Maritime and Shipping: A Practical Reference

A container ship carries standard 20 and 40 foot boxes in cell guides below deck and stacked on the hatch covers above, and its capacity is quoted in twenty-foot equivalent units (TEU). UNCTAD counted 7,044 of them at the start of 2026, totaling 389 million deadweight tonnes at an average age of 14.2 years, which works out to roughly 33 million TEU of slot capacity at the 12 deadweight tonnes per TEU that large ships run to. Size classes come from locks and berths rather than from naval architecture: a Panamax ship fits the 1914 Panama locks at 294 m long, 32 m beam and 12.2 m draft and carries roughly 4,500 TEU, a Neopanamax fits the 2016 locks at 366 m, 49 m and 15.2 m and carries 12,500–16,000 TEU, and the largest ships in service carry over 24,000 TEU and are held back by berth length, channel depth and crane outreach rather than by any canal. Nearly all of this tonnage runs liner service, which means a published rotation calling the same ports on the same weekday, roughly 900 ports across 178 economies and about 280 million container moves a year. That is the fact that separates this trade from every other segment on the sheet, because the operator knows a year in advance which ports the ship will bunker in, and it buys those bunkers on its own account.

Cost per TEU-mile falls steeply with ship size, which is why the fleet went from 4,500 TEU to 24,000 TEU in thirty years, and it is also why a boxship can only call at the handful of ports deep enough and craned for it. A 20,000 TEU ship burns on the order of 150 tonnes a day at 19 knots, and spreading that over 17,000 loaded boxes is what makes ocean freight cheap per unit. The fixed rotation makes an alternative fuel contractable: a supplier can be asked to commit tonnes per year at two named ports, which is impossible for a tramp ship. That shows up directly in the order book, with 440 dual-fuel container ships and vehicle carriers in service in early 2026 and 764 more ordered. The weaknesses are financial rather than technical. Freight rates swing by a factor of five or more within two years, the ships cost $180–275M each and are financed against those rates, and a slot filled with a methanol tank is a slot not filled with a paying box, so the 2.4x tank volume that methanol needs comes straight out of revenue. Cascading makes it worse: when new 24,000 TEU ships arrive on Asia-Europe, the displaced 14,000 TEU ships move to secondary trades and push everything below them down a rung, so oversupply spreads through the whole segment at once.

If you are placing an alternative-fuel bet anywhere in shipping, start here, because this is the segment where the company that buys the engine also buys the fuel. If you need a ship that can carry whatever cargo is offered wherever it is offered, do not buy a container ship; buy a bulker or a tanker, since a boxship without a liner network behind it has no business. If you are specifying a newbuild for a fixed Asia-Europe or transpacific rotation, dual-fuel methanol or LNG is defensible today, and the question to settle first is which two ports will supply it and at what annual volume. If the route is intra-regional feeder work under about 3,000 TEU, spend the money on hull form, coatings and speed instead, because feeder margins will not carry a $15–25M engine premium. And if you are chartering rather than owning, check who pays the bunkers before assuming the efficiency spec matters to you.

7,044 ships and 389M dwt at the start of 2026, about 33M TEU · Panamax 294 x 32 x 12.2 m and ~4,500 TEU · Neopanamax 366 x 49 x 15.2 m and 12,500–16,000 TEU · largest in service over 24,000 TEU · newbuild $180–200M at 15,000 TEU, $260–275M at 24,000 TEU, dual-fuel adds $15–25M · about 150 t/day of fuel at 19 knots on a 20,000 TEU ship · 440 dual-fuel boxships and car carriers in service, 764 on order

MSC Irina at 24,346 TEU, the largest ship in the trade; CMA CGM Jacques Saadé, a 23,112 TEU LNG dual-fuel ship delivered in 2020; Maersk's Ane Mærsk class of 16,000 TEU methanol dual-fuel ships; Maersk Halifax, a 15,000 TEU ship converted to methanol at Zhoushan in 2024; Seaspan, Danaos and Costamare, the non-operating owners that charter tonnage to the lines rather than running services themselves.

A 15,000 TEU newbuild runs about $180–200M and a 24,000 TEU ship $260–275M, with a dual-fuel engine, tanks and gas-safe machinery adding $15–25M on top; feeder tonnage of 2,500–3,000 TEU is closer to $45M. The lines own roughly half their slot capacity and charter the rest from non-operating owners on time charters of three to twelve years, so the fleet is split between companies that sell freight and companies that sell ships. Fuel is the largest single voyage cost and the arithmetic is easy to follow: a 20,000 TEU ship burns on the order of 150 tonnes a day at 19 knots, so a 26-day Asia-Europe leg is about 3,900 tonnes, and at $700–800 a tonne that is $2.7–3.1M of bunkers, or roughly $160–180 per TEU on a ship loaded to 17,000 boxes (derived from those figures here, not a published rate). Required power goes roughly with the cube of speed, so running the same ship at 22 knots instead of 19 costs about 1.5 times the daily fuel, which is why schedules stretched rather than tightened when bunkers went above $600. EU ETS adds about €220 per tonne of fuel oil at a €70 allowance, and half the emissions of a voyage into the EU are covered, so the same leg carries roughly €430k of allowances, about €25 per TEU. The order book shows what all of this adds up to: 440 dual-fuel container ships and vehicle carriers were in service in early 2026 with 764 more ordered, and the World Shipping Council counted 78% of container ship orders by deadweight as dual-fuel capable.

10 minHow Container Ships WorkPolyMatter · 1m+ views

8 minThe Economics of Owning A ShipNeu · 500k+ views

Dual-Fuel Fleet Dashboard (World Shipping Council) · Chapter 1.3 – Interoceanic Passages (Port Economics, Management and Policy)

A bulk carrier is a single-deck ship with large box-shaped holds and hatches wide enough to drop a grab into, and it carries iron ore, coal, grain, bauxite, cement and fertilizer loose rather than in packages. It is the largest fleet afloat by deadweight: UNCTAD counted 14,630 bulk carriers totaling 1.07 billion deadweight tonnes at the start of 2026, about 42% of world tonnage, at an average age of 12.9 years. The size classes are named after the places that limit them. Handysize runs 10,000–40,000 dwt and carries its own cranes so it can work ports with no shore gear; Supramax and Ultramax run 50,000–65,000 dwt; Panamax and Kamsarmax run 65,000–82,000 dwt, the latter capped at 229 m length by the bauxite jetty at Port Kamsar in Guinea; Capesize runs 100,000–200,000 dwt and is named for having to sail around the Cape of Good Hope rather than through the old Panama locks; Newcastlemax is the 300 m by 50 m limit of the coal berths at Newcastle, Australia, at roughly 210,000 dwt; and Chinamax or Valemax ships of 380,000–400,000 dwt draw 24 m and can only be handled at a short list of iron-ore terminals. Nearly all of this tonnage trades tramp, which means the next fixture decides where the ship goes.

A bulker is the cheapest way to move a tonne of anything, and it is simple: no reefer plugs, no cargo pumps, no cryogenic tanks, daily operating costs of roughly $5,000–7,500 depending on size, and newbuild prices from $30–31M for a Handysize to $72–80M for a Capesize. A Capesize burns about 38 tonnes a day at 14 knots while carrying 180,000 tonnes of ore, and bunkers are usually 40–60% of the voyage cost. Being able to take any cargo to any port is the whole commercial proposition, and it is also the reason the segment cannot decarbonize on its own. An owner who fits a methanol or ammonia engine has bought a ship that must still be able to trade to ports with no green fuel, so the premium buys an option that mostly goes unused, and on a time charter the fuel saving goes to the charterer anyway. Physical weaknesses are real too: bulk cargoes can shift or liquefy, which is why cargoes like nickel ore have sunk ships, and the wide hatch openings make the hull structurally demanding, which is why bulk carrier safety has its own IMO chapter.

If the cargo is homogeneous, low-value per tonne and does not need a box, use a bulker and size it to the smallest constraint on the route rather than to the largest ship you can finance. If either end of the trade has draft under about 12 m or no shore cranes, buy geared Handysize or Supramax tonnage and accept the higher cost per tonne-mile. If both ends are deepwater ore or coal terminals with a single counterparty, Capesize or Newcastlemax is the right answer and Valemax is worth looking at only for the handful of berths that can take it. Do not commit a tramp bulker to an alternative fuel: spend the money on hull form, energy-saving devices, coatings and speed, which apply on every voyage and need no fuel that the next port might not have. The exception is a dedicated trade with a fuel producer at one end, such as an ammonia or methanol plant shipping its own product, where the fuel and the cargo arrive at the same berth.

14,630 ships and 1.07 billion dwt at the start of 2026, 42% of world tonnage · average age 12.9 years · Handysize 10,000–40,000 dwt, Kamsarmax 82,000, Capesize 100,000–200,000, Newcastlemax ~210,000, Valemax 380,000–400,000 · newbuild $30–31M Handysize, $36–38M Kamsarmax, $72–80M Capesize · Capesize burns about 38 t/day at 14 knots · daily operating cost $5,000–7,500 · bunkers 40–60% of voyage cost on a voyage charter

Vale's Valemax and Guaibamax ships on the Tubarão–Qingdao iron ore run; the Baltic Exchange C5 and C3 routes, which price West Australia–China and Brazil–China ore freight; NYK and Mitsui O.S.K. ammonia dual-fuel bulker orders tied to specific producers; Star Bulk, Golden Ocean and Oldendorff among the larger owners, none of which holds more than a couple of percent of the fleet.

Newbuild prices run roughly $30–31M for a Handysize, $36–38M for a Kamsarmax and $72–80M for a Capesize or Newcastlemax, which is a fifth to a quarter of what a large container ship costs, and the ownership is correspondingly fragmented: thousands of owners, mostly Greek, Japanese and Chinese, with no single company holding more than a few percent of the fleet. That fragmentation is why the fuel decision stalls. Three charter structures split the fuel bill differently. On a voyage charter the owner carries a stated cargo from A to B for a rate per tonne and pays the bunkers, port costs and canal dues itself. On a time charter the charterer hires the ship by the day and buys the bunkers, while the owner keeps paying crew, insurance and maintenance. On a bareboat charter the charterer takes the ship with no crew for five to fifteen years and pays for everything, which makes the owner a financier rather than an operator. Only the voyage charter puts the fuel saving and the capital decision in the same company, and most large bulkers trade on time charter. The size of the prize is easy to check: a Capesize burning about 38 tonnes a day at 14 knots covers the 11,000 miles from Tubarão to Qingdao in roughly 33 days, so a laden and ballast round trip is about 2,300 tonnes of fuel, or $1.8M at $800 a tonne, against gross freight of $3.2–5.0M for 180,000 tonnes at the $18–28 per tonne the route has paid (derived here from those figures). Bunkers are therefore 40–60% of voyage cost, which is exactly why an owner on a time charter, who pays none of it, orders the conventional ship.

