Carbon footprint for an industrial manufacturer: energy and raw materials first

At three in the morning, with nobody on the floor, the furnace is still holding temperature. So is the compressed-air network, leaking through fittings that were fine last year, and the chiller keeping a process loop at setpoint for a line that will not restart until Monday.

An industrial carbon footprint is not an abstraction assembled from invoices. Most of it is physically on the site, running whether or not anyone is watching, or it arrived at goods-in as material. That changes both how you measure it and where the reductions are.

Keep the sites apart until the analysis is done

The instinct on a multi-site footprint is to consolidate early: one spreadsheet, one total. Resist it. Two plants making similar products can differ by a factor of two in energy per unit because one runs an older furnace or a leakier compressed-air network, and that difference is not noise, it is the finding. Consolidate at the end for reporting. Model per site, and wherever the metering allows, per line or per major machine.

Site-level scoping also settles the boundary questions industrial groups actually have: the co-owned workshop, the tenant activity in your building, the plant abroad. Follow operational control, and write the choice down where next year's team will find it.

The meters already know

An industrial scope 1 holds more than most checklists suggest: fuel for process heat, gas burners on treatment lines, furnaces, on-site vehicles and forklifts, standby generators, and refrigerant leaks from process cooling as well as air conditioning. Your maintenance contractor's recharge reports are the leak record, and on cold-heavy processes that line surprises people every time.

Electricity is scope 2, and for once collection is pleasant, because it is already metered. The work is allocation. "The plant used 4 GWh" suggests nothing; "the compressed-air system takes a fifth of it" is an action item with a name on it. Sub-metering pays for itself in analysis before it ever pays in savings.

Good to know: if a French site is a classified installation (ICPE) or falls under energy audit obligations, part of this data already exists in structured form. The energy audit, the ICPE file and the carbon footprint read the same meters. Run the collections together and each exercise gets cheaper.

The materials line decides the total

For most manufacturers, purchased raw materials and components are the largest single line, ahead of freight and far ahead of travel. Steel, aluminum, castings, plastics, electronics: each arrives carrying embodied emissions that dwarf what your machines add to it.

Collect them in physical units. Purchasing records give tonnages by material family, and physical factors for standard materials are well covered in the public databases. A monetary estimate, which is the honest fallback for diverse services, is a poor fit here: metal prices swing for reasons unrelated to carbon, and the entire point of measuring this line is to see the effect of buying recycled content or a lower-carbon grade. Euros cannot show you that.

Freight in and out comes next, in tonne-kilometers by mode from your logistics data or your forwarder. Then the long tail, services, IT, maintenance contracts, where monetary factors on the accounting export are both honest and sufficient.

An intensity the plant can steer by

Report the absolute total, because that is what reporting wants, but steer by intensity, and choose a denominator the plant actually controls: kilograms of CO2-equivalent per unit produced, per tonne of output, per machine-hour.

Per-euro intensities move with prices and product mix, which means the number can improve in a year when nothing physical changed, and can worsen in a year of good work. A per-unit intensity computed per site is comparable across years and across plants, and it turns the footprint into something a plant manager owns alongside scrap rate and energy cost.

Where the reductions actually are

The footprint will point at levers your operations people already suspected: heat recovery on furnaces and compressors, compressed-air leak campaigns, motor and drive upgrades, load scheduling. On the materials side, recycled content, supplier choice, and design that removes material from the part.

So start with the meters, keep the sites apart, and weigh the materials. An industrial carbon plan turns out to be an energy-efficiency and purchasing plan, run by people who already work here, measured by a number they can watch move.

FAQ

What are the main emission sources for an industrial company?

Purchased raw materials and components usually form the largest line, carried in as embodied emissions. Then come process energy (fuel and heat in scope 1, electricity in scope 2), freight, and refrigerant leaks. Travel and offices, dominant for service companies, are minor lines in an industrial footprint.

Should an industrial carbon footprint be measured per site?

Yes. Model each site separately and consolidate only for reporting. Differences between comparable sites are the most direct pointer at reduction levers, and site-level data aligns with the meters, the energy audits, and the ICPE files that already exist.

Which intensity metric should a manufacturer use?

Emissions per unit produced or per tonne of output, computed per site. Per-euro intensities move with prices and product mix, so they can improve while physical reality is unchanged. A per-unit figure is comparable across years and lets plant managers steer it like any other operational indicator.