C-DRONE GUIDE · 26 AUGUST 2026
Drone and Rendering Plant: ICPE Headings 2730 and 3650, Odour, Neighbours and Prices
At a rendering plant, the year's first meeting with the town council is never about the roof or the cookers : it is about odour. A neighbour smelled something on a Tuesday evening with a westerly wind, a petition is circulating, and the operator has to demonstrate, drawings in hand, exactly where the discharges leave the site, at what height, how far from the nearest houses, and what condition the scrubber and biofilter are in. Yet that information is usually scattered between an old site plan, ground-level photos taken from the fence and the maintenance manager's memory. A drone survey gathers it all in a day, in a georeferenced form an engineering consultant can use. Here is what it actually documents on this kind of site — and above all what it does not measure.
Published on 26 August 2026, reviewed on 29 August 2026 — regulations in force as of August 2026.
Rendering and animal by-products: ICPE headings 2730, 2731 and 3650
A site turning slaughterhouse co-products into animal meal and fat, collecting fallen stock or storing by-products before dispatch falls in France under several headings of the classified-installation nomenclature, often combined on one site. ICPE heading 2730 covers the treatment of animal by-products, including debris, offal and carcasses : it places the installation under the authorisation regime as soon as treatment capacity exceeds 500 kg per day, with requirements set by the order of 12 February 2003. Heading 2731 governs the storage or transit of animal by-products, with two distinct regimes : simple storage in sealed, covered containers, without handling, between 500 kg and 30 tonnes falls under registration (order of 2 October 2015) ; any other installation above 500 kg moves to authorisation, under a separate order also dated 12 February 2003. Those very low quantities explain why even a modest collection point is already a classified installation.
Above a certain throughput, the site changes regulatory category and enters the scope of the European industrial emissions directive : heading 3650, "disposal or recycling of animal carcasses or animal by-products", requires authorisation for any installation whose treatment capacity exceeds 10 tonnes per day — which covers the vast majority of industrial rendering plants. This part of the nomenclature was recently reworked : decree No 2025-617 of 3 July 2025 replaced the notion of "animal waste" with "animal by-products", transposing directive (EU) 2024/1785, and an order of the same day sets the general requirements for installations under heading 3650. That text drives everything that follows : it requires an odour management plan, capture at source and ducting of emissions, discharge through stacks allowing good dispersion, and bans dilution of effluents except where necessary for treatment. In parallel, Regulation (EC) No 1069/2009 of 21 October 2009 classifies animal by-products into categories 1, 2 and 3 according to their health risk and sets out the authorised recovery or disposal routes.
Odour and neighbours: what the 3D model brings to a dispersion study
On this kind of site, the number-one sensitive issue is neither noise, nor traffic, nor even fire risk : it is odour, and the neighbourhood disputes it feeds. A study by Francesca Tagliaferri, Marzio Invernizzi and Selena Sironi, published in 2024 in Chemical Engineering Transactions, illustrates the point perfectly : the authors modelled, using the CALPUFF dispersion tool, the odour impact of a rendering plant while distinguishing its various sources — production workshop, cooking tanks, effluent treatment plant. The results show that the cooking tanks push odour up to the perception threshold (1 European odour unit per cubic metre, at the annual 98th percentile) as far as the nearest residential areas, and that two targeted measures — cooling the tanks and covering the effluent basins — cut the impact by up to 80 % (see the study on Google Scholar). In other words : the effectiveness of an odour management plan hinges on precisely identifying the sources, not on a blanket response.
That is exactly where the drone helps, and only there. A photogrammetric flight produces a georeferenced orthophoto and a 3D model of the site giving, in a few hours and without anyone working at height, the exact position and real height of every emission point (stacks, roof vents, bay extraction outlets, meal silo vents, uncovered basin surfaces), the footprint and height of the buildings — which govern the aerodynamic wake effects that pull a plume back down — and the measured distance to the nearest homes and sensitive premises. These are the input data of a dispersion model, today often reconstructed from an old site plan that no longer reflects the plant's successive extensions. The same image set serves, at a public meeting or in a formal notification file to the authorities, to show a reality nobody sees from the road : a crisp vertical view beats a long paragraph of justification.