10 minHow a Bulk Carrier Works – Loading, Cargo, and Design Features3D Living Studio · 100k+ views

4 minWhy Are Bulk Carriers Named Handy Size and Cape Size?Sailor's Edge · 10k+ views

5.4 – Maritime Transportation (The Geography of Transport Systems) · The Impact of Distance and Narrow Waterway on Voyage Cost: Cost Formulation and Implementation on a Dry Bulk Carrier (Journal of ETA Maritime Science)

A tanker carries liquid cargo in tanks built into a double hull, moves it with its own pumps through a deck manifold, and is divided into crude carriers, product carriers with coated or stainless tanks, and chemical carriers with segregated tank groups. UNCTAD counted 12,931 oil tankers totaling 683 million deadweight tonnes at the start of 2026, at an average age of 21.1 years, which is the oldest of the major fleets by a wide margin. The trade is enormous: 2.04 billion tonnes of crude oil and another 2.28 billion tonnes of products and other liquids were loaded in 2024, roughly 36% of all seaborne cargo. Size classes come from waterways and terminals. An MR carries 45,000–55,000 dwt of clean products and can enter most ports, an Aframax of 80,000–120,000 dwt is named after the Average Freight Rate Assessment scale it was priced on, a Suezmax of roughly 150,000–160,000 dwt is the largest laden ship the Suez Canal's 20 m draft limit will pass, and a VLCC of 200,000–320,000 dwt carries about 2 million barrels and has to go around the Cape. The Strait of Malacca's 25 m depth sets the ceiling above that, which is why ULCCs over 320,000 dwt have almost disappeared.

Tankers are the one deep-sea segment where the fuel incentive is usually not split, because most spot crude and product cargo moves on voyage charter and the owner buys the bunkers. That matters at this scale: a VLCC costs $120–130M to build, about $8,000–9,500 a day to operate, and burns 45–50 tonnes a day laden, against $49–52M for an MR and $83–88M for a Suezmax. Cargo value per tonne is high enough that freight is a small share of it: crude at $75 a barrel is about $550 a tonne, so a $15 per tonne freight rate is under 3% of what is in the tanks, which means charterers care more about reliability and vetting than about a few dollars of fuel. Safety has genuinely improved, and the mandatory double hull is most of the reason: spills over 7 tonnes are down more than 90% from the 1970s, with the 2020s averaging about 7 a year. The weaknesses are the age profile and the politics. Sanctions on Russian and Iranian crude have kept several hundred older tankers trading outside mainstream insurance and inspection, which props up the average age at 21 years, holds down demolition, and makes vetting by the oil majors, rather than class or flag, the real quality filter.

If you are moving crude or refined product in parcels above about 30,000 tonnes, a tanker is the only option, and the size to pick is the largest one both terminals and the route can take. If the discharge port is draft-restricted or the parcel is under 60,000 tonnes, use MR or LR1 tonnage; if the route is Arabian Gulf to Asia with deepwater terminals at both ends, use a VLCC and accept that Suez is closed to you laden. If you own and trade the ship on voyage charters, fit every efficiency measure that pays back inside five years, because you buy the fuel and you keep the saving. If you charter the ship out on time, expect no fuel-driven premium and specify to the charterer's vetting requirements instead. Consider dual fuel only under a long charter to a major with its own emissions target, since the route mix will otherwise change faster than any bunkering network can follow.

12,931 tankers and 683M dwt at the start of 2026, average age 21.1 years · 2.04 Bt of crude plus 2.28 Bt of other liquids loaded in 2024, about 36% of seaborne cargo · MR 45,000–55,000 dwt, Aframax 80,000–120,000, Suezmax ~150,000–160,000, VLCC 200,000–320,000 and about 2M barrels · newbuild $49–52M MR, $83–88M Suezmax, $120–130M VLCC · VLCC burns about 45–50 t/day laden and costs $8,000–9,500/day to operate · spills over 7 tonnes down more than 90% since the 1970s

Worldscale, the annual nominal freight scale that crude and product fixtures are quoted against as a percentage of flat rate; the Baltic Exchange TD3C route for Arabian Gulf to China VLCC cargoes; Frontline, DHT and Teekay among the listed owners; the oil majors' vetting regimes, which decide which ships a charterer will accept; MARPOL Annex I, which made double hulls mandatory and retired the single-hull fleet.

An MR newbuild runs about $49–52M, a Suezmax $83–88M and a VLCC $120–130M, against daily operating costs of roughly $8,000–9,500 for the VLCC. Spot freight is quoted in Worldscale points rather than dollars: an annual schedule sets a nominal flat rate in $/tonne for every route, fixtures are done at a percentage of it, and the owner converts the result into a time charter equivalent per day to compare against a period charter. Because the dominant structure is the voyage charter, the owner pays the bunkers on most crude cargoes, and a VLCC burning 45–50 tonnes a day at $800 a tonne is spending $36,000–40,000 a day on fuel against $8,000–9,500 of operating cost, so fuel is roughly 40–60% of the voyage bill and the owner is the one who benefits from cutting it. That is why tanker owners fit energy-saving devices, premium coatings and hull cleaning programs without needing a charterer to ask, and it is also why the alternative-fuel share of the tanker order book sits near 10–20% rather than the container fleet's. Route mix is the blocker: a ship fixed today for the Arabian Gulf and next month for the US Gulf cannot rely on any port supplying methanol or ammonia, so the premium buys an option that mostly goes unexercised. Dual fuel gets ordered here only when a charterer with its own target signs a long enough charter to pay for it.

8 minHow An Oil Tanker Works And Designed3D Living Studio · 5m+ views

12 minThe Economics of Oil TankersMicro · 100k+ views

Oil Tanker Spill Statistics 2025 (ITOPF) · Efficiency in Dirty Tanker Market (Journal of ETA Maritime Science)

A gas carrier moves its cargo as a liquid at atmospheric pressure and very low temperature, or under pressure at closer to ambient, and the containment system is most of what makes the ship expensive. LNG rides at -162 °C, LPG at -42 °C, ammonia at -33 °C, and ethylene at -104 °C, so every tank is either a membrane lining an insulated hold, a self-supporting sphere or prismatic tank, or an IMO Type C pressure vessel. The modern LNG standard is 174,000 m³, about 78,000 tonnes, and Qatar's Q-Flex and Q-Max ships run 210,000 and 266,000 m³; the LPG fleet is built around VLGCs of 84,000–93,000 m³ and mid-size carriers of 35,000–40,000 m³, which are also the ships that carry ammonia. There are roughly 800–900 LNG carriers and around 1,700 LPG and chemical gas carriers trading, against LNG trade of a bit over 400 million tonnes a year. What makes the segment unusual is that the cargo boils: a 174,000 m³ ship losing 0.10% a day gives up 174 m³, or about 78 tonnes at LNG's 450 kg/m³, which is close to what the ship needs to make 17–19 knots. LNG carriers have therefore burned their own cargo since the 1960s, which makes them the only merchant ships that arrived with an alternative fuel already installed.

The containment system decides the ship. Membrane tanks from GTT hold most of the LNG order book because they use the hull form efficiently and give boil-off rates of 0.07–0.10% a day against 0.10–0.15% on older designs, and every percentage point of boil-off is cargo the buyer does not receive. Propulsion followed the same logic: steam turbines burned boil-off at about 30% efficiency for forty years, dual-fuel four-strokes reached the mid-40s, and the two-stroke ME-GI and X-DF engines now ordered reach about 50%, which turns surplus boil-off into a reliquefaction or a sales opportunity rather than a loss. The weaknesses are cost and cyclicality. A 174,000 m³ newbuild runs $250–260M against $120–125M for a VLGC, and needs a ten to twenty year charter to finance, and when deliveries run ahead of new liquefaction the spot market collapses, as it did from over $400,000 a day in late 2022 to under $20,000 within about three years. Crew supply is a real constraint too, because gas carrier certification takes years to build.

If you are shipping LNG, LPG or ammonia, the ship type is decided by the cargo, so the real questions are size and charter length. If the trade is a fixed liquefaction plant to a fixed import terminal, order a 174,000 m³ membrane ship against a long time charter, because that structure is what a bank will lend against. If the trade is US Gulf to Asia and you need the Panama Canal, check the Neopanamax booking situation before assuming the routing, since droughts have pushed ships around the Cape more than once. If the cargo is ammonia, look hard at ammonia dual fuel: the crew already handles the molecule, the containment already exists, and the loading terminal already has it, which removes all three of the usual blockers at once. If the volumes are small or seasonal, use pressurized Type C tonnage of 5,000–20,000 m³ and accept the higher cost per tonne rather than committing to a large ship you cannot keep employed.

Roughly 800–900 LNG carriers and about 1,700 LPG and gas carriers · LNG standard 174,000 m³ and about 78,000 tonnes, Q-Max 266,000 m³, VLGC 84,000–93,000 m³ · cargo at -162 °C LNG, -42 °C LPG, -33 °C ammonia · boil-off 0.07–0.10%/day on modern membrane tanks, about 78 t/day on a 174,000 m³ ship · newbuild $250–260M for a 174,000 m³ LNG carrier, $120–125M for a VLGC · propulsion efficiency 30% steam, mid-40s dual-fuel four-stroke, about 50% two-stroke

GTT's NO96 and Mark III membrane systems, which cover most of the LNG order book; the Moss spherical tanks visible on older LNG carriers; QatarEnergy's Q-Flex and Q-Max ships and its 100-plus vessel newbuild program; MAN ME-GI and WinGD X-DF two-strokes on ships ordered since about 2016; Exmar and BW ammonia dual-fuel mid-size gas carriers; the IMO IGC Code, which governs construction and equipment for every ship in this segment.

