Two catalogues, the same rating, one machine at 9.6 tonnes and one at 21.5. Steel mass decides the carbon number, and the building the crane sizes decides the rest. Here is the arithmetic, the CBAM line between 8426 and 7308, and the ten lines that make two offers comparable.
Two catalogues, both for a 10 tonne double girder at 22 metre span. One lists the crane at 9,585 kg. The other lists 21,500 kg. Same capacity, same span, same duty description on the front page. That is a factor of 2.2 in steel, and steel is where most of the carbon sits.
So when a buyer asks us for the carbon figure on a double girder, the first honest answer is that the number is not fixed by the tonnage. It is fixed by the design class, the mill that rolled the plate, and the building the crane is about to be installed in. The second honest answer is that the crane itself is only about a quarter of the total. The runway beams, the columns and the rail carry the rest, and the crane decides how heavy all of them have to be.
This piece works through a 20 tonne, 22.5 metre, A6 double girder. All the arithmetic is ours, done the way we do it for a customer enquiry, and it is labelled as our planning figures rather than a published statistic. The external numbers, the steel intensities, the section weights, the CBAM code list, come with sources.
Around 43 tonnes of CO2e from cradle to the factory gate, for a machine that weighs 16.5 tonnes. That is a little over 2 tonnes of CO2e per tonne of lifting capacity. A 10 tonne single girder comes out near 1.3 to 1.5. The double girder is not twice as dirty, but it is 50 to 60 per cent dirtier per tonne of capacity, and the reason is unglamorous: there is simply more plate.
Table 1 builds the number up by component. It assumes blast furnace plate at the global average intensity, which is the default for Chinese supply unless somebody pays for something else.
| Component | Mass | Factor | CO2e | Why the factor differs |
|---|---|---|---|---|
| Two box girders | 9,450 kg | 2.34 kg/kg | 22.1 t | Plain plate, blast furnace route, worldsteel 2025 average |
| End carriages and wheels | 1,850 kg | 2.5 kg/kg | 4.6 t | Castings and forged wheels carry more processing energy than plate |
| Trolley frame and 20 t hoist | 3,300 kg | 2.9 kg/kg | 9.6 t | Gearboxes, motor windings, wire rope, bearings |
| Walkway, handrail, festoon, cable | 1,050 kg | 2.5 kg/kg | 2.6 t | Light sections plus galvanising and PVC |
| Cab and electrical panels | 850 kg | 4.0 kg/kg | 3.4 t | Copper, electronics, glass, plastics |
| Paint and consumables | 250 kg | 2.5 kg/kg | 0.6 t | Epoxy system, zinc primer, thinner |
| Total | 16,500 kg | — | 43 t | 2.15 t CO2e per tonne of rated capacity |
Two things in that table surprise people. The hoist and trolley are 20 per cent of the mass but 22 per cent of the carbon, because machined and wound components are more energy intensive per kilo than plate. And the paint is under 2 per cent, which is worth saying out loud because paint is the item most often raised in a sustainability meeting and the least worth arguing about.
Not because of the second beam. Because the load path changes. On a single girder the hoist trolley runs on the bottom flange, so the beam works as a combined bending and torsion member and the manufacturer keeps it shallow. On a double girder the trolley runs on a rail on top of two girders, the load is shared between them, and each girder is sized for bending plus the local wheel load of the trolley. The pair ends up heavier than one beam, and the extra weight is not proportional to capacity.
Table 2 is the comparison most often quoted in the trade, and our own measurements land close to it.
| Capacity and span | Single girder bridge | Double girder bridge | Ratio | Comment |
|---|---|---|---|---|
| 5 t × 15 m | 1,200 kg | 2,100 kg | 1.75× | Double girder rarely justified below 5 t |
| 10 t × 20 m | 2,800 kg | 4,600 kg | 1.64× | The crossover point; both designs are viable |
| 20 t × 25 m | 6,500 kg | 9,800 kg | 1.51× | Above 15 t the single girder mostly stops being offered |
| 50 t × 30 m | Not viable | 22,000 kg | — | 22 t of plate is 51 t of CO2e before anything is bolted on |
The ratio falls as capacity rises, which is counter-intuitive until you remember that the end carriages and the trolley are a bigger share of a small crane. On a 5 tonne machine the second girder nearly doubles the bridge. On a 50 tonne machine it adds half.
