A record 226.1 million cgt on order, and the cranes that decide whether those ships get built are the ones nobody photographs.
On 2 November, Clarksons Research put the world newbuilding orderbook at 226.1 million compensated gross tonnes. That is past the September 2008 record of 225.5 million, worth roughly USD 700 billion, and up 24 percent in twelve months. The photographs that go with that number are always the same: a 1,800 tonne goliath on rails, a hull block the size of a building hanging under it.
The cranes that will decide whether those ships actually get delivered are not in the photographs. They are the 20 to 50 tonne overhead machines inside the panel line, the sub-assembly hall and the pre-outfitting shop. Those run A6 to A7 duty, lift every three to six minutes, and they set how many blocks arrive at the dock each day. A goliath can only erect what the halls have already built.
This matters to anyone buying an overhead crane for a fabrication hall, and to anyone being told their shipyard order will be delivered on schedule. China holds 68 percent of this orderbook and has to lift its annual output from 22.7 million cgt delivered in 2025 to 44.0 million cgt scheduled in 2028. That is very close to a doubling of the work that moves through covered halls on overhead cranes, not a doubling of drydock capacity.
Clarksons and BIMCO measure at different dates and get slightly different totals. Both point the same way. Clarksons has October at a record 226.1 million cgt. BIMCO had 203.2 million on 1 September, up 13 percent from 180.0 million a year earlier. The gap between the two is date and coverage, not disagreement about direction.
| Measure | Oct 2026 | Previous peak, Sep 2008 |
|---|---|---|
| Orderbook | 226.1m cgt | 225.5m cgt |
| Value | about USD 700bn | not comparable in USD |
| Orderbook as share of fleet | 25% by gross tonnage | 53% |
| Orderbook in deadweight | 601m dwt | 657m dwt |
| Containerships | 31% of orderbook cgt | about half that share |
| Gas carriers | 15% | 5% |
| Bulkers | 17% | 31% |
| Alternative-fuel capable | 43% of orderbook cgt | 2% |
| China share of orderbook | 68% | about 34% |
| Korea and Japan share | 17% and about 8% | larger, on a much smaller book |
Read the share-of-fleet line twice. Twenty-five percent against fifty-three percent is why this record is not 2008 with different dates. Yards are not carrying a bubble of speculative tonnage relative to a fleet that has barely grown. They are carrying four years of committed work for owners who have already signed financing, which is a very different reason to add crane capacity.
The mix shifted too, and the mix is what changes crane duty. Containerships are now 31 percent of the book and gas carriers 15 percent, against 5 percent eighteen years ago. Alternative-fuel capable ships are 43 percent of the backlog against 2 percent at the last peak. Gas carriers and dual-fuel tonnage are not built like bulkers. More outfitting work, more pipe, more containment structure, more hours spent inside a covered hall under an overhead crane instead of at a quayside.
Because they are extraordinary machines, and because they are easy to photograph. The last eighteen months produced a run of orders that would have been unthinkable in 2019.
| Yard and equipment | Specification | What it changes |
|---|---|---|
| Hanwha Ocean, Okpo, goliath ordered Jan 2026, steel cut 4 Oct 2026 | 1,800 t capacity, 205 m span, fabrication 9–12 months, typical contract USD 40–60m | Doubles the 900 t hook capacity at the drydocks and allows in-air turnover of blocks up to 1,000 t |
| Hengli Heavy, Changxing Island super drydock | 2 × 2,000 t, 256 m span, 12,700 t deadweight each, single lift above 10,500 t, lift height 100 m, plus one 1,000 t unit | Yard claims more than 50% gain in drydock construction efficiency |
| GENMA order, July 2026 | One 2,000 t and one 1,200 t shipbuilding gantry, eight single jib portal cranes, one plate unloader | The portal cranes, not the gantries, do the daily plate and outfit work |
| Yangzi Hongyuan, delivered Aug 2026 | 1,800 t and 1,200 t goliaths | Became the engineering baseline for the Korean unit now under construction |
| Samsung Heavy, Geoje | 2 × 1,200 t goliaths | Tandem lifts for super-mega blocks |
A 1,800 tonne goliath is a nine to twelve month fabrication job on its own, and at USD 40 to 60 million it costs more than the entire overhead crane fleet in a mid-size fabrication hall. It is not the machine a shipyard buys when it needs to lift 30 tonne panels onto a jig.