A 174,000 m³ LNG carrier costs $250–260M, a VLGC about $120–125M, and a mid-size ammonia carrier roughly $85–90M, so the capital per ship is the highest in merchant shipping outside cruise. Almost none of it is speculative: producers and portfolio players such as QatarEnergy, Shell and TotalEnergies take ships on ten to twenty year time charters, and the charter is what makes the ship financeable rather than the other way round. Freight is a small share of the delivered cargo, which is why charterers pay for reliability rather than for cheapness. A 174,000 m³ cargo is roughly 3.7 million MMBtu, so at $10/MMBtu it is about $37M, while a 30-day voyage at $80,000 a day is $2.4M, or around 6% of the cargo value (derived here from those figures). Fuel works differently in this segment because the fuel is the cargo. Charters carry a guaranteed boil-off rate alongside the usual speed and consumption warranty, so an owner whose insulation underperforms pays for the gas that did not arrive, and that puts the efficiency incentive back on the owner in a way no other time charter does. Alternative fuel is easier here than anywhere else for the same reason: the ship already handles a cryogenic hydrocarbon, and for the ammonia carriers the fuel and the cargo come off the same loading arm.

8 minHow LNG Carriers (Gas Tankers) Work - Design Types, Loading & Discharge3D Living Studio · 500k+ views

8 minInside an LNG Cargo Ship: Carrying Gas at -162°Cmarineinsight · 500k+ views

IGC Code (IMO) · NO96 technologies (GTT)

Roll-on/roll-off ships carry wheeled cargo that drives aboard over a ramp instead of being lifted, which covers pure car and truck carriers, freight ro-ros, ropax ferries carrying both vehicles and passengers, and pure passenger ferries. The decks are long undivided spaces with hoistable car decks, and that geometry is the segment's defining feature: it makes loading fast, and it means water on a vehicle deck has a free surface that can capsize the ship, which is why ro-ro damage stability has its own SOLAS rules written after the Herald of Free Enterprise and Estonia losses. The World Shipping Council counts over 900 vehicle carriers in service with more than 100 on order, the largest carrying up to 10,800 cars, while modern deep-sea tonnage typically runs 7,000–9,300 car equivalent units. The ferry side is much larger in people terms: Interferry puts the global industry at roughly 4.27 billion passengers and 373 million vehicles a year, comparable to commercial aviation. Short sea shipping, meaning intra-regional voyages that compete directly with road haulage, is roughly 60% of EU maritime freight tonnage and is the trade most of this tonnage serves.

No other segment runs routes this fixed. A ferry may return to the same berth twice a day for twenty years, and a car carrier runs a scheduled round trip between the same two or three loading ports, so both know exactly where they will take on fuel or electricity. That is why full battery-electric operation works here and nowhere else, and why the World Shipping Council counted 94% of vehicle carrier orders by deadweight as dual-fuel capable in May 2026. The costs are the other side of the same shape. Ro-ro cargo is bulky and light, so revenue per tonne of displacement is poor, and a ropax carries hotel staff, catering and passenger safety systems that a freight ship does not, which pushes crew cost well above other segments. Fire is the live risk: vehicle decks are hard to compartmentalize, and electric vehicles have made it worse, as the Felicity Ace and Fremantle Highway car carrier fires showed.

If the crossing is under about two hours with berths at both ends that can take a 1–5 MW shore connection, specify battery-electric, because charging twice an hour is only possible on a fixed berth and this is the one segment that has it. If the crossing runs two to eight hours, use a hybrid: batteries for peak shaving, maneuvering and harbor legs, and a conventional or dual-fuel engine at sea, which typically cuts 10–25% without depending on any new fuel. If it is a deep-sea car carrier on a fixed round trip, order LNG or methanol dual fuel now and specify ammonia-ready, since the route calls the same ports every rotation and the car makers behind the cargo have their own supply-chain targets. If the route is tendered under a public service contract, match the asset life to the contract: an eight to ten year contract will support a battery ship, a two-year one will not. And if you are carrying unaccompanied trailers rather than passengers, check whether a lo-lo container feeder is cheaper per unit before committing to ro-ro tonnage at all.

Over 900 vehicle carriers in service, more than 100 on order, largest up to 10,800 cars · modern deep-sea tonnage 7,000–9,300 CEU · roughly 4.27 billion ferry passengers and 373 million vehicles a year · short sea about 60% of EU maritime freight tonnage · 94% of vehicle carrier orders dual-fuel capable by deadweight, May 2026 · battery-electric practical under about two hours between charges · hybrid peak shaving cuts 10–25%

MF Ampere, the first battery-electric car ferry, in service in Norway since 2015, with roughly 80 electric car ferries operating there now; Ellen in Denmark, which runs a 22 nautical mile route on a 4.3 MWh pack; Höegh Autoliners' Aurora class LNG dual-fuel car carriers of 9,100 CEU built ammonia-ready; the Felicity Ace and Fremantle Highway fires, which put electric vehicles on car decks onto every operator's risk register; EMSA's STARSS study on carrying alternative fuel vehicles on ro-ro ships.

A large LNG dual-fuel car carrier runs roughly $110–125M, against about $75M for a conventional 6,500 CEU ship a few years earlier, and charter rates for that size went from around $20,000 a day in 2020 to over $100,000 in 2023 and 2024 as car exports outran the fleet. Freight is a small share of what is on the deck, which is why the manufacturers behind the cargo will pay for a cleaner ship: 6,500 cars on a 30-day voyage at $100,000 a day is $3M, about $460 a car against a vehicle worth $25,000–35,000, so a fuel premium that adds 20% to freight moves the retail price by well under half a percent (derived here from those figures). Ferries work on a different model again. Many routes are tendered by a public authority on eight to ten year contracts, and that contract, not the fuel price, is what makes a battery ship financeable, because the revenue is known for long enough to amortize a pack costing $400–700 per kWh installed. Fuel is typically 25–40% of a ropax operator's cost and crew is most of the rest, so electrification helps the fuel line while doing nothing for the largest one. The shore side is where the money often goes: a 1–5 MW charging connection with an automatic coupler frequently costs more than the batteries, and the grid upgrade behind it is the item that slips the schedule.

10 minHow Ro-Ro Ships Transport Cars Across the Ocean3D Living Studio · 500k+ views

3 minThis electric ferry uses a very long extension cordTom Scott · 1m+ views

Ferry Industry Facts (Interferry) · Comparison of External Costs of Diesel, LNG, and Electric Drive on a Ro-Ro Ferry Route (Promet - Traffic&Transportation)

A large marine two-stroke is a crosshead diesel that fires every revolution and turns slowly enough to drive the propeller directly, with no reduction gear between the two. Air enters through scavenge ports cut around the bottom of the liner and leaves through one hydraulically actuated exhaust valve in the cylinder head, so the charge sweeps up the bore in one direction and a turbocharger driven by the exhaust supplies it. Bores run from roughly 300 mm to 960 mm at rated speeds of about 55 to 130 rpm, which covers 2 to 80 MW on a single shaft; WinGD's current program spans 520–920 mm bores and 56–120 rpm. Fuel consumption is 165–185 g/kWh on the 42.7 MJ/kg reference fuel, an efficiency of roughly 47–52% with the large-bore engines at the top of the band, and that makes it the most efficient heat engine in production. The crosshead carries the piston side load on a guide and seals the combustion space away from the crankcase, which is why the engine can burn residual fuel carrying 3.5% sulfur, vanadium and catalyst fines without ruining the system oil. Modern engines are electronically controlled rather than camshaft-driven, so injection and exhaust valve timing can be retuned for whatever speed the ship actually runs at.

The efficiency and the fuel tolerance are the two reasons nothing has displaced it: a ship can bunker the cheapest residue sold in any of about a thousand ports and convert about half of it to shaft power. Direct drive matters more than it sounds: a propeller gets more efficient as it gets bigger and slower, so a large hull needs an 8–10 m propeller turning at 70–100 rpm, and this engine is built to turn at exactly that speed. That removes the reduction gear worth 1–2% of the delivered power, removes the clutch, and lets the engine itself run astern. Time between major overhauls is 20,000–30,000 hours against a 25-year hull life. The costs are size and emissions. A 12-cylinder 950 mm engine weighs on the order of 2,300 tonnes and stands about 17 m tall, so it dictates the engine room and much of the hull's midbody. And the efficiency does nothing for carbon, because CO2 is set by the fuel: burning a tonne of fuel oil releases about 3.1 tonnes of CO2 whatever the engine does with it. The specific failure modes are cold corrosion of the liner when a ship runs long periods at low load, and abrasive wear from aluminum-silicon catalyst fines that the fuel treatment plant failed to remove.

Assume this engine unless something specific rules it out, because it is the baseline every other option on this sheet is priced against. If the ship is a deep-sea bulker or tanker that takes whatever cargo is offered and cannot know its next ports, fit a conventional two-stroke and spend the money on hull form, energy-saving devices and coatings instead, since those cut fuel without needing a fuel that may not be there. If the trade is fixed and a supplier will commit tonnes per year, the dual-fuel version of the same engine is the upgrade, and it keeps full liquid-fuel capability as the fallback. If the ship needs more than about 25 MW, this is the only engine family that reaches it, so the question is only which fuel it burns. Below about 10 MW, and on any vessel that needs a shallow engine room or frequent maneuvering, medium-speed four-strokes through a gearbox usually win on installed height and cost.

Bore roughly 300–960 mm · rated speed about 55–130 rpm · 2–80 MW on one shaft · 165–185 g/kWh on 42.7 MJ/kg fuel, about 47–52% efficiency · time between overhauls 20,000–30,000 hours · about 3.1 t CO2 per tonne of fuel oil burned · a 12-cylinder 950 mm engine weighs around 2,300 t

MAN B&W ME-C engines (now sold under the Everllence brand) and the 950 mm G95ME-C series used on large container ships; WinGD's X-series, descended from Sulzer, including the 12X92DF; the 14RT-flex96C that powered Emma Maersk at about 80 MW; licensed construction by HD Hyundai Engine & Machinery, Doosan Enerbility, Japan Engine Corporation and CSSC yards, which build most of the world's units under license.

The design houses and the builders are different companies, and that is the unusual thing about this industry. MAN Energy Solutions and WinGD design the engines and collect license royalties; almost every engine is actually built by a Korean, Chinese or Japanese licensee next to the yard that will install it, which keeps engine prices tied to Asian shipbuilding capacity rather than to the designers' pricing power. A large two-stroke installation runs on the order of $200–500/kW, so a 70 MW main engine is a $15–35M line on a $180–200M container ship, roughly a tenth of the newbuild price. Everything after that is fuel: a 15,000 TEU ship at 20 knots burns something like 150–200 tonnes a day, which at $550/t VLSFO is $80–110k a day and usually exceeds the vessel's time-charter rate. That ratio is why a 1% efficiency gain is worth chasing and why owners de-rate engines for slow steaming rather than sell the ship. Running costs beyond fuel are small by comparison: cylinder oil consumption of roughly 0.6–1.0 g/kWh at $1,500–2,500 per tonne, plus scheduled overhauls every 20,000–30,000 hours. The party that captures the fuel saving depends on the charter, so on a time charter the owner buying a more efficient engine is spending capital to reduce someone else's bunker bill.