Now compare those two catalogues again. A 10 tonne double girder at 22 metres is listed at 9,585 kg by one Italian builder and 21,500 kg by one Indian builder. Both are real machines, both will pass a load test, and the difference is design philosophy: how much deflection is allowed, how conservative the fatigue allowance is, whether a cab and a full walkway are included, and what steel grade the girders are rolled from. If you are comparing carbon numbers from two suppliers and one is half the other, this is usually the reason, and no amount of green steel will close a 2.2× gap in mass.
Yes, on the same drawing it does. The World Steel Association's 2025 sustainability report puts the global average at 1.92 tonnes of CO2 per tonne of crude steel, with a blast furnace route at 2.34 and a scrap based electric route at 0.69. Widening the boundary to include methane, nitrous oxide and upstream mining moves those to 2.18, 2.66 and 0.71. Roughly a 3.4 to 1 spread on the same mass of steel.
| Steel route | CO2 per tonne of crude steel | Widened boundary | Energy | Where you meet it |
|---|---|---|---|---|
| Blast furnace, basic oxygen | 2.34 t | 2.66 t | 23.88 GJ/t | Most Chinese plate and section |
| Direct reduced iron, electric | 1.47 t | 1.66 t | 23.30 GJ/t | Middle East, India, some European mills |
| Scrap, electric arc | 0.69 t | 0.71 t | 9.84 GJ/t | Turkey, Italy, Spain, US long products |
| Global average | 1.92 t | 2.18 t | 20.95 GJ/t | The safe default when nobody declares |
Apply that to the 12 tonnes of fabricated steel in our 20 tonne crane, girders plus end carriages plus walkway, and the plate alone swings between 8.8 and 28 tonnes of CO2e. The whole machine swings from about 18 to 43 tonnes. That is a 2.4× spread on identical drawings, identical dimensions and identical capacity.
Which raises the awkward question every buyer eventually asks. If it is the same crane, why would anyone pay a premium for the lower figure? Because on a European grid that 25 tonne difference is worth about 25 certificates at the current price. CBAM does not apply to cranes, so nobody is actually sending that invoice. But the number still appears in the buyer's Scope 3 inventory, and a procurement team with a 2030 target is starting to treat it as a real cost.
The building. This is the part of the conversation that gets skipped, and it is the part with the most steel in it. A double girder at 20 tonnes produces wheel loads in the 170 to 190 kN range, and those loads size the runway beam, the columns and the rail. None of that steel appears on the crane quotation.
Table 3 gives typical runway beam sections against crane capacity. These come from a published AISC and AS 4100 design reference and they match what we see specified on real projects.
| Crane capacity | Runway span | Typical section | Mass | CO2e per 200 m of runway |
|---|---|---|---|---|
| 5 t | 6 to 9 m | 310UB32 | 32 kg/m | 15 t |
| 10 t | 9 to 12 m | 360UB44.7 | 44.7 kg/m | 20.9 t |
| 15 t | 12 to 18 m | 410UB53.6 | 53.6 kg/m | 25.1 t |
| 20 t | 18 to 24 m | 460UB67.1 | 67.1 kg/m | 31.4 t |
| 30 t | 24 to 30 m | 530UB82 | 82 kg/m | 38.4 t |
| 50 t and above | 30 m and above | Built-up plate girder | 120 kg/m and up | 56 t and up |
Now put a whole bay together. A 100 metre long building with a runway on each side, 26 columns, a QU70 rail. Our own planning figures again, and the assumptions are in the table so you can substitute yours.