There is a second reason the goliaths dominate the coverage. They are the last step in a chain, so any gain in dock throughput gets attributed to them. Hengli's own framing is that assembling ultra-large modules as whole units raises drydock efficiency by more than 50 percent. That is true. It is also only achievable if the halls upstream can produce a 10,000 tonne module in the first place, which is a question about many smaller cranes working in parallel, not one very large one.
Count lifts rather than tonnes. A goliath at a super drydock might complete four to ten lifts in a shift. It is moving hull sections that took weeks to build. An overhead crane in a panel line completes a lift every three to six minutes and does it for two or three shifts a day.
| Stage | Typical capacity | Lifts per shift | What it sets |
|---|---|---|---|
| Plate yard and cutting shop | 10–30 t with magnet or vacuum spreader | 60–100 | Plate feed rate into the line |
| Panel line | 20–50 t, 20–35 m span, A6 | 25–40 panel moves | How many panels a day become sub-blocks |
| Sub-assembly hall | 32–50 t, A6–A7 | 10–18 sub-blocks | Block completion rate |
| Block assembly bay | 50–100 t, 25–40 m span, A7, often tandem | 4–8 blocks | Maximum block weight that can travel to the dock |
| Super drydock goliath | 1,800–10,000 t | 4–10 | How fast the hull leaves the dock |
The arithmetic in that table is our own planning estimate, not a published statistic. The structure of it is not controversial though, and any yard engineer will recognise it. One extra lift at the dock needs five to ten more blocks delivered, and each of those blocks needs dozens of upstream crane moves. Add dock lifting capacity without adding hall capacity and the goliath spends its shift waiting.
That is why the hall crane is where a capacity expansion actually starts, and why it is usually ordered late anyway. A new goliath makes a press release. Six 50 tonne A7 overhead cranes in a new sub-assembly bay make a procurement line item.
This is the part of the story that turns a market statistic into a procurement schedule. Chinese yards are not being asked to add a few percent. They are being asked to roughly double within three years.
| Figure | Value | Source and date |
|---|---|---|
| Global deliveries, 2027 | 64m cgt, up 24% on 2026 and 18% above the 2010 record of 55m | Clarksons Research, Nov 2026 |
| Global deliveries, 2028 | 72m cgt, 60% above 2025 | Clarksons Research, Nov 2026 |
| China, delivered 2025 | 22.7m cgt | BIMCO analysis, Oct 2026 |
| China, scheduled 2028 | 44.0m cgt, nearly double 2025 | BIMCO analysis, Oct 2026 |
| Order coverage on current capacity | China 4.2 years, Korea 1.9 years, Japan 1.7 years | BIMCO, highest output 2021–2025 as capacity |
| Order coverage if capacity means peak 2021–2030 output | China 2.1 years, Korea and Japan 1.6 years | BIMCO, Oct 2026 |
| Hengli alone, deliveries | 99 ships in 2027 and 199 in 2028, against 62 completed in the first nine months of 2026 | Banchero Costa, Nov 2026 |
| Hengli expansion budget | CNY 13.5bn, about USD 1.94bn, with about CNY 10bn for a high-end vessel plant plus a curved hull block workshop and a pre-outfitting workshop | Caixin, Jan 2026 and Splash247, Apr 2026 |
Those last two rows are the ones to sit with. A curved hull block workshop and a pre-outfitting workshop are buildings full of overhead cranes. They are not drydocks. The money being committed in 2026 is going into exactly the covered halls where the tonnage gets assembled, because that is where the bottleneck is now.
The counter-argument is worth stating plainly, because BIMCO made it: capacity added for a 2028 delivery peak can sit underused if contracting slows after 2027. That is a real risk for a yard. For a crane buyer in a yard it changes the financing, not the engineering. It argues for buying cranes sized to a hall you can keep busy, not for buying the biggest machine the orderbook will justify this quarter.