14 minHow a Ship Engine Works - 2-Stroke Marine Diesel Engine3D Living Studio · 1m+ views

6 minWorlds largest ship engine - 14 Cylinder - 14RT Flex96C Tier IIMaritime Report · 1m+ views

Basic principles of ship propulsion (MAN Energy Solutions) · Lowspeed Engines 2026 (WinGD)

A medium-speed engine is a four-stroke diesel running between about 250 and 1,200 rpm, with bores from roughly 200 mm to 640 mm and outputs from a few hundred kilowatts to about 25 MW. Because it turns ten times faster than a two-stroke, it cannot drive a propeller directly, so it works through a reduction gearbox or drives a generator that feeds an electric propulsion motor. Fuel consumption is typically 175–190 g/kWh, an efficiency of about 43–48%, with the best current designs near 50%. Every merchant ship also carries three or four auxiliary engines of 0.5–4 MW each, running at 720, 750, 900 or 1,000 rpm so the alternator produces 50 or 60 Hz directly, and those are what supply the ship's electrical load in port and at anchor. The same engine family therefore appears twice on most ships: as main propulsion on ferries, ro-ro, cruise and offshore vessels, and as the generating sets on everything else.

The strength is packaging and part-load behavior. A four-stroke is roughly a third the height and a fifth the weight of a two-stroke of the same power, which is what lets a ro-ro carry cargo over the engine room, and splitting the plant into four or six units means you run two at 85% load rather than six at 30%, where fuel consumption per kWh is far worse. Redundancy comes free from that arrangement, and it is why dynamic positioning vessels and cruise ships are diesel-electric. The weaknesses are efficiency and maintenance: 3–6 percentage points behind a two-stroke on fuel, a gearbox or a generator-plus-motor chain that costs another 5–10% of the power in conversion losses, and far more moving parts per megawatt, so top-end overhauls come every 12,000–20,000 hours instead of on a two-stroke's schedule. Most auxiliaries also burn distillate rather than residual fuel, which costs $150–250 per tonne more.

Pick medium-speed main engines when the hull cannot give up the height a two-stroke needs, when the ship maneuvers constantly, or when the load swings hard: ferries, ro-ro, offshore supply, tugs, dredgers and cruise. If the load profile is a flat deep-sea passage at one speed, use a two-stroke instead and accept the taller engine room, because the efficiency gap compounds over 250 sea days a year. Go diesel-electric when the propulsion load varies over a wide range or when thrusters need the same power at different times, since only then does the conversion loss buy anything back. For auxiliaries you have no choice, so the decisions are how many sets, and whether to add a shaft generator so the main engine carries the electrical load at sea. If the ship spends long periods at berth in a port with shore power, size the auxiliaries for sea load and plan the connection rather than adding another genset.

Speed roughly 250–1,200 rpm · bores about 200–640 mm · 0.5–25 MW per engine · 175–190 g/kWh, about 43–48% efficiency · gensets at 720, 750, 900 or 1,000 rpm for 50 or 60 Hz · three or four auxiliaries of 0.5–4 MW on a typical merchant ship · top-end overhaul every 12,000–20,000 hours · distillate costs $150–250/t more than residual fuel

Wärtsilä 20 through 64 and the Wärtsilä 31; Everllence (MAN Energy Solutions) 32/44CR and 49/60DF; Caterpillar MaK M32 and M46DF; Bergen B and C series; HiMSEN gensets from HD Hyundai, Daihatsu and Yanmar auxiliaries; diesel-electric cruise plants with six engines feeding azimuthing pods.

Installed cost runs roughly $300–600/kW, above a two-stroke per kilowatt, and a diesel-electric plant adds generators, switchboards, drives and motors on top, so the machinery package on a DP offshore vessel can approach a fifth of the newbuild price. The money comes back only if the duty cycle is variable enough that running fewer engines near full load beats one engine at part load. Auxiliary engines are a small capital item and a large operating one: a large container ship's hotel and reefer load of 3–8 MW burns something like 3–5 tonnes of gas oil a day at berth, which at $700–800/t is $2,000–4,000 a day paid by whoever holds the bunker account. That number is the whole business case for shore power and for battery peak shaving, and it is also what the EU ETS and FuelEU now price. Aftermarket is where the engine builders make their margin: Wärtsilä and Everllence both earn more from parts, service agreements and overhauls than from the initial sale, which is why service network coverage in the ship's trading area belongs in the selection criteria rather than in the footnotes.

8 minHow a Ship Engine Works - 4 Stroke Marine Diesel Engine3D Living Studio · 100k+ views

3 minHow Electricity is Generated on Ships?marineinsight · 50k+ views

Regulations for Emissions from Marine Vessels (US EPA) · Marine Power and Propulsion (MIT OpenCourseWare)

A dual-fuel engine burns natural gas with a small spray of liquid pilot fuel to ignite it, and keeps the full diesel system so it can run on fuel oil when gas is unavailable. There are two ways to get the gas in, and the choice decides almost everything about the emissions result. Low-pressure engines admit gas at 5–16 bar during the scavenging phase, so it mixes with the air before ignition and burns as a lean premixed Otto-cycle charge; high-pressure engines inject gas at about 300 bar directly into the combustion chamber after the pilot has already lit, so it burns as a diffusion flame on the diesel cycle. Premixing is what causes methane slip: some of the charge sits in the piston ring crevices and the cool layer against the liner where the flame never reaches, and it leaves unburned. Measured slip is about 2.5 gCH4/kWh on a low-pressure two-stroke and 5.5 gCH4/kWh on the common medium-speed four-stroke, against 0.2 gCH4/kWh on a high-pressure two-stroke, which works out to roughly 1.5–3.5% of the gas for the Otto-cycle engines and 0.2–0.4% for the diesel-cycle ones.

Burning methane instead of fuel oil removes essentially all the sulfur oxides and most of the particulates, and the lean premixed engines meet IMO Tier III NOx in gas mode with no aftertreatment at all. On carbon, gas has about 25% less carbon per unit of energy than fuel oil, which is where the headline number starts. What it ends at depends on the slip and on the warming horizon you count it over. The industry life-cycle study behind most vendor claims puts the well-to-wake cut at 10–20% for the Otto-cycle engines and up to 23% for the high-pressure two-stroke on a 100-year basis. The ICCT ran the same comparison on a 20-year global warming potential of about 86 for methane, and found no benefit for any engine type: the high-pressure two-stroke came out 4% worse than marine gas oil, and the popular medium-speed four-stroke 70% worse. Both are arithmetically correct and they disagree only about the time horizon, so ask which one a claim used before believing it. The other costs are physical: LNG holds about 21 GJ/m³ against 39 for fuel oil, so tanks and insulation take roughly 2.3 times the space, and the cryogenic fuel system, gas-safe machinery spaces and double-walled piping are most of the capital premium.

Specify high-pressure diesel cycle if the emissions claim has to survive 20-year methane accounting or a customer's own supply-chain target, and budget roughly $3–5M more for the 300-bar gas supply system. Specify low-pressure Otto cycle if the driver is cheap gas per gigajoule and compliance with the current 100-year accounting, and ask for the measured slip at the loads this ship actually runs, since slip roughly doubles below about 40% load. If the ship is an LNG carrier, the decision is already made, because cargo boil-off is free fuel and the tanks exist. If the ship is a tramp bulker with no fixed rotation, do not take LNG, because bunkering reaches only about 185 ports and a ship that cannot get gas burns diesel while still paying for the tanks. And if the route is inside the EU, run the FuelEU number with the engine's real slip factor rather than the marketing figure, because the penalty is charged on well-to-wake intensity including methane.

Slip about 2.5 gCH4/kWh low-pressure two-stroke, 5.5 gCH4/kWh medium-speed four-stroke, 0.2 gCH4/kWh high-pressure two-stroke · roughly 1.5–3.5% of gas slipped on Otto cycle, 0.2–0.4% on diesel cycle · methane 20-year global warming potential about 86 · well-to-wake cut 10–20% Otto cycle and up to 23% high-pressure diesel cycle on a 100-year basis · no benefit on a 20-year basis, with the medium-speed four-stroke 70% worse than gas oil · LNG holds about 21 GJ/m³ against 39 for fuel oil · about 185 bunkering ports

MAN B&W ME-GI high-pressure two-strokes on container ships and LNG carriers; WinGD X-DF low-pressure two-strokes including the 12X92DF, with the iCER exhaust-recirculation option added to cut slip; Wärtsilä 50DF and MaK M46DF four-strokes on LNG carriers and cruise ships; the EU-funded GREEN RAY project measuring and reducing slip on ships in service.

The premium on a large container ship is roughly $15–25M against the conventional version, most of it in the tanks, the cryogenic fuel handling and the gas-safe engine room rather than the engine itself. It pays back through the price spread per gigajoule and through carbon pricing. LNG has generally traded at $8–14/GJ against about $14/GJ for VLSFO at $550 a tonne, so the fuel saving alone is thin and swings with the gas market, which is why the 2022 gas price spike pushed several LNG-capable ships back onto fuel oil. Carbon pricing is the more reliable half: EU ETS at €70 an allowance costs about €220 per tonne of fuel oil burned, so a 20% cut is worth roughly €44 a tonne on top of the fuel spread. Slip is now priced too, since FuelEU counts methane in the intensity number, and that changes the ranking between the two combustion cycles rather than just the size of the prize. Resale is the open risk: an LNG-fueled ship sold in 2035 is worth whatever a buyer then thinks of methane, and nobody is pricing that today.

