Ask a conservation officer, anywhere in Europe, whether you can work by drone on a listed building (classé or inscrit): you will almost always get a polite version of the same answer. That is not the question I am examining: this article is about listed building cleaning by drone, and what heritage architects actually require.
Heritage consent concerns the intervention, not the machine. What is assessed is what will come into contact with the historic fabric, with what energy, in what quantity, leaving what residues — and whether the intervention is needed at all. The aircraft is merely a means of application: it matters for the method statement, airspace and species protection, but not for the conservation question itself.
You might think this deflates the premise. It is in fact the strongest argument in its favour, provided you follow it through to the end. Conservators’ main objection to cleaning facades and roofs has never been about how the workers reached the surface. It has always been about water and pressure. And an airborne applicator is physically incapable of delivering the pressure regime behind most of the damage. The conservation benefit comes from a constraint, not a capability.
Listed building (classé or inscrit): what the heritage authority actually examines
The institutional landscape is varied — the Architectes des Bâtiments de France (in French), the Italian Soprintendenze, the German Denkmalschutzbehörden, the Dutch Rijksdienst voor het Cultureel Erfgoed, Historic England and Historic Environment Scotland all operate under different texts. Yet the assessment logic converges, because it stems from the same doctrinal source: ICOMOS’s Venice Charter and the texts that followed it, which established minimal intervention, compatibility, reversibility where possible and full documentation as operational principles.
CEN/TC 346 has since turned many of these principles into published European standards. The ones that matter here are EN 15898 (conservation terminology), EN 16096 (condition survey and report of built cultural heritage), EN 16682 (measurement of moisture content in materials constituting immovable cultural heritage) and, even more directly, EN 17138:2019 — Methods and materials for cleaning porous inorganic materials. A cleaning specification for protected masonry that does not draw on this last document is not a conservation specification.
In practice, the requirements are consistent:
- Characterisation of the substrate. Not ‘sandstone’, but which sandstone, which binder, in what condition, whether it has already been consolidated or coated.
- Justification of need. Why cleaning is needed rather than leaving it as it is.
- The gentlest effective method, demonstrated rather than merely asserted.
- Test areas (trial panels), assessed after a realistic interval, not on the same day.
- A before-and-after condition report, under EN 16096 or a national equivalent.
- Named products, with their regulatory status, which, for biocides in the EU, means authorisation under the Biocidal Products Regulation (BPR) for the product type and use concerned.
Nothing on this list is satisfied by a description of the aircraft.
High-pressure cleaning: why pressurised water is the real objection
The mechanism is worth understanding properly, because it explains why the prohibition is so consistent from one national guidance document to another.
Surface loss. Many limestones develop a hardened surface — a skin enriched in calcium carbonate, sometimes called calcin, formed by dissolution and redeposition over decades. It is harder and less permeable than the underlying stone, and it is thanks to it that a weathered block of ashlar has come through two centuries. Remove it and you expose an open, friable core. The US National Park Service guidance is blunt about the threshold: start at 100 psi or with a garden hose, and as a rule do not exceed 300–400 psi; if grains of stone or sand appear in the water at ground level, the pressure is already too high. An ordinary pressure washer runs at 100–150 bar, or about 1,500–2,200 psi.
Differential erosion of the mortar. Historic lime mortars are deliberately weaker than the units they bind. That is the whole point: the joint is the sacrificial element, designed to absorb movement and give way before the stone. High-pressure cleaning preferentially erodes the softest material in the assembly, that is, precisely the element the wall depends on.
Water volume and salts. This is the most frequently underestimated mechanism. Saturating masonry mobilises the soluble salts held in the pore network. As the wall dries, these salts recrystallise — at the surface as efflorescence, which is cosmetic, or below the surface as subflorescence, which generates crystallisation pressure in the pores and causes disaggregation and spalling. The NPS guidance states that masonry can take more than a week to dry after cleaning, and that water cleaning should not be carried out in the few days before the average date of the first frosts. In continental and Alpine Europe, this seasonal window is narrow; in Mediterranean Europe, the salt load of coastal masonry increases the risk whatever the season.