Shipyard duty is not factory duty with a coat of paint. The published ranges from marine crane suppliers are fairly consistent, and they are much wider than what a general fabrication shop buys.
| Application | Capacity | Span | Duty |
|---|---|---|---|
| Light workshop and component handling | 5–20 t | 10–25 m | A5, FEM 2m |
| Panel line assembly | 20–50 t | 20–35 m | A6, FEM 3m |
| Block assembly and engine installation | 50–100 t | 25–40 m | A7, FEM 4m |
| Heavy lift and drydock service | 100–200 t and above | 30–45 m | A7–A8, FEM 4m–5m |
| Graving dock crane, dual hoist | 50–150 t main, lighter auxiliary | 30–50 m | A7, with dual hoist for engines and propellers |
| Plate yard with magnet or vacuum spreader | 10–30 t, plates up to 40 m long | 20–40 m | A6–A8 depending on shift pattern |
| Outfitting and commissioning | 10–32 t | 15–35 m | A5–A6, slow creeping speed for placement |
Two things in that table catch buyers out. The first is span. A panel line hall 35 metres wide needs a crane that covers it, and span drives bridge weight, wheel load and runway cost faster than capacity does. We quote 32 tonne machines on 22 metre spans and 32 tonne machines on 34 metre spans, and they are not close in price or in what the building has to carry.
The second is the speed requirement. Outfitting lifts are light and slow. Engine installation is heavy and very slow. A crane bought for tonnage alone will be geared for a speed nobody needs and will still lack the creeping speed the fitters actually use. If you are specifying for a hall that does both, say so in the RFQ and expect a two-speed hoist or a VFD.
Duty class is the single specification that a shipyard most often gets wrong, usually by buying a factory crane because it was cheaper and looked adequate on paper.
| Class | FEM equivalent | Operating hours a year | Where it belongs in a yard |
|---|---|---|---|
| A5 | FEM 1Am–2m | under 200 | Maintenance shops and light assembly only |
| A6 | FEM 2m | up to about 1,000 | Plate handling and outfitting cranes, single shift |
| A7 | FEM 3m | 1,000–2,000 | Graving dock and block assembly, two to three shifts |
| A8 | FEM 4m | over 2,000, frequent near-rated lifts | Continuous steel handling and dedicated plate yards |
The operating-hours column comes from published marine crane selection guidance, and it maps to a rule most yards learn the hard way. A hall that runs two shifts is an A7 hall, whether or not the RFQ says so. The crane will be asked to do two-shift work and will be maintained on a single-shift assumption. Then the gearbox oil analysis at year five reads like a machine twice its age.
The price difference between adjacent duty classes is real but not enormous, typically somewhere in the mid-teens percent at the same capacity. The cost of re-rating or replacing a crane after three years of heavier service is not in the mid-teens. If you genuinely do not know the shift pattern, specify A7 and let the supplier talk you down with an argument, not the other way round.
A shipyard is a corrosive environment with a lifting problem attached. If the building is within a few hundred metres of salt water, the paint system on a standard industrial crane is under-specified before it is delivered.
The published marine specification is not exotic. Structures get extra surface preparation and a high-build epoxy system rated C4 to C5 under ISO 12944. Electrical enclosures are IP55 or better. Fasteners, cable trays and handrails move to 304 or 316 stainless to avoid galvanic corrosion. None of this is optional at a coastal yard, and all of it is cheaper at the factory than as a retrofit.
| Upgrade over a standard industrial crane | Cost effect on crane price | What it buys |
|---|---|---|
| ISO 12944 C4 paint system | 8–15% | Coastal industrial atmosphere, roughly 5–15 km from the water |
| ISO 12944 C5-M paint system | 15–30% | Direct marine atmosphere on a quayside or dock-side bay |
| Stainless fasteners, trays and handrails | 5–10% | Removes the galvanic couple that eats zinc-plated bolts first |
| IP66 instead of IP55 enclosures | 8–12% | Driven rain and washdown, not just dust |
| Marine-grade festoon cable jacket | 3–6% | Chloroprene or equivalent instead of PVC, which hardens and cracks |
Those percentages are our planning figures from recent quotations, not a published index. The direction is well established and any supplier quoting a coastal yard should be able to defend their numbers line by line.
The failures we see on cranes that were specified for a dry inland factory and installed near water are rarely structural. They are the small things: a limit switch enclosure with a gasket that has gone hard, a brake lining that has absorbed moisture over a shutdown, a cable jacket that cracked at the festoon trolley, bolts at a handrail joint that were zinc plated and are now rust blooms. A bridge girder painted to C5-M is a boring, expensive decision. It is also the reason the machine is still worth rebuilding at year fifteen.
All of this demand lands on an industry that is short of people, and that constraint is upstream of the cranes as well as inside the yards.