1 minDual Fuel Process - Engine on Gas | WärtsiläWärtsilä Corporation · 100k+ views

4 minMitigating Methane Slip in LNG Engines | GREEN RAYREVOLVE · 1k+ views

The climate implications of using LNG as a marine fuel (International Council on Clean Transportation) · Life Cycle GHG Emissions Study on the Use of LNG as Marine Fuel (SEA-LNG)

A methanol dual-fuel two-stroke injects liquid methanol into the cylinder at around 500 bar and lights it with a 3–5% spray of pilot diesel, burning on the diesel cycle with no premixing and therefore no slip worth measuring. Methanol is a liquid at ambient temperature and pressure, so the fuel system is pumps and pipes rather than cryogenics, which is the main reason this became the first alternative fuel with a converted two-stroke actually trading. Its flashpoint is about 12 °C, well under the 60 °C that SOLAS assumes, so it counts as a low-flashpoint fuel and needs cofferdams, inerting, double-walled piping and gas-safe machinery spaces under the IGF Code framework. It carries about 16 GJ/m³ against 39 for fuel oil, so a ship needs roughly 2.4 times the tank volume for the same range. Combustion is clean in the local sense, with no sulfur, very little particulate and NOx roughly 30–50% below diesel operation, but the carbon number depends entirely on where the molecule came from.

Handling is the strength. Methanol needs no pressure vessel and no insulation, it is already traded in bulk through existing chemical tanker and terminal infrastructure, and a spill disperses in water rather than persisting like fuel oil. Conversion of a large in-service two-stroke has been demonstrated, which no other alternative fuel can claim. The weakness is the molecule: almost all methanol bunkered today is made from natural gas, and its well-to-wake intensity is around 100 gCO2e/MJ against about 91 for fuel oil, so a ship burning it emits slightly more, not less. Bio-methanol and e-methanol cut 60–95%, but world output of those is under about 1 Mt a year while replacing even 5% of shipping's energy would take roughly 25 Mt. Methanol is also toxic if swallowed or absorbed through skin, burns with a nearly invisible flame, and holds less than half the energy per tonne of fuel oil, so bunkering takes longer and happens more often.

Choose methanol when the route is fixed, a supplier will sign a multi-year green offtake, and the ship has room for 2.4 times the tank volume without losing cargo that matters. If the ship is a liner container vessel or a car carrier whose operator buys its own bunkers and has customers with their own emissions targets, this is currently the most practical alternative fuel to specify. If nobody will commit green molecules, do not pay the premium expecting an emissions result, because on fossil methanol the ship is slightly worse than a conventional one and the only thing bought is an option on future supply. For an existing ship, methanol is the one conversion with a track record, at roughly $10–20M and about three months of yard time, so it is worth pricing against simply ordering a newbuild later. Against LNG, methanol costs less to install and has no methane problem, and costs more per gigajoule once the fuel is actually green.

Pilot fuel 3–5% · injection around 500 bar · flashpoint about 12 °C · about 16 GJ/m³, roughly 2.4x the tank volume of fuel oil · NOx 30–50% below diesel operation · fossil methanol about 100 gCO2e/MJ against 91 for fuel oil · bio and e-methanol cut 60–95% · green output under 1 Mt/yr against roughly 25 Mt for 5% of shipping's energy · retrofit about $10–20M and three months

MAN B&W ME-LGIM two-strokes, sold under the Everllence brand, in service on the Laura Maersk and the 16,000 TEU Ane Maersk class; WinGD's X-DF-M; the four-stroke Everllence 175DF-M and Wärtsilä 32 Methanol for ferries and offshore vessels; Maersk's conversion of the 15,000 TEU Maersk Halifax at Zhoushan in 2024; the European Green Pioneer and other chemical-tanker installations that run on methanol cargo.

The newbuild premium is roughly $10–15M on a large container ship, below LNG because the tanks are ordinary steel and there is no cryogenic plant, and a retrofit runs $10–20M with about three months out of service. Fuel is where the case is decided. VLSFO at $550 a tonne is about $14/GJ, while green methanol has been quoted at $50–80/GJ, three to five times as much. Working that through for a ship burning 150 tonnes of fuel oil a day: 150 t at 40.2 GJ/t is roughly 6,000 GJ, so the daily fuel bill goes from about $84,000 to $300,000–480,000 (arithmetic done here, not a published figure). No charterer absorbs that voluntarily, which is why the ships that have been ordered belong to liner operators who buy their own bunkers and can pass a surcharge to shippers with their own targets. Carbon pricing closes part of the gap rather than all of it: EU ETS costs about €220 per tonne of fuel oil burned and the FuelEU penalty is €2,400 per tonne of fuel-oil-equivalent deficit, and a compliant ship's surplus can be pooled across a fleet, which is worth more than the fuel saving on any single vessel. The producers capture most of the value in a tight market, so an owner without a signed offtake is buying the ship and hoping.

3 minClimate-neutral methanol engine for huge Maersk container vesselEverllence · 100k+ views

11 minWhy cargo giant Maersk is investing millions in methanol-fuelled ships | Transforming BusinessDW News · 100k+ views

Marine Methanol: Future-Proof Shipping Fuel (Methanol Institute) · Handling and Properties of Methanol as a Marine Fuel (National Laboratory of the Rockies)

An ammonia engine is a two-stroke dual-fuel diesel that injects liquid ammonia into the cylinder and lights it with about 5% pilot diesel, because ammonia has a very high autoignition temperature and a narrow flammability range and will not ignite on compression alone. Ammonia carries no carbon, so combustion produces no CO2 beyond what the pilot fuel contributes, and the engine's efficiency is close to the same engine on diesel. It liquefies at -33 °C at atmospheric pressure or at about 10 bar at ambient temperature, and holds roughly 13 GJ/m³ against 39 for fuel oil, so a ship needs about three times the tank volume. Two problems come with it. Ammonia is acutely toxic, with an occupational limit around 25 ppm over a working day and immediate danger to life in the low hundreds of ppm, so the whole fuel system is double-walled and every space it passes through needs detection, ventilation and a water curtain. And nitrogen in the fuel can form nitrous oxide, which warms about 265 to 273 times as much as CO2 per tonne, so a small slip cancels a large part of the carbon saving.

On green ammonia the well-to-wake cut is 85–95%, which is deeper than anything else that will scale to a 60 MW main engine, and the fuel needs no carbon source, unlike e-methanol. Everllence's test results on the 4T50ME-X-LGIA research engine and the 7S60ME-LGIA report nitrous oxide well under 5 ppm and total greenhouse emissions at full load of about 6.5% of the diesel case, so more than a 90% saving including the pilot oil. Treat that as the best case from the engine builder: it is measured at high load on a diesel-cycle engine, and N2O formation rises where combustion is cooler and less stratified, which is exactly what low load and transients look like. Unburned ammonia slip is a second problem, not a greenhouse one but a toxicity and secondary-particulate one, and it is handled by injector design plus the selective catalytic reduction unit the engine carries anyway for NOx. The molecule matters as much as with methanol: conventional ammonia comes from steam methane reforming and is worse than fuel oil on a well-to-wake basis, and green output today is a rounding error against the roughly 185 Mt a year the world makes for fertilizer.

Take ammonia only on a newbuild, on a dedicated trade, with a producer at one end and a plan for bunkering at the other, because there is no commercial ammonia bunkering network and pilot transfers are all that has happened so far. It fits gas carriers that already load ammonia as cargo and bulkers moving between a green ammonia plant and a fixed discharge port, and it fits nothing that trades wherever the next fixture sends it. If the trade is not that specific, choose methanol, which is easier to handle and has ships in service, or stay conventional and spend on efficiency. Before signing, settle three things that are not engineering: whether the ports on the rotation will accept an ammonia-fueled ship at all, what P&I terms cover a release, and where the crews get trained, since the STCW competence standard does not yet exist and the guide's rule of thumb of $2,000–6,000 per seafarer across two or three crews is the real schedule constraint. Retrofit is not an option, since no conversion of a large in-service two-stroke has been demonstrated.

Pilot fuel about 5% · liquid at -33 °C or about 10 bar · roughly 13 GJ/m³, about 3x the tank volume of fuel oil · occupational exposure limit around 25 ppm · N2O warms 265–273 times as much as CO2 · measured N2O under 5 ppm and total GHG about 6.5% of the diesel case at full load in builder testing · green ammonia cuts 85–95% well-to-wake, conventional ammonia is worse than fuel oil · world ammonia output about 185 Mt/yr, almost all fertilizer

Everllence (MAN Energy Solutions) B&W ME-LGIA, first full-scale testing on the 7S60ME-LGIA at Mitsui; WinGD X-DF-A with first engines delivered from 2025; Fortescue's Green Pioneer, the first dual-fuel ammonia vessel to get flag and class approval; ammonia-fueled gas carrier orders at Korean and Chinese yards; the Singapore MPA ammonia bunkering pilot.

An ammonia dual-fuel newbuild costs roughly 15–25% more than the conventional version, and unlike LNG or methanol almost none of that premium is offset by cheaper fuel. Ammonia holds 18.6 GJ per tonne, so conventional ammonia at $500 a tonne is about $27/GJ and green ammonia at $1,000 a tonne is about $54/GJ, against roughly $14/GJ for VLSFO (arithmetic done here from the tonne prices, not a published $/GJ series). Scale is the harder number: shipping burns on the order of 12 EJ a year, which at 18.6 GJ per tonne would take more than 600 Mt of ammonia, over three times what the world currently makes for every purpose combined. So the near-term market is a small number of dedicated ships, and the value sits with whoever owns green ammonia production and the terminal, not with the shipowner. What makes any of it bankable is regulation rather than fuel economics: the FuelEU penalty of €2,400 per tonne of fuel-oil-equivalent deficit, plus pooling, means one deeply compliant ship can carry the compliance of a fleet, which is currently the clearest way an owner gets paid for the premium. Insurance and liability are still being priced, and an unquantified P&I position is what stops projects that have already cleared the technical review.

2 minTesting of the first two-stroke ammonia engineEverllence · 50k+ views

6 minThe Next Clean Fuel Hype: AmmoniaSabine Hossenfelder · 100k+ views

Update on the development of ammonia-based propulsion systems based on Everllence two-stroke technology (Everllence) · Comparison of ammonia with methanol, liquefied natural gas and conventional marine transportation fuels through life cycle cost and emissions analysis (Argonne National Laboratory)

A fuel cell converts hydrogen and oxygen to electricity and water without combustion, so it produces no NOx, no sulfur oxides and no particulates at the stack. Two types matter at sea. Proton exchange membrane cells run at 60–80 °C, reach 50–60% electrical efficiency, start in seconds and follow load quickly, but they need high-purity hydrogen and will not tolerate sulfur or carbon monoxide. Solid oxide cells run at 600–1,000 °C, reach 55–60% electrical and up to about 85% if the waste heat is used aboard, and can reform methanol, LNG or ammonia internally, but they take hours to warm up and dislike thermal cycling. Marine installations are built from 100–200 kW modules paralleled into 0.1–5 MW plants, which is auxiliary or small-vessel power rather than main propulsion on anything large. The binding constraint is usually the fuel rather than the cell: liquid hydrogen holds about 8.5 GJ/m³ against 39 for fuel oil, so tanks and insulation take seven or eight times the space.

