the fabric.

Floors

Never lift a lovely old floor unnecessarily, since centuries of foot traffic are part of its significance. The biggest floor problem is raised external ground levels, so always check the external ground is below internal floor level. A floor on sand or earth that has gone damp is usually just trapped moisture from something laid over it, and allowing air circulation will dry it out. Where a floor genuinely must be replaced, limecrete floors are a game-changer, built up from a structural, insulating, capillary-breaking base of foamed recycled glass aggregate, which needs no plastic membrane because the large voids stop capillary draw, then a geotextile to hold back silt rather than moisture, compacted, then limecrete of lime and sand with no cement laid over the underfloor-heating pipes. The result is thermal mass, full breathability and a slow-response floor that pairs well with a heat pump. Dig down gently so as not to undermine the wall footings, and allow around a hundred days for the limecrete to dry.

The summary mantras were fabric first, deal with draughts before insulating, prioritise airtightness in exposed situations, render where appropriate, always get a condensation-risk analysis and proper advice, use experienced tradesmen, treat insulation as a major intervention, and design in future inspection.