| Element | Quantity | Steel | CO2e | Share |
|---|---|---|---|---|
| Double girder crane, complete | 1 machine | 16,500 kg | 43 t | 24% |
| Runway beams, 460UB67.1 | 2 × 100 m | 13,420 kg | 31 t | 17% |
| Columns and bracing | 26 columns, 12 m | 34,000 kg | 80 t | 45% |
| Crane rail and clips, QU70 | 200 m | 10,560 kg | 25 t | 14% |
| Bay total | — | 74,480 kg | 179 t | Crane is 24% |
Part of that column steel would be built anyway. A factory needs columns. But the crane decides how heavy they are, and it decides whether the runway beam is a 460 section or a 530. So the honest framing is not that the crane owns 179 tonnes of carbon. It is that the crane sizes about 58 tonnes of steel it never touches, and that steel is invisible on every quotation in the market.
There is a real project that shows the scale. A steel mill in the southern United States, described in a published case study, frames the plant with more than 36,000 tonnes of structural steel and runs over six miles of crane runway girder and rail for roughly 50 cranes. Six miles of runway at 100 kg per metre is about 966 tonnes of steel in the runways alone. No CBAM question there, since it is a domestic mill, but it shows how much structure a crane population drags into a building.
It does, and this is the one lever that pays twice. European style double girders are commonly quoted at 15 to 30 per cent lower mass than a general purpose double beam design, with maximum wheel pressure 10 to 35 per cent lower. Lower wheel pressure means a smaller runway section, and a smaller runway section means lighter columns.
Run the bay again with those reductions. Table 5 uses 15 per cent and 30 per cent mass reduction, paired with 15 per cent and 35 per cent wheel pressure reduction, and steps the runway down one and two sections respectively.
| Element | As drawn, 16.5 t | 15% lighter | 30% lighter |
|---|---|---|---|
| Crane CO2e | 43 t | 36.6 t | 30.1 t |
| Runway section and CO2e | 460UB67.1, 31 t | 410UB53.6, 25.1 t | 360UB44.7, 20.9 t |
| Columns CO2e | 80 t | 71.6 t | 63.7 t |
| Rail and clips CO2e | 25 t | 25 t | 25 t |
| Bay total | 179 t | 158 t | 139 t |
| Change against baseline | — | −21 t, −12% | −40 t, −22% |
A 30 per cent lighter crane takes 40 tonnes of CO2e out of the project, and only 13 of those tonnes come off the crane. The other 27 come off the building. That is why the mass per tonne on a datasheet is worth more attention than the motor efficiency class, and why a buyer who only compares crane prices is measuring a third of the cost.
Small, but not zero, and buyers ask. Table 6 puts the leftovers next to the machine so the proportions are visible.
| Item | Basis | CO2e | Share of gate-to-gate total |
|---|---|---|---|
| Sea freight, Shanghai to Rotterdam | 20,000 km at 10 to 16 g per tonne-km, 16.5 t | 3.3 to 5.3 t | 7 to 11% |
| A 30 tonne double girder, same route | Same distance, 30 t shipped | 6.0 to 9.6 t | Breakbulk rather than container, so add handling |
| Coating system | 250 kg of epoxy and primer | 0.6 t | 1.4% |
| Hoist and trolley | 3,300 kg of machinery | 9.6 t | 22% |
One reference point for the hoist. A published environmental product declaration for a 5 tonne low headroom rope hoist, 277 kg of hardware at 6.5 m lift, reports a total life cycle climate impact of 2,159 kg CO2e over 20 years of typical use, with raw material processing as the largest single phase. That is a real declaration from a major European manufacturer, covering the whole life, not just the factory gate. It also shows how few of these exist: as of this month there is no published EPD for a standard 5 to 100 tonne double girder overhead crane.