Efficiency at part load is the real advantage. A diesel genset at 25% load burns far more per kWh than at 85%, while a fuel cell's efficiency is flat or slightly better as load falls, which suits harbor operation, hotel load and dynamic positioning. There is also no noise and no vibration, which matters on research vessels and passenger craft. Against that, installed cost is roughly $2,000–5,000/kW against $200–400/kW for a diesel genset, a factor of ten, and stacks are consumables with lives on the order of 20,000–40,000 hours before replacement. Nothing exists near the 20–70 MW a large merchant ship needs, and hydrogen bunkering exists at a handful of demonstration berths. As with every other entry here, the emissions result belongs to the molecule: a fuel cell running on hydrogen made by steam methane reforming is worse well-to-wake than a diesel burning fuel oil.

Use fuel cells where the vessel is small, returns to one berth, and someone else is paying for the fuel infrastructure: harbor ferries, research vessels, and offshore support where the charterer will fund it. Fit them as auxiliary or harbor power on a larger ship if the requirement is zero emissions at berth and shore power is not available, and compare that against simply plugging in, which costs far less where the terminal has a connection. If the duty is a deep-sea voyage, do not consider hydrogen fuel cells, because the tank volume alone rules it out at today's storage densities. If the attraction is quiet variable-load power rather than zero emissions, a battery hybrid usually delivers most of the benefit at a fraction of the capital. And if the vessel already handles ammonia or methanol, a solid oxide cell running on that fuel avoids the hydrogen storage problem entirely, at the price of a plant that cannot be cycled off between trips.

PEM 50–60% electrical efficiency at 60–80 °C · solid oxide 55–60% electrical and up to about 85% with heat recovery, at 600–1,000 °C · marine plants of 0.1–5 MW built from 100–200 kW modules · installed cost roughly $2,000–5,000/kW against $200–400/kW for a genset · stack life about 20,000–40,000 hours · liquid hydrogen holds about 8.5 GJ/m³ against 39 for fuel oil

Norled's MF Hydra, the first ferry running on liquid hydrogen, with two 200 kW PEM stacks; MV Sea Change on San Francisco Bay, a 360-passenger hydrogen ferry with about 600 kW of PEM; the EU ShipFC project putting a 2 MW ammonia solid oxide plant on the offshore vessel Viking Energy; Sandia National Laboratories' SF-BREEZE and Zero-V feasibility studies, which set out the volume and cost limits that later designs worked within.

Nothing about this pencils on fuel savings alone, so almost every vessel in service was paid for by a public program: Norwegian ferry tenders that specify zero emissions, California grants and harbor-craft rules, and EU Horizon funding for the demonstrators. The capital gap is the reason. A 1 MW fuel cell plant at $2,000–5,000/kW is $2–5M against $200–400k for a diesel genset of the same output, and the stack has to be replaced once or twice over a 25-year hull life. Fuel makes it worse before it makes it better: green hydrogen at $4–8/kg is roughly $33–67/GJ against $14/GJ for VLSFO, and the fuel cell's efficiency advantage over a medium-speed diesel recovers only a fraction of that. What could change it is volume in road and stationary markets, since marine stacks are the same cells in a marinized enclosure and the cost curve is set elsewhere. See `hydrogen-and-efuels` for how the fuel is made and what it costs, and `power-electronics` for the converters between the stack and the propulsion motor.

7 minHydrogen: The Savior of the Shipping IndustrySciShow · 100k+ views

3 minHydrogen-powered ferry prepares to launch in San FranciscoReuters · 1k+ views

Hydrogen and Fuel Cells for Maritime Applications: From Sandia Feasibility Studies to First Demonstrations (Sandia National Laboratories) · ShipFC (Clean Hydrogen Partnership)

A marine battery installation is lithium-ion cells in class-approved racks with liquid or air cooling, gas venting to open deck, and water mist or deluge suppression, feeding the ship's DC or AC switchboard through converters. It gets used two ways. Full-electric vessels carry 0.5–6 MWh and charge at the berth through a 1–10 MW connection, often with a shore-side buffer battery because the local grid cannot take that peak directly; crossings of under about one to two hours are what this suits, since the pack has to be recharged in the few minutes a ferry spends alongside. Hybrid vessels carry a smaller pack alongside conventional gensets and use it for peak shaving, for spinning reserve so a second engine does not have to idle, and for load smoothing on dynamic positioning, which typically cuts 10–25% of fuel with no change of fuel. The physical limit is energy density: a battery room stores on the order of 50–100 kWh per cubic meter, roughly 0.2–0.4 GJ/m³ against 39 GJ/m³ for fuel oil, and even allowing for an electric drivetrain being about three times as efficient end to end, the useful energy per cubic meter is still around 30 times worse (that comparison is arithmetic done here, not a published figure).

Batteries do the two things a diesel plant does badly: they respond in milliseconds, and they cost nothing to have available. That removes the running spare genset, keeps the remaining engines near their best load, and makes the dynamic positioning system far more stable, which is where the 10–25% comes from on offshore vessels. Maintenance falls with engine hours, and a full-electric ferry has no exhaust, no vibration and no fuel handling at all. The weaknesses are volume, cost and life. Installed cost is $400–700/kWh, so a 6 MWh ferry pack is a $2.4–4.2M item before the shore connection, and the pack is a wearing part with roughly 3,000–7,000 cycles to 70–80% of its original capacity, so a ferry cycling twice an hour replaces it partway through the hull's life. Charging infrastructure is a per-berth cost that somebody has to build, and on a route with two terminals that means two of them.

Go full electric when the crossing is under about one to two hours, both terminals can be fitted with chargers, and the schedule leaves five to ten minutes alongside; that combination describes Norwegian and Danish car ferries and almost nothing else. Go hybrid when the load swings hard, which means offshore supply vessels on dynamic positioning, tugs, and harbor craft, and size the pack to the swing rather than to any range target. If the vessel runs a steady deep-sea passage, do not fit a large battery, because there is no load variation to harvest and the volume goes to waste. If the driver is emissions at berth rather than at sea, compare a pack against shore power first, since a cable connection is usually cheaper than a battery that has to be charged from the same grid. And check who owns the berth: on a public ferry contract the authority usually pays for the charger, while on a commercial route the shore side is the part with no clear owner.

Full-electric packs typically 0.5–6 MWh · charging connections of 1–10 MW in a few minutes alongside · hybrid peak shaving cuts 10–25% of fuel · installed cost $400–700/kWh · roughly 3,000–7,000 cycles to 70–80% of original capacity · battery rooms store about 50–100 kWh/m³, roughly 0.2–0.4 GJ/m³ against 39 GJ/m³ for fuel oil · full electric suits crossings under about one to two hours

MF Ampere, the first battery-electric car ferry, working a Norwegian crossing since 2015 with shore-side buffer batteries at both quays; the Norwegian and Danish electric ferry fleets built out under public tenders that specified zero emissions; Incat Tasmania's China Zorrilla for Buquebus, with a pack of about 40 MWh; hybrid retrofits across the North Sea offshore supply fleet; California's Commercial Harbor Craft rule, which pushes hybrid and zero-emission propulsion onto tugs and crew boats by regulation.

The energy cost genuinely falls, and it is worth doing the arithmetic rather than trusting the headline. Marine gas oil at $800 a tonne is about $19/GJ, or roughly $0.067 per kWh of fuel energy, and at about 40% from tank to shaft that is around $0.17 per kWh at the propeller. Electricity at $0.10/kWh, through charger, battery and drive at roughly 90%, lands near $0.11 per kWh, so the energy bill falls by about a third, and much further where hydro power sells for half that (steps shown because this is derived, not quoted). A hybrid retrofit shows the same shape at smaller scale: a 500 kWh to 1 MWh pack costs $1–3M installed, and on an offshore vessel burning 10–15 tonnes of gas oil a day a 15% saving is $1,200–1,800 a day, so roughly $300–500k a year at 250 operating days and a three to six year payback. That is real but not fast enough to explain the order book, which is mostly driven by public ferry tenders and by rules like California's harbor craft regulation. The margin sits with the pack integrators and the shore-charging suppliers rather than with the shipyard, and battery prices track the automotive cost curve, so the marine premium over automotive packs is the number to watch.

12 min100% Electric Ferry Crossing | Fully Charged 4kEverything Electric CARS · 100k+ views

2 minWorld's largest electric battery powered ship switched on | ABC NewsABC News (Australia) · 50k+ views

Exploring the cost and emissions impacts, feasibility and scalability of battery electric ships (Lawrence Berkeley National Laboratory) · Commercial Harbor Craft (California Air Resources Board)

A nuclear ship replaces the fuel tanks and the diesel with a reactor raising steam for a turbine, or with a reactor driving generators and an electric propulsion motor. The physics has been settled for seventy years: naval fleets have run pressurized water reactors since the 1950s, and Russia's Arctic fleet alone had accumulated roughly 400 reactor-years by 2021. A merchant application needs 25–70 MW at the shaft, which is a reactor of about 100–200 MW thermal, and a core that runs several years between refuelings rather than the 20,000 hours between engine overhauls. Four civil ships have been built: NS Savannah from 1962, Germany's Otto Hahn from 1968, Japan's Mutsu, which never traded, and Russia's Sevmorput from 1988, a 61,900 tonne container and lighter carrier with a KLT-40 reactor of 135 MWt driving 32.5 MW to the propeller. Current work is design studies rather than ships, and this entry is about the marine application; see `nuclear-reactors` for the reactor types themselves.

The attraction is that the fuel bill and the carbon exposure both go to roughly zero, and a ship that does not stop to bunker can also run faster, because a nuclear plant's cost barely changes with output while a diesel's fuel bill rises with the cube of speed. What has stopped it every time is not the reactor. NS Savannah was a technical success and was never commercially competitive: she was designed as a showpiece with accommodation for 60 passengers and only 9,250 tons of deadweight, she carried a complement of 124 crew, and the whole project consumed over $90 million before the Maritime Administration retired her in 1971. A 1963 dispute over pay differentials between deck and engineer officers shut her reactor down and left her idle at the dock for almost a year, which is a crewing problem rather than a nuclear one. Otto Hahn sailed 650,000 nautical miles on 126 voyages in ten years with no technical trouble and was converted to diesel in 1982 because she was too expensive to run. Sevmorput was laid up for years and only returned to service in 2015 after Rosatom overhauled her rather than scrap her, inside a fleet that was still drawing a state subsidy of about 40% as late as 2011.