Korea is the clearest case. Employment in the Korean shipbuilding industry fell 44 percent between 2014 and 2024, and skilled workers accounted for only about 28 percent of the workforce, according to an OECD peer review published in April 2026. Foreign workers now make up roughly 15 percent of total employment. Korean yards are short of experienced welders, painters and pipefitters, and younger workers are not lining up to replace them. HD Korea Shipbuilding & Offshore Engineering had 526 vessels worth about USD 75.6 billion on order as of late August 2026, and Hanwha Ocean 144 vessels worth USD 33.9 billion at the end of July. Those books are being delivered by a shrinking workforce.
The comparison that Korean yards keep making is with China, which took 63 percent of global newbuilding orders by compensated gross tonnage last year against Korea's 21 percent. Labour flexibility is part of the explanation, but so is the crane fleet. A yard that has to hire 20 percent more welders to hit an output target is constrained by hiring. A yard that lifts bigger blocks with fewer people is constrained by steel and cranes, which are easier to buy.
Crane builders face the same problem from the other side. Experienced welders, machinists, electrical and hydraulic specialists and engineers are in short supply across the equipment industry, and custom specifications stretch schedules further. A crane is a fabrication job before it is a product. When the orderbook for shipyard cranes stretches, the reason is often welding hours, not steel availability.
For a hall crane, the numbers are unglamorous and worth writing down. The table uses our published price bands for double girder machines and then applies the shipyard specification premium.
| Capacity and class | Span | Standard industrial, FOB | With shipyard spec, FOB |
|---|---|---|---|
| 20 t, A6 | 22 m | USD 38,000–55,000 | USD 44,000–72,000 |
| 32 t, A7 | 25 m | USD 55,000–85,000 | USD 63,000–110,000 |
| 50 t, A7 | 25 m | USD 75,000–130,000 | USD 86,000–170,000 |
| 75 t, A7 | 28 m | USD 110,000–180,000 | USD 127,000–234,000 |
| 100 t, A8, dual hoist | 30 m | USD 150,000–280,000 | USD 173,000–364,000 |
The comparison that matters is at the bottom of the scale, not the top. Six 50 tonne A7 hall cranes with shipyard specification come to roughly USD 0.5 to 1.0 million. One 1,800 tonne goliath is USD 40 to 60 million. The goliath is fifty to a hundred times the price and moves a hundredth of the lifts. Both are necessary. Only one of them is a routine purchase a yard can phase into a capex plan without a board decision.
If you are specifying a machine for a hull block or pre-outfitting hall, the double girder crane range covers the 5 to 100 tonne band and the same structure is used for the larger plate yard units. If the hall is open-sided rather than covered, the gantry crane range is the one to look at.
Most of the specification risk in a shipyard crane order sits in ten lines. Write them down and the quotation comparison becomes possible instead of rhetorical.
| Item | What to state | Why it matters here |
|---|---|---|
| 1 | Duty class and hours a year, not just capacity | A7 versus A6 is a mid-teens percentage on price and a decade on structure life |
| 2 | Span measured between what, exactly | Rail centres, building columns and hook coverage are three different numbers |
| 3 | Hook height above rail and below rail | Block height plus sling plus clearance decides the building, not the crane |
| 4 | Lifts per shift and mean load as a share of rated capacity | A crane rated 50 t that lifts 12 t all day wears differently from one lifting 45 t twice a shift |
| 5 | Distance to open water and the ISO 12944 category you want | C4 and C5-M are different paint systems at different prices |
| 6 | Paint system, dry film thickness and colour standard | Retrofitting a paint system on site costs several times the factory upgrade |
| 7 | Enclosure IP rating and the stainless schedule | Bolts and gaskets are the first failures, not girders |
| 8 | Control arrangement, including tandem or synchronised lifting | Tandem block lifts need matched speeds and a written procedure, not just two cranes |
| 9 | Voltage, frequency and the state of the hall supply | A new hall's transformer may arrive later than the crane |
| 10 | Documentation: load test certificate, welding standard, declaration of conformity, spare parts list | The most common reason a crane sits idle after energisation is a missing drawing |
Two of those items deserve a sentence each. On item 4, a shipyard crane's load spectrum is unusually polarised: mostly light panels, occasionally something at 90 percent of rated capacity. That is a fatigue question, and it is the reason duty class exists as a concept. On item 8, tandem lifting is normal in block assembly. Two cranes lifting one block need matched hoist speeds, a defined load share and a documented procedure, and none of that is implied by buying two identical machines.