The saying that circulates among stone conservators is that about ninety per cent of the risk lies in removing the last ten per cent of the soiling. All cleaning, however gentle, causes some surface loss — the whole judgement lies in its extent and its justification.
Drone facade cleaning: what changes when the applicator flies — and what does not
The chemistry does not change. A quaternary ammonium biocide behaves exactly the same way whether it comes out of a backpack sprayer or a nozzle twelve metres up. The idea that aerial application is inherently gentler is true only for one precise, mechanical reason: the physics of flight rules out the alternative.
A water jet generates a reaction force equal to the mass flow rate multiplied by the jet velocity, that velocity being approximately √(2ΔP/ρ). At 100 bar and 10 L/min, that represents about 24 N — nearly 2.4 kgf of continuous horizontal thrust against a small aircraft. At 150 bar and 20 L/min, it approaches 58 N. These are not mere trim disturbances: they become dominant, which is why free-flying systems are confined to low-flow, low-pressure application. The platform structurally caps the energy that can be applied to the historic fabric. For a conservation officer, that is a significant and verifiable constraint.
But two new variables come with it.
Rotor downwash. A multirotor generates a strong, continuous airflow directed at the surface, forming a wall jet with recirculation zones near the facade. On sound ashlar, this is of no consequence. On render that is coming away, friable sandstone, eroded joints, a flaking limewash or a previously consolidated surface that is starting to scale, a directed airflow is a mechanical load applied exactly to the material most likely to detach. I have found no published data on rotor downwash speed thresholds for friable historic surfaces, and I would treat any precise figure with suspicion — but the mechanism is in no way speculative, and stand-off distance becomes a conservation parameter rather than a piloting preference.
Control of deposition. On a flat elevation, coverage remains manageable. On a modelled elevation — cornices, drips, string courses, carvings, tracery —, the shadow created by rotor downwash leaves untreated pockets, while gravity produces a dose gradient down the elevation and wash-off from projecting elements onto the courses below. Spray drift onto leaded stained glass, historic ironwork, polychromy or neighbouring planting poses a containment problem that a ground-based operator solves by masking, and that a drone operator largely cannot.
Moss removal and lichens: the objection that has nothing to do with the drone
Before all this comes a prior question, which heritage advisers raise more and more often and which sales talk rarely addresses: should this growth be removed at all?
The assumption that biological cover damages stone is only partly supported. The literature now treats the subject as a genuine controversy — recent review work in the microbiology press frames it explicitly in terms of biodeterioration or bioprotection, and the answer seems to depend on the species and the substrate rather than on a general rule. Work on the effect of lichens on sandstone hardness has shown that the impact depends on the species rather than being uniformly negative, and studies on oolitic limestone have argued for a bioprotective role, with the thallus shielding the surface from rain, thermal cycling and dissolution.
The practical consequence is uncomfortable. Removing an established lichen cover from a soft limestone may expose a surface that has been protected for decades and is neither sound nor hardened. Cleaning then triggers a deterioration process that the growth was holding in check. Whether this applies to a given building depends on the stone type, the lichen species, the exposure and the existing condition — and that is precisely why characterisation comes before the choice of method, and why ‘the facade is green’ is not a condition diagnosis.
There is a second, more nuanced version of the same argument. On many listed buildings (classé or inscrit), weathering and biological patina are part of the recognised value of the masonry. Cleaning then becomes an aesthetic intervention that must be justified as such, not a neutral act of maintenance.
Biocide residues raise another question that I can only flag without answering: the long-term behaviour of quaternary ammonium residues in a porous heritage substrate, and their interaction with later conservation treatments such as consolidants, limewashes or sacrificial protective coatings, is not, to my knowledge, well characterised in the published conservation literature. That uncertainty is in itself a reason to be cautious.
