Thermal Comfort Retrofits in Historic Urban Buildings

Historic buildings don't underperform. They run on different physics. Thick masonry stores heat and releases it slowly, a property known as thermal mass. Windows leak air on purpose, feeding ventilation that a modern sealed building has to fake with fans. Occupants adjusted their clothes and their daily rhythm to match the building long before anyone owned a thermostat, and none of that shows up as a flaw on an energy audit. It shows up as a building doing exactly what it was built to do, just not what a 2025 energy code wants it to do.

So when a retrofit team shows up with the standard playbook of adding insulation, tightening the envelope, and swapping in new HVAC, that playbook breaks. Moisture gets trapped in walls that were designed to breathe, a failure mode hygrothermal analysis is specifically designed to predict. Mechanical systems chew through plasterwork that can't be replaced at any price. Energy models built for airtight new construction misread how a 200-year-old wall actually loses and stores heat, and the numbers come out wrong before anyone's touched a wrench.

The UK alone carries around 400,000 listed buildings and buildings sitting inside conservation areas, and historic buildings as a category account for roughly 5% of total UK emissions. That's not a figure the country can ignore on the way to net zero. There is also no universally agreed method for assessing energy efficiency potential in historic buildings, according to a review by Martinez-Molina and colleagues, which means every project starts without a shared baseline. The challenge isn't hitting energy targets. It's hitting them without most of the tools that make hitting them straightforward anywhere else.

Regulations practitioners are actually navigating now

The EU's revised Energy Performance of Buildings Directive, in force since May 28, 2024, sets Minimum Energy Performance Standards requiring the worst-performing 16% of floor area to comply by 2030, and 26% by 2033. That's a hard deadline attached to a hard number, and it applies to buildings that predate the light bulb.

EU households burn through more than 75% of their energy on heating, cooling, and hot water. Buildings are the primary lever for decarbonization, and the directive reflects that by setting binding renovation targets.

The directive carves out flexibilities for heritage buildings, and member states have to spell out exemption criteria in their national renovation plans. How the carve-out actually applies still varies country to country, and no member state is operating from a finished rulebook yet.

The UK runs on a separate track entirely, with Minimum Energy Efficiency Standards (MEES) setting the domestic compliance framework. Listed building consent governs what can and can't be touched, and the tension between energy obligations and conservation status gets resolved project by project, not settled once at the national level. Practitioners end up working two systems at once, energy rules and heritage consent, and those two systems don't always agree on what "good" looks like.

How preservation constraints limit available technical options

Start with the wall. External insulation is the easiest and most effective envelope fix in new construction, but on a historic façade it is usually banned outright, which removes the default tool before the project even begins.

Interior work isn't much easier. Ornate plasterwork, original joinery, and historic floor structures all block the routes ductwork and conduit would normally take. Guidance on historic building HVAC work is consistent on the core principle: minimize intrusive modification to original material.

Mid-century modern buildings present additional difficulties: poured concrete, floor-to-ceiling glass, low ceilings, and decorative exterior elements combine to restrict retrofit options significantly. Guidance for this typology is sparse, and the profession's default answer has too often been demolition instead of retrofit.

Thermal bridges are widespread in pre-modern construction: stone lintels, solid masonry returns, and wall-to-floor junctions with zero insulation. Fixing those from the interior eats floor area with no clean geometric workaround. A review of 69 case studies on historic building energy retrofits, published in Heritage in February 2024, identified a recurring cluster of compounding challenges: heritage evaluations, alteration restrictions, overlapping authorities, technical limits, higher costs, and knowledge gaps. The practical implication is that a heritage significance assessment must be completed before any solution is drawn up, not after.

Internal insulation: the default with real limitations

Since the exterior is usually off-limits, internal insulation applied to the inside face of the exterior wall ends up as the only envelope fix available in most historic retrofits. That default status doesn't make it a complete solution on its own.

Interstitial condensation forms at the now-colder original wall face, and mold follows. These are documented risks that run worse in solid-wall historic construction than in cavity-wall modern buildings. Skipping a properly designed vapor control layer and a wall-specific hygrothermal assessment means the insulation intended to protect the fabric ends up degrading it instead.

Material choice and achievable U-value carry more weight here than in new builds. A study on a 1924 Seoul building tested seven insulation materials, including EPS, XPS, PU, PF, glass wool, mineral wool, and vacuum insulation panels, and all seven met Korea's national U-value requirement of 0.240 W/m²·K at 120mm thickness. Vacuum insulation panels and phenolic foam pulled ahead at thinner depths, which matters when every centimeter of floor area is already spoken for.

Reversibility is another critical filter. Heritage frameworks commonly favor interventions on designated buildings that can be undone if a future owner wants them removed. Not every insulation system is designed with removal in mind, which is a practical constraint worth assessing alongside thermal performance. Internal insulation is one effective layer in a broader strategy, not a complete approach by itself.

Interior finish replacement as a preservation-compatible fix

Sometimes the original interior finish is already gone, or was never heritage-significant to begin with. That creates an opportunity to substitute high-performance composite materials, gypsum-based products engineered for thermal storage and insulation, without touching the exterior or the primary structure at all.