Plenty, and buyers should hear it from a supplier rather than discover it later.
The orderbook is a record in cgt, but it is only 25 percent of the existing fleet, against 53 percent in 2008. That means the tonnage on order is not a bubble relative to demand for shipping, but it also means the delivery schedule is heavily back-loaded into 2027 and 2028. If contracting slows in 2027, the yards that added the most hall capacity carry the most empty floor space. BIMCO said as much in October, and they were not being alarmist, just arithmetic.
The fuel mix is the second uncertainty. Forty-three percent of the book is alternative-fuel capable, and those ships are not evenly distributed across yards. A yard tooled for bulkers is not automatically tooled for membrane containment work, which needs clean, dry, tightly controlled halls and different crane handling. The crane is the easy part of that transition. The building is not.
The third is price. Steel, motors and controls all moved in the last four years, and crane prices moved with them. Some of the quotes a yard receives in 2027 will be for equipment whose steel was bought at 2024 prices and whose delivery slot is 2028. That is normal in a capacity-constrained market, and it is worth asking a supplier which cost basis their validity period assumes.
What we would not do is order crane capacity speculatively on the strength of an orderbook headline. A hall crane ordered for a building that is not funded is a depreciating asset in a warehouse. Order the crane for the hall that is already committed, specify the duty class honestly, and let the market statistics stay in the market.
Clarksons Research put it at 226.1 million compensated gross tonnes in October 2026, slightly above the September 2008 record of 225.5 million and worth roughly USD 700 billion. It grew 24 percent in twelve months and is about three times the 2020 level. BIMCO measured 203.2 million cgt on 1 September, up 13 percent year on year, on a slightly different basis. The difference between the two figures is timing and coverage rather than direction.
Both, and they do different jobs. Goliath and shipbuilding gantry cranes at the dock handle whole blocks, from 900 tonnes up to the 2,000 tonne units now being built for Chinese and Korean yards. Overhead cranes inside the halls handle the panels, sub-blocks and outfitting work that make those blocks possible, typically 20 to 100 tonnes on 20 to 40 metre spans. A yard with a new goliath and an old hall crane fleet has bought lifting capacity it cannot feed.
A6 as a minimum for plate handling and outfitting, and A7 for block assembly and graving dock service. Continuous plate yard work can justify A8. Published marine crane guidance maps A6 to about 1,000 operating hours a year, A7 to 1,000–2,000 hours, and A8 to more than 2,000 with frequent near-rated lifts. If the hall runs two or three shifts, the honest answer is A7 whatever the original plan said.
Around USD 75,000 to 130,000 FOB for a standard industrial double girder machine on a 25 metre span, and roughly USD 86,000 to 170,000 once the marine specification is added: C5-M paint, IP66 enclosures, stainless fasteners and a marine-grade festoon jacket. Installation is additional, typically 20 to 40 percent of the equipment value depending on the building and the site. See the double girder crane specification pages for the configurations behind those ranges.
Four to eight weeks for a standard single girder machine, six to ten weeks for a box girder, and sixteen to twenty-two weeks for a semi-custom unit. Add ocean freight and customs for an export order, and two to four weeks on site for erection, connection and load testing. The very large goliaths are a different category entirely: the 1,800 tonne unit now being fabricated for Hanwha Ocean needs nine to twelve months of structural fabrication before shipment.
It is mixed. It means shipyards have funded work and will keep buying equipment, which is good for availability of designs and spare parts. It also means fabrication slots tighten, welding labour is scarce and lead times stretch, exactly as happened in the general crane market in 2025 and 2026. Buyers should order against a committed building programme, specify duty class and corrosion category honestly, and treat any delivery promise shorter than the current market as something to verify in writing.
Related articles:
Crane Order Books: What 2026 OEM Backlogs Mean for Your Delivery Date, 5 to 100 Ton
Transformer Lead Times: Why Your Crane Arrives Before the Power Does, 5 to 100 Ton
Shipping an Overhead Crane: What Landed Cost Really Looks Like in 2026
Double Girder Crane Cost: Equipment, Installation and 10 Year TCO, 10 to 100 Ton
Gantry Crane Cost: Equipment, Foundations and 10 Year TCO, 5 to 100 Ton
If you are specifying cranes for a fabrication hall, send us the span, the hook height and the shift pattern and we will tell you which duty class the work actually requires. Talk to our applications team.