Treat this as a research position rather than a procurement option today. If you are a shipowner, the question to ask first is not about the reactor but whether the ports on the route will accept the ship at all, because entry is a port state decision and past nuclear merchant ships were refused in several places. Then ask who insures it: the 1962 Brussels convention on the liability of operators of nuclear ships never entered into force, so there is no international third-party liability regime for a commercial nuclear vessel, and P&I clubs have nothing standard to sell. Then ask where the crew comes from, since STCW has no competence standard for reactor operators at sea and the naval pipeline is not available to merchant fleets. If those three answers exist for a specific trade, the case is strongest on a large ship that runs a fixed high-speed liner route at high utilization, since that is where fuel spend is largest and the capital is spread over the most sea days. If any of the three is missing, the technology choice is irrelevant.

25–70 MW at the shaft for a merchant ship, from a reactor of roughly 100–200 MWt · Russian Arctic fleet about 400 reactor-years by 2021 · NS Savannah: 74 MWt reactor, 20,000 shp, 21 knots, 9,250 tons deadweight, 60 passengers, 124 crew, over $90M project cost, retired 1971 · Otto Hahn: 650,000 nautical miles on 126 voyages in 10 years, converted to diesel in 1982 · Sevmorput: 61,900 t, 135 MWt, 32.5 MW to the propeller, laid up and returned to service in 2015 · advanced SMR overnight capital cost estimates of $5,500–10,000/kWe

NS Savannah, preserved and decommissioned by the US Maritime Administration; the Otto Hahn and Mutsu programs, both ended; Sevmorput and Rosatom's Atomflot icebreaker fleet, the only nuclear commercial vessels still operating; the NuProShip project, whose first phase screened 99 small modular reactor developers in January 2025 and shortlisted three designs, with DNV, the Norwegian Maritime Authority and Knutsen among the partners.

The operating case is easy and the capital case is not. Idaho National Laboratory's cost meta-analysis puts advanced small modular reactor overnight capital at roughly $5,500–10,000 per kilowatt electric, and a container ship's power plant is on the order of 60 MWe, so the reactor and its containment alone come to something like $330–600M (arithmetic done here from those two figures). A conventional 15,000 TEU ship costs $180–200M complete. So the plant costs two to three times the entire ship it replaces, and that gap has to be closed by fuel and carbon savings over 25 years plus a higher service speed. INL's breakeven modeling finds the numbers can work at capital costs above current marine SMR estimates, and that a carbon price widens the advantage, because carbon cost on a fossil ship overtakes fuel cost within a few years of a rising price. Three costs sit outside that model and are the ones that killed the earlier ships: insurance for a risk nobody has priced, crews who have to be trained and paid a premium, and decommissioning at the end. Whoever moves first also pays for the regulatory framework, since the IMO code for nuclear merchant ships dates from 1981 and port state acceptance would have to be negotiated one country at a time.

11 minWhy Can’t Cargo Ships Go Nuclear?Casual Navigation · 50k+ views

12 minNS Savannah: The First Nuclear-Powered Merchant ShipSideprojects · 50k+ views

Considerations for Maritime Nuclear Technologies, Economic Viability and Public Acceptance (Idaho National Laboratory) · Nuclear-Powered Ships (World Nuclear Association)

Residual fuel oil is what is left at the bottom of the crude barrel after the distillates are taken off. It is thick enough that a ship heats it to 130–150 °C before it will pump and atomize, and it carried up to 3.50% sulfur until 2020. MARPOL Annex VI Regulation 14 cut the global cap to 0.50% on 1 January 2020, and emission control areas had already been at 0.10% since 2015. An owner can comply two ways: buy 0.50% very low sulfur fuel oil, which is a blended product, or keep burning cheap 3.50% high sulfur fuel oil and fit an exhaust gas cleaning system, which the same annex allows as an equivalent means. A scrubber sprays the exhaust with water inside a tower, so sulfur dioxide dissolves and is neutralized and the washwater carries the sulfur away as sulfate. Open-loop units take raw seawater at roughly 45 m³ per MWh and discharge it overboard, relying on the ocean's own alkalinity; closed-loop units recirculate fresh water dosed with caustic soda and hold a small bleed-off aboard. Either type removes over 98% of the SOx and none of the CO2.

The reason to fit a scrubber is the price spread. High sulfur fuel oil has traded roughly $80–150 a tonne under VLSFO for most of the period since 2020, and a ship burning 25,000 tonnes a year turns that spread straight into cash. High sulfur fuel is also more consistent than VLSFO, which is a blend and has caused stability, compatibility and cat-fine problems when two stems mix in a tank. The weakness is that the equipment moves pollution from the air into the water: open-loop washwater is acidic and carries polycyclic aromatic hydrocarbons, nitrates, vanadium and nickel, and a growing list of ports and coastal states now prohibit discharging it in their waters. The scrubber also burns 1–2% more fuel to drive its pumps and removes no CO2, so under EU ETS and FuelEU a scrubber-fitted ship carries slightly more carbon cost than the ship it was supposed to save money against.

Treat this as a trade with a payback period, not an emissions measure. If the ship burns more than roughly 20,000 tonnes a year, has eight to ten years of trading life left, and stays out of the ports that ban open-loop discharge, a scrubber at $2–6M pays back in one to three years as long as the spread holds above about $100 a tonne. If the ship spends much of its time in restricted waters, specify hybrid or closed-loop and budget for the caustic soda supply and the bleed-off tank, or skip the scrubber and buy compliant fuel. If the ship burns under about 10,000 tonnes a year, or is within a few years of demolition, buy VLSFO and spend the capital on hull coating and energy-saving devices instead. And if the reason for the project is a decarbonization target, this is the wrong tool, because it cuts sulfur and nothing else.

0.50% global sulfur cap since 2020 and 0.10% in emission control areas · scrubber $2–6M installed · HSFO usually $80–150/t under VLSFO · payback 1–3 years at 20,000–25,000 t a year · over 98% of SOx removed and no CO2 · 1–2% extra fuel for the pumps · about 45 m³ of seawater per MWh in open loop

Alfa Laval PureSOx, Wärtsilä, Yara Marine Technologies and Pacific Green as the main suppliers; Singapore, Fujairah, the Panama Canal, California and much of China's coastal and inland water among the places that prohibit open-loop discharge; MARPOL Annex VI Regulations 4 and 14, which set the cap and allow the equivalent-means route.

The return is entirely in the spread, and the arithmetic fits on one line. A large container ship or VLCC burning about 100 tonnes a day for 250 sea days uses 25,000 tonnes a year; at a $100 spread that is $2.5M of gross saving, and after the 1–2% extra fuel and the consumables roughly $2.2M net, so a $4M installation pays back in under two years. That number is derived here rather than published, and every term in it moves. The spread was near $300/t in January 2020, fell below $50/t by the middle of that year when demand collapsed, and has mostly run $80–150/t since. It is also self-limiting, because every scrubber fitted adds high sulfur demand and narrows the spread that justified fitting it. On a time charter the charterer buys the bunkers, so the owner recovers the capital through a scrubber premium on the daily hire rate, which tracks the spread and disappears with it. The risk that writes the investment off is regulatory rather than commercial: more than 5,000 ships now carry scrubbers, most of them open-loop, and each new port-level discharge ban shrinks the water in which they are worth having.

4 minScrubber System in Ships - Wärtsilä Open Loop Exhaust Gas Cleaning SystemWärtsilä Corporation · 50k+ views

19 minIMO 2020 SULPHUR CAP LIMIT|EXHAUST GAS CLEANING SYSTEM|Marine Engineering Hub · 5k+ views

IMO 2020 – cutting sulphur oxide emissions (International Maritime Organization) · Global scrubber washwater discharges under IMO's 2020 fuel sulfur limit (International Council on Clean Transportation)

LNG is natural gas chilled to -162 °C, at which point it holds about 450 kg/m³ and roughly 21 GJ per cubic meter against 39 for fuel oil, so a ship needs about 2.3 times the tank space for the same range once insulation and tank structure are counted. Bunkering happens three ways. Truck-to-ship moves 20–60 m³ a load and suits a small ferry and nothing larger; ship-to-ship from a dedicated bunker vessel runs 500–2,000 m³ an hour and is the only practical way to fill a deep-sea ship; shore-to-ship comes by pipeline from an import terminal. About 185 ports now offer LNG with roughly 60 bunker vessels behind them, concentrated in Rotterdam, Singapore, Zeebrugge, Barcelona, Fujairah, the US Gulf and the Chinese coast. Counts vary with the definition: tallies of ports with LNG bunkering capability, including planned and occasional supply, run past 250, while the number that will reliably sell you a cargo today is smaller. Fuel tanks aboard non-gas ships are usually IMO Type C pressure vessels, cylindrical or bilobe and designed for several bar, so boil-off raises the tank pressure instead of venting; the largest container ships use membrane tanks instead, and boil-off runs roughly 0.1–0.5% of the contents a day depending on the tank. The transfer is governed by the IGF Code on the ship side and ISO 20519 at the interface, which between them require inert-gas purging, a cool-down step, an emergency shutdown link between the two vessels, and a hazardous zone kept clear around the manifold.

LNG is the only alternative marine fuel with a real network behind it, and the fuel is a traded commodity with a published price rather than a bilateral offtake. It removes almost all SOx and most particulates, so no scrubber is needed, and a low-pressure Otto engine meets NOx Tier III without aftertreatment. The weakness is that the price is a gas price and gas moves independently of oil: TTF went above $70/GJ in 2022, which made LNG about five times VLSFO per unit of energy and sent dual-fuel ships back to burning diesel for a year. Handling also has to be exact in ways oil handling does not, because a cryogenic spill embrittles ordinary steel and the vapor cloud stays heavy and near the deck until it warms. And LNG only cuts 10–23% well-to-wake even when everything works, with the number set by the combustion cycle and its methane slip rather than by the fuel.

