Counting the Carbon a Film Leaves Elsewhere

1. The big number is a model, not a property of every blockbuster

A Screen New Deal, published by Arup with the BFI and BAFTA albert in 2020, gives the screen industry one of its most repeated figures. In the report’s analysis, an average tentpole production, defined as a film with a budget above US$70 million, generated 2,840 tonnes of carbon-dioxide equivalent. Transport accounted for about 51 per cent, mains electricity and gas for 34 per cent, and diesel generators for 15 per cent.

The unit is per production. Repeating it as a daily figure would change its meaning by orders of magnitude. It is also an average assembled for the report, not a measurement that can be assigned to any named film merely because its budget crosses the threshold.

Used properly, the figure identifies leverage. Replacing plastic bottles while leaving travel and energy decisions untouched cannot address the largest categories in this model. Used carelessly, the same figure becomes a dramatic decoration: one large number cited before a list of small actions, with no route between them.

That is the first rule of environmental accounting. A footprint should change which decisions receive attention. If it only changes the language around an unchanged production, measurement has become scenery.

2. A calculator has a field of view

The albert calculator is more serious than a checklist of symbolic gestures. Its 2025 methodology includes filming and non-filming spaces, travel and accommodation, materials, disposal and post-production. It applies published emission factors to activities such as kilometres flown, hotel nights and hours in an edit suite. A draft footprint can be prepared in pre-production and compared with the completed record.

That structure matters because it converts “be sustainable” into quantities that a production can budget and revise. It also creates a field of view. Anything without a requested input, reliable activity record or suitable emission factor will be estimated, simplified or missed.

Albert’s 2025 ACCELERATE report is unusually candid about this. It describes gaps in travel, temporary power and waste, notes that collection is often retrospective, and repeats the last external audit’s error breakdown: 52 per cent estimates, 39 per cent incorrect inputs and 10 per cent aggregation. It also points to weak integration with budgets and to data entry being assigned to staff without sufficient experience or oversight.

A calculator’s output is not a camera pointed at an objective total. A footprint is assembled from activity data, emission factors, category rules and omissions. Two productions can improve while their reported totals become less comparable because one has collected more complete data.

The honest claim is therefore more specific than “this production was sustainable”. It says which activities were counted, which factors were used, where primary records replaced estimates and which material effects remain outside the inventory.

3. Greenwashing begins in the gap between a result and its publicity

Magali Delmas and Vanessa Burbano define firm-level greenwashing through the combination of poor environmental performance and positive environmental communication. The useful part of that definition is the comparison. A green statement is not misleading merely because it describes one small improvement. The problem begins when the statement invites the audience to infer a larger environmental performance that has not been demonstrated.

For film and television, familiar phrases leave essential fields blank. Lower-carbon compared with which production plan? Reduced by 40 per cent across which lifecycle stages? Carbon neutral after which reductions, purchased instruments and exclusions?

The same test protects worthwhile work from lazy cynicism. Reuse, rail travel, renewable power and better data collection can produce real reductions. Calling every partial measure greenwashing makes evidence irrelevant. The task is to distinguish a bounded improvement from an unbounded reputation claim.

A certification logo cannot carry that explanation by itself. The public needs the boundary that made the logo possible.

4. Virtual production changes the counterfactual

Virtual production is the clearest example of why a comparison needs a carefully built alternative. Studio Ulster’s 2023 report estimates that incorporating virtual-production techniques can reduce emissions by 20 to 50 per cent relative to traditional methods, especially when at least 30 per cent of a hybrid production is filmed in an LED volume. Travel, location fuel and physical set construction can all fall.

The headline is promising, and the report states limits that publicity can easily lose. Its case study concerned one roughly £80,000 short film made in a virtual studio. The film was not produced a second time on location. Researchers constructed traditional and hybrid scenarios with advice from an experienced location manager and assistant directors, then compared those counterfactual schedules with the recorded studio production.

That is a modelled comparison rather than a controlled pair of finished films. The report also says that existing production calculators do not yet capture the manufacture and disposal of LED panels, software and cloud services, rendering, data storage and data movement well. It recommends supplementing screen-production accounting with information-technology accounting.

The result survives those qualifications. If an LED stage genuinely replaces flights, road freight, generators and disposable construction, a large reduction is plausible. If it supplements a location schedule, encourages additional scenes or runs on a carbon-intensive grid, the balance changes. “Virtual” names a technique. It does not specify the counterfactual it replaced.

5. Distribution moves the boundary into the home

Digital distribution creates another boundary problem. A 2021 Carbon Trust white paper estimated the footprint of one hour of video-on-demand streaming in Europe at about 55 grams of CO2e. The viewing device was usually the largest component, while changing video resolution had only a small effect in the model.

The study was funded by Netflix and developed in consultation with DIMPACT, a collaboration between University of Bristol researchers and media companies. Funding does not make its result false. It is part of the evidence a reader needs when judging a study about an industry sponsor’s service.

The unit also matters. An estimate per viewing hour answers a different question from the annual footprint of a platform, the embodied emissions of a television or the effect of releasing more programmes to more subscribers. Efficiency per hour can improve while total hours rise. Conversely, assigning the whole electricity use of a household television to a production company would answer a boundary question in a different, contestable way. The claim has to keep the unit it was calculated to support.

6. Put carbon where the production can still change

A footprint calculated after wrap can describe a production and improve the next one. A carbon budget set during development can alter the production now. The distinction is the same one filmmakers already understand from money and time: a report delivered after the decision cannot govern the decision.

A credible production claim needs a defined lifecycle and organisational boundary. It should publish both the absolute footprint and any intensity measure, because tonnes per finished hour can fall while total tonnes rise. Alternatives need comparison before locations, travel and workflows are locked. Estimates, measured activity and avoided-emissions scenarios should remain separate, as should direct reductions and any compensating credits.

The GHG Protocol’s product standard follows raw materials, manufacture, transport, use and disposal across a lifecycle. Its guidance on avoided emissions likewise treats comparative claims as a separate exercise requiring an explicit baseline. Screen production needs that discipline because its most attractive technologies are often substitutions. Each one removes a visible activity and creates a less visible chain.

The useful question for an LED wall, electric generator or digital delivery system is what disappeared because it was used, what appeared elsewhere and whether the comparison includes both. Carbon accounting matters most before a production plan hardens, when an estimate can still change a location, schedule or workflow. After wrap, the same number mostly describes decisions already made.