The boundary must be stated plainly. The drone measures no odour concentration : the reference remains standardised dynamic olfactometry under EN 13725, with on-site sampling and a laboratory human panel. A study by Javier Burgués, Silvia Doñate, María Deseada Esclapez, Lidia Saúco and Santiago Marco, published in 2022 in Science of the Total Environment, mounted a 1.3 kg electronic nose under a drone, drawing air through a ten-metre tube and analysing it in flight across an array of 21 gas sensors, then benchmarked its odour concentration predictions against dynamic olfactometry at a wastewater treatment plant (see the study on Google Scholar). Such systems are improving fast, but they remain research instrumentation : on a French industrial site in 2026, regulatory odour data comes from the laboratory, not from the drone.
Discharge stacks, scrubbers and biofilters: clearance and condition
The order of 3 July 2025 applicable to heading 3650 installations is explicit : emissions must be captured at source, ducted and discharged through stacks allowing good dispersion, dilution being banned unless it serves a treatment purpose. Compliance with that logic is partly verified by eye, at a height nobody climbs day to day. A targeted inspection flight documents each stack outlet and its immediate surroundings : a taller building or plant room built nearby that disturbs vertical dispersion, a cowl or deflector added over the years, an emergency vent left open, vegetation or stored material masking a low-level discharge. It also documents the condition of the structures themselves : corrosion of the shell and flanges, an open seal at a sampling port, a degraded caged ladder or sampling platform — a point that directly determines whether the next statutory test by an accredited body is even feasible. This service extends the one described in our guide to drone inspection of industrial chimneys and silos.
The scrubbers and biofilters treating air extracted from workshops and reception bays deserve the same pass. On a scrubber, a close flight reads external corrosion of the tower, the condition of ports and sumps, a run mark revealing a leak, obstructed access. On an open-surface biofilter — a common layout in rendering —, the vertical view is especially telling : at a glance it reveals compaction zones in the media, craters and preferential paths through which air escapes untreated, dry patches, encroaching vegetation or moss, the condition of the humidification rails. These are precisely the faults that sink abatement performance without any control system noticing. Here too the limit is clear : the image says nothing about pressure drop, actual media moisture or abatement efficiency, which are the province of fixed probes, operational monitoring and standardised measurement ; it directs an intervention, it does not conclude one.
Workshop roof, steam network, basins and fence: the rest of the site
The roof of a workshop cooking and pressing animal by-products ranks among the most punished in the whole food industry. From below it constantly takes the condensation of fat-laden cooking vapour ; on top it receives vent discharges, greasy deposits around extractors and an often ammoniacal or sulphurous atmosphere that accelerates corrosion of steel sheeting, fixings and flashings. Nobody climbs it willingly, and the critical details — duct penetrations, extractor outlets, gutters, translucent panels grown brittle — are exactly the ones that fail first. A study by Rafael Lemos, Ricardo Cabral, Diogo Ribeiro and co-authors, published in 2023 in Applied Sciences, built a database of more than 8,000 high-resolution aerial images captured by drone over large-footprint industrial buildings and developed a deep-learning method for automatically detecting corrosion on their roofs (see the study on Google Scholar) — proof, on this precise kind of building, that drone coverage has become an analysis input and not just an illustration. This service falls under our drone roof inspection offer, in line with what we describe for the upstream end of the chain in our guide to drone inspection of a slaughterhouse.
The thermal pass here has two useful targets. First the external steam network, ubiquitous on a rendering site : stretches of insulation torn off or waterlogged, a steam trap stuck open betrayed by an abnormal thermal signature downstream, losses on rooftop headers. Second the shell of the cookers and dryers where these are accessible from outside, along with condensers and air-cooled condensers. The camera reads a surface temperature, nothing more : it calculates no efficiency, does not replace an energy audit run with meters, and above all in no way substitutes for the periodic requalification of pressure equipment by an approved body, which covers most cookers and dryers in this sector. On a site where steam is the leading energy cost, it nonetheless remains an excellent prioritisation tool, whose payback can be quantified using the method set out in our guide to calculating the ROI of an industrial drone inspection.