No. Cranes are HS 8426, Chapter 84. Annex I of Regulation (EU) 2023/956 lists cement, electricity, fertilisers, iron and steel, aluminium and hydrogen, and the iron and steel entry covers all of Chapter 72 plus a set of Chapter 73 headings: 7301, 7302, 7303, 7304, 7305, 7306, 7307, 7308, 7309, 7310, 7311, 7318 and 7326. Chapter 84 is not there. A crane imported into the EU does not trigger a CBAM declaration, and the carbon paperwork is a commercial matter between you and your customer, not a customs condition.
The project is a different question, because the steel around the crane probably is in scope.
| Item on the project | CN heading | In CBAM Annex I | Practical effect |
|---|---|---|---|
| The crane | 8426 | No | No declaration, no certificates |
| Runway beams, columns, brackets | 7308 | Yes | Counts towards your 50 tonne threshold |
| Crane rail | 7302 | Yes | Depends how the line is declared; ask the forwarder |
| Rail clips, anchor bolts | 7318 | Yes | Small mass, same paperwork |
| Plate and sections bought loose | 72 | Yes | Excludes 7204 scrap and the listed ferro-alloys |
The threshold is 50 tonnes of CBAM goods per importer per calendar year, aggregated across all the CN codes in Annex I. Since 1 January 2026 an importer above that line has to be an authorised CBAM declarant. The first declaration and certificate surrender falls due on 30 September 2027, covering the 2026 import year, and any carbon price already paid in the country of origin can be deducted.
Our bay above, with 58 tonnes of runway, column and rail steel, sits just over the line on its own. A single small crane in an existing building with 10 tonnes of runway steel sits under it and nobody files anything. That is the practical dividing line, and it is worth calculating before the purchase order rather than after.
Certificate prices for 2026, published quarterly by the Commission: €75.36 for the first quarter, €75.28 for the second, €82.32 for the third. That is roughly USD 95 per tonne of CO2 at early October rates. From 2027 the price is set weekly, which means a project that slips six months can find its allowance cost has moved.
Because customs has seen that trick before, and the rule is written to stop it. General Interpretative Rule 2(a) of the Harmonized System says that a reference in a heading to an article includes that article presented unassembled or disassembled. The World Customs Organization text is explicit, and the Court of Justice confirmed the principle again in Case C-107/22 in April 2023: components presented together, needing only assembly by bolts, rivets or welding to become the finished article, are classified as the finished article.
So a double girder arriving in six pieces, two girders, two end carriages, a trolley and a cab, is still a crane at 8426. Bolting and welding count as assembly, not as further working. Splitting the shipment changes the freight bill and the port handling, not the customs classification, and declaring welded girders as 7308 structures when they are part of a crane supplied under one contract is a misdeclaration with a duty consequence.
Where the split does happen is legitimately: when the buyer sources the runway, the columns and the rail separately, from a steel fabricator, on their own contract. That steel is 7308 and 7302. It is CBAM goods. It is also, as Table 4 showed, the majority of the project's carbon. So the answer to the CBAM question on a crane project is almost always the same: not the crane, look at the building.
Yes, by a wide margin, once the crane is running at A6 or above. This is the number nobody asks for, and it dwarfs everything in the previous sections.
| Duty class | Equivalent full load hours per year | kWh per year | 20 year CO2e, EU grid 206 g | 20 year CO2e, China grid 537 g |
|---|---|---|---|---|
| A5 | 1,600 | 37,800 | 156 t | 405 t |
| A6 | 2,000 | 47,200 | 194 t | 506 t |
| A7 | 2,600 | 61,400 | 253 t | 658 t |
| Embodied, for comparison | — | — | 43 t | 43 t |
At A6 on a European grid the electricity is 4.5 times the steel. On a Chinese or Indian grid it is closer to 12 times. The grid factors are published: the European Environment Agency puts EU electricity generation at 206 g CO2e per kWh for 2024, down 11 per cent in a year, and China's Ministry of Ecology and Environment reports a national average of 0.5366 kg per kWh for 2022.