Protected species and drone regulations: the constraints nobody budgets for
Protected species. Historic buildings — bell towers, roof spaces, cavities behind render, blocked-up flues — are prime roosts for bats (in French) and nesting sites for birds. Bats are strictly protected throughout the EU under Annex IV of the Habitats Directive and the EUROBATS agreement, with their breeding sites and resting places protected whether or not animals are present at the time of the works. Birds and occupied nests are protected under the Birds Directive. Guidance for historic buildings generally requires works to be scheduled around roost occupancy, and demands a recent survey before any absence is assumed. A hovering aircraft producing noise, downwash and a chemical plume next to a bell-tower louvre raises a disturbance issue, not a mere formality, and it is better dealt with before than after.
Airspace and authorisation. Protected sites are frequently located in geographical zones designated under the EASA framework, and operations near buildings and uninvolved people generally fall outside the Open category, which requires an operational authorisation based on a risk assessment. That is an administrative lead time, running in parallel with heritage consent — not in its place.
When drone cleaning is really not suitable
A balanced assessment must clearly name the exclusions.
Aerial chemical treatment is not the right tool when the surface is friable, delaminating, or has been consolidated and is now scaling, because both rotor downwash and wetting stress the failing material. It is not suitable for sacrificial and decorative layers — limewashes, sacrificial protective coatings, earth and cob renders, exterior polychromy, murals —, where water or a surfactant can destroy what you are trying to conserve. It is not suitable for salt-laden or gypsum-rich masonry, where any additional water input is the main risk, whatever the pressure. It is not suitable for sculpture or highly modelled elevations, where coverage is hit-and-miss and drift impossible to contain.
Nor is it suitable where the prescribed method requires controlled contact — poultices (including peel-off pastes such as Arte Mundit), nebulisation or misting systems applied over hours, laser ablation under EN 16782 —, because these techniques rely on contact time, containment and continuous conservator judgement that an airborne applicator cannot provide.
And it is not suitable where verification relies on touch. A conservator who taps render to find hollow delamination, or who feels whether a stone face is sound, is carrying out a test that no camera reproduces. When the condition diagnosis itself requires hands on the fabric, the access must allow hands on the fabric.
Listed building drone cleaning: where it really has its place
The credible scope is narrower than marketing claims and wider than sceptics admit: large, simple, robust historic masonry in good condition, where the problem is a phototrophic biofilm rather than a sulphation crust, and where close access would be mobilised for cleaning alone. Late nineteenth- and twentieth-century brick and engineering-brick elevations, industrial heritage, hard slate or clay tile roofs on listed buildings (classé or inscrit), large areas of sound ashlar — these are the candidates. For these substrates, there are water-based stone cleaners that their manufacturer presents as compatible with drone application (for example SCALP PIERRE AQUA 89), which does not remove the need for test areas or for the opinion of the conservation officer.
If you are preparing an application, the order of steps is fixed and the drone comes last. Characterise the substrate and record its condition. Establish that cleaning is necessary and what value the existing surface carries. Choose the gentlest method that achieves the defined objective, and test it on test areas assessed after a proper interval. Name the product and its authorisation. Commission the species survey. Then, and only then, specify how the product reaches the wall — and it is at this stage that the aerial option is assessed like any other, on stand-off distance, exposure to rotor downwash, drift containment and monitoring.
To be honest, the drone has not solved any conservation problem. It has made a particular category of low-intrusion treatments logistically viable on buildings where mobilising close access for a light, recurring task was previously disproportionate. It is a real but limited gain, and it evaporates entirely whenever the right answer is to leave the surface as it is.
Sources
- EN 17138:2019 — Conservation of cultural heritage: methods and materials for cleaning porous inorganic materials (in French)
- EN 15898:2019 — Conservation of cultural heritage: main general terms and definitions
- EN 16782:2016 — Cleaning of porous inorganic materials: laser cleaning techniques (in French)
- NPS Preservation Brief 1 — Cleaning and Water-Repellent Treatments for Historic Masonry
- Building Conservation Directory — Can Masonry Cleaning Ever Be Considered Truly Safe?
- Biofilms on stone monuments: biodeterioration or bioprotection? — Trends in Microbiology
- Lichen impact on sandstone hardness is species-specific — Earth Surface Processes and Landforms
- Bioprotective role of lichens on oolitic limestone buildings
- EU Habitats Directive
- EUROBATS — Protection of overground roosts for bats
- Historic England — Building Works and Bats