Three further areas are covered in passing, with no extra flight. The pre-treatment basins handling effluent — flotation, grease removal, buffer, sometimes a lagoon — are surveyed by photogrammetry to quantify silting and the actually available useful volume, check bank and edge structures, and spot a torn floating cover : the method is identical to the one detailed in our guide to drone inspection of a wastewater treatment plant. The reception yard and enclosed bays are documented to check what the neighbourhood actually sees and smells : bay doors left open during unloading, containers parked outside cover, cleanliness of collection-truck wash areas, traffic direction. Finally the fence line and biosecurity : the entire perimeter covered in a single pass, breaches, vegetation leaning against it, gates and fly-tipping spots on the boundary — a reading detailed in our drone surveillance and security offer. If the nuisance reported by neighbours also concerns noise from extraction fans or air-cooled condensers, our guide to noise mapping and sound source localisation describes the complementary approach.
Mission method, coactivity and prices
The mission fits into half a day for a medium-sized site, a full day where the 3D model must feed a dispersion study and cover every building with its heights. It starts with a briefing with the QHSE or environment manager : the list of emission points to record, stacks and biofilters to inspect closely, time slots when condenser vapour will not spoil the imagery, collection-truck traffic, a clean take-off point well away from the bays, visitor protocol and biosecurity instructions. The flight then runs a high photogrammetric pass for the orthophoto and the georeferenced 3D model, close manual-flight shots of each stack, scrubber and biofilter, a thermal pass over the steam network and the roof, then a survey of the basins. As with any mission on an industrial site, the prevention plan and coactivity with the teams are settled before flight day, not in the car park at 8 am.
Two checks frame the quote. On the airspace side, a rendering plant usually sits in an isolated business zone or a rural area, a favourable layout, but it is not unusual for one to lie near an aerodrome or under a restricted zone : checking Géoportail is systematic before any commitment, and the specific arrangements described in our guide to flying a drone in French cities apply as soon as homes back onto the fence. On the cover side, up-to-date professional liability insurance is a non-negotiable prerequisite for working on a site operating under authorisation. Finally, a reminder of what none of this ever replaces : standardised olfactometric measurement, testing of air and water discharges by an accredited body, the site's sanitary control plan, and the periodic requalification of pressure equipment.
2026 prices (excl. VAT) : €600 to €1,200 for a georeferenced 3D model and orthophoto of the site, recording emission points, building heights and distances to the nearest homes, delivered in a format an engineering consultant can use ; €1,200 to €2,200 for a full campaign combining the 3D model, close inspection of stacks, scrubbers and biofilters, roof coverage and a thermal pass over the steam network ; €350 to €700 for a targeted inspection of a single discharge stack or biofilter ; €300 to €600 for a standalone photogrammetric survey of the pre-treatment basins ; €250 to €500 per visit for recurring monitoring on the same flight plan. A travel charge applies beyond a 30 to 50 km radius. Request a quote stating the heading and regime the site falls under, the number of discharge points to document and the goal (dispersion study, formal notification file, preparing an inspection visit or a maintenance plan).
Frequently asked questions
Can a drone measure odour at a rendering plant?
No, not as part of an aerial imagery service. Odour measurement is the domain of standardised dynamic olfactometry : sampling at a hood or a stack, laboratory analysis by a human panel under standard EN 13725, within a tight window after sampling. Research systems carrying an electronic nose under a drone do exist and are improving, but they are specialised instrumentation, not a standard photo flight. What the drone brings is of a different order : the geometry of the problem — the real position, height and outlet of each emission point, building footprints and heights, exact distance to the nearest homes —, that is, the input data for a dispersion model, not its results.
Is an overflight compatible with the site's biosecurity protocol?
Yes, and that is one of its selling points : the aircraft does not enter dirty zones, does not move between reception bays and never touches a vehicle or a container. Take-off is from a clean point agreed with the operator, well clear of collection-truck traffic ; the equipment is cleaned and disinfected under the site's visitor protocol before and after the mission, like any other incoming kit. The remote pilot follows the required visitor route, including airlocks and mandated protective equipment. None of this changes the sanitary control plan or the obligations arising from Regulation (EC) No 1069/2009 on animal by-products : the drone complies with them, it does not replace them.
Does the survey report replace statutory stack emission testing?
No, at no point. Stack concentrations and mass flows — dust, hydrogen sulphide, ammonia, organic compounds — are measured by standardised sampling, carried out by an accredited body at the frequency set in the site's prefectural order. The drone documents the condition of the structures and the clearance of the outlet : shell corrosion, an open seal, a degraded access platform, an obstacle added near a vent, biofilter media that is compacted or overgrown. These are dated, located visual findings, useful for preparing an environmental inspectorate visit or prioritising maintenance ; they produce no emission figure whatsoever.