Which creates an odd situation. Buyers ask for the embodied figure because it is the new thing on the questionnaire, and it is the easiest to put in a table. The number that will actually decide whether the plant hits its 2030 target is the one on the electricity bill, and it is controlled by duty class, cycle times and the local grid. Not by which mill rolled the plate.
It is the biggest single lever on a double girder, and it gets specified almost casually. Duty class drives plate thickness, motor sizing, brake rating and the maintenance interval, and a step from A5 to A7 is a step up in mass of roughly 20 to 30 per cent on the same capacity and span. The extra mass then compounds into the building.
| Specification | Crane mass | Embodied | 20 year electricity, EU | Combined |
|---|---|---|---|---|
| 20 t, A5, honest light duty | 13,500 kg | 35 t | 156 t | 191 t |
| 20 t, A6 specified by default | 16,500 kg | 43 t | 194 t | 237 t |
| 20 t, A7 for a two shift plant | 18,900 kg | 49 t | 253 t | 302 t |
The gap between a correctly sized A5 and a default A7 is 111 tonnes of CO2e, and 58 of that is electricity the plant never needed because the crane was never going to run two shifts. Overspecifying duty class is the most expensive habit in this business, in money and in carbon, and it is invisible because nobody puts "A7 because we always do" on a specification sheet.
We have written about the sizing logic elsewhere, and it is worth a look before the enquiry goes out rather than after the drawing is approved. The double girder crane range we build covers 5 to 100 tonnes and we will size to A5 where A5 is honest, because a heavier machine is not a safer machine once the duty is overstated.
More than half of the original burden, on paper, and you cannot net it. That is the awkward part of module D in EN 15804.
| Item | Basis | Value | Comment |
|---|---|---|---|
| Recovered steel at end of life | 16,500 kg at 85 to 98% recovery | 14,000 to 16,200 kg | Structural steel recovery is well established |
| Avoided primary production | 2.34 minus 0.69 = 1.65 t CO2 per tonne | 23 to 27 t CO2e | Reported in module D, outside the system boundary |
| Dismantling and transport | Module C1 to C4 | 1 to 3 t CO2e | Cutting, crane hire, road transport to the yard |
| Reuse instead of recycling | Girder reused as a runway beam elsewhere | Up to 43 t avoided | Needs a design that allows it; rare in practice |
Here is the catch. Module D is mandatory in EN 15804, but it sits outside the product system boundary by definition. Most procurement rules will not let you subtract it from the A1 to A3 figure you report, and the ones that do require you to show the calculation both ways. So when a supplier advertises a "carbon neutral crane" on the strength of a recycling credit, they are netting a future benefit against a present emission, which is not the same as reducing anything.
The credit is real. It is just not yours to spend at the procurement stage. What you can do is specify a machine whose girders can be unbolted rather than cut up, because that is what turns 27 tonnes of credit into 43 tonnes of avoided production.
Ten lines. None of them require a consultant, and together they make two offers comparable without anyone having to trust a marketing number.
| Line | What to ask for | Why it changes the answer |
|---|---|---|
| 1 | Total machine mass, broken down by component | Mass is 90% of the embodied carbon and the one number suppliers omit |
| 2 | Steel mill and route for plate and sections | Blast furnace versus electric is a 3.4 to 1 spread |
| 3 | Mill test certificates with heat numbers | The only way to verify line 2 after the fact |
| 4 | Maximum wheel load per corner, at rated load | Sizes the runway, which is the bigger half of the project carbon |
| 5 | Duty class with the assumed cycles per hour and hours per day | A7 by habit costs 111 t CO2e over 20 years |
| 6 | Installed motor power by mechanism | Converts to the electricity number in Table 8 |
| 7 | Whether an EPD exists, and under which programme | If none exists, say so; an estimate labelled as an estimate is fine |
| 8 | Shipping mass and packing, per unit | Freight is 7 to 11% of the gate-to-gate total |
| 9 | Coating system and dry film thickness | Small carbon, large maintenance consequence |
| 10 | Bolted versus welded girder connections | Decides whether the crane can be reused rather than scrapped |
If a supplier cannot answer line 1, the rest of their carbon claim is decoration. If they can answer lines 1 through 4, you have enough to build a defensible number in an afternoon, and enough to notice when the next quotation quietly doubles the mass.