A 2024 study in Energy and Buildings tested this on a Korean historic campus building under two scenarios: composite applied to walls alone, and composite applied across walls plus ceilings and roofs. Applying the composite to interior surfaces produced energy savings of 20 to 30% against the pre-retrofit baseline. The fuller scenario cut annual heating and cooling consumption by 31 to 42%, and the gap between the model and actual measured use came out to 13.06%, tight enough to use the simulation for decisions rather than treat it as an approximation.

What makes this approach attractive to conservation officers is that the work stays entirely inside, it is reversible in principle, and it coincides with material replacement that interior finishes require eventually regardless. The approach works best in lighter, frame-based historic structures. In buildings with very thick stone walls and already-high thermal mass, additional storage material contributes relatively little.

Mechanical systems where ductwork routes don't exist

Thick stone, ornate ceilings, and floor voids with limited clearance mean that the routing assumptions HVAC engineers rely on in new construction simply don't apply. The system has to fit the building, and selecting the wrong one early is the single most expensive mistake on most historic retrofit projects.

Variable refrigerant flow (VRF) systems have become a leading solution for exactly that reason. They move refrigerant instead of air or water, require minimal structural penetration, and can run multiple zones off one outdoor unit. A BIM-based evaluation of a 1933 museum building found VRF was the strongest performer available, cutting the average PMV index for the whole building during the cooling period by up to 31%.

Mini-split and multi-split units follow the same logic: small penetrations, no ductwork, and straightforward removal later. The trade-off is visual impact, since a wall-mounted unit in a room with original cornicing requires negotiation with whoever holds heritage authority over the building.

Radiant systems, whether underfloor or ceiling-embedded, can work if the floor is being replaced anyway. They are usually off the table where original floorboards, tile, or stone carry protected status. Sequencing also matters critically: selecting the mechanical system before the envelope has been assessed means the system runs continuously compensating for heat loss that a better envelope would have reduced, which inflates both capital and operating costs.

Passive measures practitioners frequently undervalue

Most projects prioritize mechanical systems and insulation first and treat shutters and curtains as a low-priority addition. The evidence points in the opposite direction.

Secondary glazing and thermal shutters cut heat loss by an amount comparable to full window replacement, while leaving the original glazing untouched, which heritage consent almost always requires anyway. The lower-intervention option frequently delivers performance close to the higher-intervention alternative that isn't legally available.

The whole-building approach that practitioners in the field consistently advocate treats passive measures as the first layer of intervention, not the fallback. Characterizing how the envelope behaves, how air moves through it, and how sunlight hits it across the day before specifying any active system allows the team to establish a lower peak load, which in turn reduces the size and cost of the mechanical system specified next.

Measuring whether a retrofit actually improved comfort

PMV (Predicted Mean Vote) and PPD (Predicted Percentage of Dissatisfied) are the standard metrics and hold up well for comparing retrofit scenarios against each other. The 1933 museum study used PMV reduction as its primary performance metric for exactly that reason.

Both metrics have known blind spots in historic buildings, particularly in large spaces with heavy thermal mass. The gap between what PMV predicts and what occupants actually report widens during the cooling season because the thermal behavior of a stone structure doesn't conform to the model's assumptions, and neither does the behavior of its occupants.

Pairing occupant surveys with sensor data addresses this by establishing a neutral comfort temperature specific to the building, rather than relying on the model alone. People in heritage spaces tend to arrive with adapted expectations, different clothing choices and activity levels than standard formulas assume, which means a comfort model built on those formulas can produce systematically misleading results.

Building energy models built on BIM should be checked against real energy bills before any retrofit path is committed to, and the 13.06% error rate from the Korean campus study provides a reasonable benchmark for what a validated model should look like. Post-occupancy evaluation is equally necessary. A retrofit that performs well in simulation but poorly in actual use has failed, and without post-occupancy measurement that failure remains invisible to the team responsible for it.

Sequencing interventions so each one supports the next

The failure mode that repeats across these projects follows the same pattern: something was installed before the building was fully understood. Insulation goes in without hygrothermal modeling. HVAC gets sized for a tighter envelope that was never actually delivered. Passive measures get cut because the budget was committed to the mechanical system first.

The correct starting point is always characterization: a historic fabric survey, an energy audit, a hygrothermal assessment, and a heritage significance assessment. Those four together define what is achievable before anyone specifies a product or system.

Envelope work comes before mechanical work because a better envelope reduces peak load, which reduces the size and cost of the mechanical system required to meet it. Combined envelope and HVAC upgrades have demonstrated meaningful reductions in both energy consumption and overall project cost, though outcomes in 2025 still fall short of zero-emission standards in most cases.

Passive measures come before active ones. Secondary glazing, shutters, and thermal curtains reduce the heating and cooling load that insulation and mechanical systems then have to cover. Sizing a mechanical system against a building before its passive performance has been improved means purchasing capacity that would never have been needed.

Heritage consent needs to run alongside the technical process, not be added at the end. Bringing in a conservation officer after the technical solution is already fixed means redesigning under deadline pressure. Involving them earlier changes what gets proposed in the first place, which produces better outcomes than negotiating changes to a design that is already locked.