The market data makes the timing awkward for anyone waiting for clarity. Hot rolled band in China was USD 422 per tonne ex-works in mid-October, in Western Europe USD 835, in the United States USD 1,310, according to the SteelBenchmarker report of 14 October 2026. European structural sections were quoted around €750 to €820 per tonne ex-works earlier in the year. When plate prices move by a third between regions, a design that uses 30 per cent less of it stops being an environmental gesture and starts being a purchasing decision.
About 43 tonnes from cradle to the factory gate for a 16.5 tonne machine on blast furnace steel, and about 18 tonnes if the same drawings are rolled from scrap based electric steel. Sea freight to Europe adds 3 to 5 tonnes. Per tonne of lifting capacity that is roughly 2.2 tonnes of CO2e, against 1.3 to 1.5 for a single girder of the same rating. The mass of the machine moves the number more than any component choice inside it.
No. CBAM covers the goods in Annex I of Regulation (EU) 2023/956, which is Chapter 72 iron and steel plus headings 7301 to 7311, 7318 and 7326, and the aluminium, cement, fertiliser, electricity and hydrogen sectors. Cranes are HS 8426 in Chapter 84 and are outside. The runway beams, columns and rail bought for the building are 7308, 7302 and 7318, and those are in scope. The importer obligation starts above 50 tonnes of CBAM goods per calendar year, with the first declaration due 30 September 2027.
No. General Interpretative Rule 2(a) classifies an article presented unassembled or disassembled as the complete article, provided only assembly is needed, and the Court of Justice applied that principle again in Case C-107/22 in April 2023. A crane in six pieces is still a crane at 8426. Declaring fabricated girders as 7308 structures under a single crane contract is a misdeclaration. Buying the runway and columns separately from a fabricator is a different, legitimate transaction, and those are CBAM goods.
Design class, not capacity. One catalogue is a light European box girder with a compact trolley and no cab. The other is a heavier general purpose machine with a full walkway, a cab and more conservative deflection limits. Both will pass a load test. The difference is 2.2 times the steel, 2.2 times the embodied carbon, and roughly double the wheel load, which then sizes the runway. Always compare mass per tonne of capacity, not price per tonne of capacity.
At A6 and above, yes, by four to five times on a European grid and by more than ten times on a Chinese or Indian one. A 20 tonne A6 machine draws about 47,000 kWh a year, which is 194 tonnes of CO2e over 20 years at the EU average of 206 g per kWh, against 43 tonnes of embodied carbon. The only way to move that number is duty class, cycle time and the local grid. Buying greener steel is a rounding error next to specifying the right class.
Machine mass by component, steel mill and route, mill certificates with heat numbers, and maximum wheel load. With those four answers any competent engineer can reproduce the calculation in an afternoon, and the result will survive an auditor's question about method. Our double girder cranes are quoted with the deadweight breakdown as standard because it is the number every other calculation depends on. If a supplier will not provide it, treat their carbon figure as marketing.
Related articles:
Double Girder Crane Energy Consumption: Where the Kilowatt Hours Go, 5 to 100 Ton
Crane Carbon Footprint: What a Single Girder Crane Emits Before It Lifts Anything
Double Girder Crane Maintenance and Overhaul: The 20 Year Schedule
Double Girder Crane Installation and Commissioning Guide
Shipping an Overhead Crane: Export Logistics, Incoterms and Landed Cost
Double Girder Overhead Cranes: 5 to 100 Tonne, 10.5 to 31.5 m Span