If the ship runs a fixed rotation, both ends of it supply LNG, and gas stays under about $12/GJ, this is the alternative fuel with the least execution risk, because the engine, the tank and the supply chain all exist today. If the ship is a tramp trader taking whatever cargo is offered, do not specify it, since the fuel will not be where the next fixture goes. If the emissions claim has to survive 20-year methane accounting, buy a high-pressure diesel-cycle engine and check the slip figures in entry 8 before believing the cut. If truck-to-ship is the only mode the port offers, work the fill against the port stay first, because a 6,000 m³ bunker at 40 m³ a truck is 150 truck movements and no liner schedule absorbs that. Where none of this fits, methanol needs no cryogenics but reaches about 30 ports, and certified biofuel blends need no capital at all.

About 21 GJ/m³ against 39 for fuel oil, roughly 2.3x the tank volume · about 185 ports and 60 bunker vessels · ship-to-ship 500–2,000 m³/h · truck-to-ship 20–60 m³ a load · boil-off 0.1–0.5% a day · LNG fuel system $15–25M on a large newbuild · gas under about $12/GJ for the fuel case to work

Rotterdam and Singapore as the two largest bunkering hubs; Gasum, Titan, Shell and TotalEnergies operating bunker vessels, the largest of which now run 18,000–20,000 m³; ISO 20519 and the SGMF bunkering guidance; simultaneous cargo and bunkering operations, now permitted at the main hubs so the fill does not extend the port stay.

Two capital stacks have to be built at once, which is why the network grew slowly. On the port side a 5,000–20,000 m³ bunker vessel costs roughly $50–100M and needs several ships a week to be worth operating, while a break-bulk berth at an existing import terminal is cheaper and only serves ships that can come alongside there. On the ship side the gas fuel system adds $15–25M to a large newbuild. Delivered LNG prices off the TTF or JKM hub plus liquefaction, delivery and a bunkering margin, typically $2–4/GJ over the hub, so a $10/GJ hub gives $12–14 delivered against about $14 for VLSFO, and the fuel saving is thin to negative in a normal year. The more reliable part of the return is the carbon bill: full EU ETS phase-in at €70 an allowance costs about €220 per tonne of fuel oil burned, and LNG emits roughly a quarter less CO2 per unit of energy, so the ETS exposure falls by about that much. From 2026 the EU counts methane and nitrous oxide as well as CO2, which takes part of that back on an Otto-cycle engine and almost none of it on a diesel-cycle one. Nearly all of the buying is by liner operators and car-carrier owners, because they buy their own bunkers and therefore keep the saving they paid for.

5 minFull expert video about bunkering liquefied natural gas (LNG)Gasum · 10k+ views

9 minStep by step LNG Bunkering by DNVLars Petter Blikom · 10k+ views

Guidance on LNG Bunkering to Port Authorities and Administrations (European Maritime Safety Agency) · International Code of Safety for Ship Using Gases or Other Low-flashpoint Fuels (IGF Code) (International Maritime Organization)

Methanol is a liquid at ambient temperature and pressure, 792 kg/m³ and 19.9 MJ/kg, which works out at about 16 GJ per cubic meter and roughly 2.4 times the tank volume of fuel oil for the same energy. That makes it the easiest alternative marine fuel to supply: no cryogenics, no pressure vessel, ordinary steel tanks in the double bottom or the wings, and an existing chemical barge can deliver it after a change of procedures and a port permit rather than a rebuild. It is still a low-flashpoint fuel at about 12 °C, so the IGF Code applies and the ship needs inerted tanks, cofferdams, double-walled fuel piping and detection through the machinery spaces. About 30 ports have supplied it as a bunker fuel, against roughly a thousand that sell fuel oil, though many more already handle methanol as a chemical, so the port count can grow quickly once there is demand to serve. The constraint is the molecule rather than the port. World methanol production is about 110 Mt a year and almost all of it is made from natural gas or Chinese coal, which puts it near 100 gCO2e/MJ well-to-wake against about 91 for fuel oil, so a ship burning it emits slightly more; bio-methanol and e-methanol cut 60–95%, and certified output of those runs under about 1 Mt a year.

Handling is the strong point. Methanol bunkers like a chemical cargo, the two-stroke engine is in catalog production, a 15,000 TEU ship has already been converted in service, and a spill dissolves and biodegrades instead of forming a slick. Against that, methanol is poisonous by ingestion and skin absorption, its flame is nearly invisible in daylight, and it needs alcohol-resistant foam because ordinary foam dissolves in it. The real weakness is supply, since certified green methanol output runs under 1 Mt a year against a fleet that burns hundreds of millions of tonnes of fuel, and a long-term offtake costs $50–80/GJ when fuel oil costs about $14. The failure mode is specific and has already happened to owners, which is that the ship is delivered, the green offtake was never signed, and the ship burns fossil methanol or diesel, so the emissions cut is zero and the premium bought an option instead of a reduction.

Pick methanol when you can name the producer, the plant and the annual tonnes, and when a cargo customer will pay a premium on the freight for the certified cut. That combination realistically means a liner route with fixed calls, which is why nearly all methanol tonnage on order is container ships, car carriers and short-sea ro-ro. If you cannot contract green molecules on the route, do not pay the premium for a methanol ship, because a dual-fuel engine burning gas-derived methanol is slightly worse than the conventional ship it replaced. If the ship trades wherever the next fixture takes it, buy certified biofuel blends instead and take the fuel decision at the next newbuild. On the supply side, if you already operate a chemical berth and bunker barges in a port that liner services call at, methanol is the cheapest new-fuel business to enter, because the tankage and the barges already exist.

About 16 GJ/m³, roughly 2.4x the tank volume of fuel oil · about 30 bunkering ports · flashpoint about 12 °C · gas-derived methanol near 100 gCO2e/MJ against 91 for fuel oil · bio and e-methanol cut 60–95% · world output about 110 Mt a year, certified green under 1 Mt · green methanol $50–80/GJ against about $14 for VLSFO

Laura Maersk, the first methanol dual-fuel container ship, delivered in 2023, followed by the 16,000 TEU Ane Maersk; the first ship-to-ship methanol bunkering in Singapore in 2023, since repeated at Rotterdam, Ulsan and Shanghai; European Energy's Kassø plant in Denmark at about 42,000 t/yr of e-methanol, which is less than one large container ship's annual fuel; MAN ME-LGIM and WinGD X-DF-M engines.

Price per gigajoule is most of the argument. Gas-derived methanol runs $350–450 a tonne, which is $18–23/GJ against roughly $14 for VLSFO, and certified bio-methanol or e-methanol at $1,000–1,600 a tonne works out at $50–80/GJ, three to five times fuel oil. Volume is the other half, and the arithmetic is worth doing once: a 16,000 TEU ship burning 100 tonnes of fuel oil a day for 280 sea days uses 28,000 tonnes a year, or 1.13 PJ, and at 19.9 GJ a tonne that is about 57,000 tonnes of methanol, so fewer than 20 such ships would consume the entire world's certified green methanol output. That figure is derived here rather than published, and it explains why the order book outran the fuel. Announced project capacity is many times the capacity that has taken a final investment decision, and Maersk signed letters of intent for hundreds of thousands of tonnes a year that mostly did not become plants, then began ordering LNG dual-fuel ships alongside its methanol ones. The ship-side capital is the small number here: a methanol fuel system costs well under an LNG one because the tanks are structural steel, and a two-stroke conversion runs $10–20M and about three months of yard time. What pays for the fuel premium is a green freight product sold to a cargo owner with its own Scope 3 target, and the volume of freight sold that way is still a few percent of the trade.

3 minMethanol as a Marine FuelMethanol Institute · 10k+ views

5 minWhat is green methanol? | BBC NewsBBC News · 10k+ views

Marine Methanol: Future-Proof Shipping Fuel (Methanol Institute) · Innovation Outlook: Renewable Methanol (IRENA and Methanol Institute)

Ammonia is liquid at -33 °C at atmospheric pressure, or at about 8–10 bar at ambient temperature, and at 682 kg/m³ and 18.6 MJ/kg it holds roughly 13 GJ per cubic meter, about three times the tank volume of fuel oil for the same energy. The molecule contains no carbon, so burning it produces no CO2; what it does produce is nitrogen oxides, some nitrous oxide, and unburned ammonia. The commodity itself exists at industrial scale, with roughly 180–190 Mt made a year, about 20 Mt of it moving by sea, and something like 200 terminals already able to load and discharge it. Bunkering is what does not exist: there is no commercial supply network for ammonia as a marine fuel, only a handful of pilot ship-to-ship transfers. Toxicity sets everything else, because ammonia is detectable by smell at a few parts per million, immediately dangerous to life or health at 300 ppm, and lethal at a few thousand. An ammonia-fueled ship therefore carries double-walled fuel piping with a ventilated or inerted annulus, gas-safe machinery spaces, water curtains and deluge, a release capture system that scrubs escaping vapor into water, catch trays under the joints, detection throughout, escape sets for every crew member and a designated safe haven. The IGF Code does not yet cover ammonia, so ships are approved through the SOLAS alternative-design route, and the IMO approved interim guidelines for ships using ammonia as fuel in December 2024.

Green ammonia cuts 85–95% well-to-wake before slip is counted, and it is the only near-zero marine fuel whose production and shipping industry already runs at hundreds of millions of tonnes a year. Storage is also far more practical than hydrogen, at -33 °C and 13 GJ/m³ against -253 °C and 8.5. The weaknesses start with nitrous oxide, which warms roughly 270 times as much as CO2 per tonne, so converting even a small share of the fuel nitrogen to N2O takes back much of the reduction; unburned ammonia slip is separately a pollutant and a fine-particulate precursor. Most ammonia made today comes from steam methane reforming, which puts conventional ammonia above fuel oil on a well-to-wake basis, so bunkering it would raise a ship's emissions rather than lower them. And even conventional ammonia costs $22–32 per gigajoule against about $14 for fuel oil, so the fuel bill roughly doubles before any green premium.

Order ammonia when the ship is a newbuild on a dedicated trade with a producer or an ammonia terminal at one end: an ammonia carrier burning its own cargo, a gas carrier, or a bulker on a fixed run to a plant. If the rotation calls at ports that have not yet decided whether to permit ammonia bunkering, assume the ship burns diesel most of the time and price the capability as an option rather than as a fuel plan. Do not specify it for a tramp trader under any circumstances, because there is nowhere to buy the fuel and no prospect of that changing in the ship's first decade. Budget crew training at $2,000–6,000 per seafarer across two or three crews, and expect training capacity and port permits to set the schedule rather than engine delivery. If the requirement is a deep cut this decade on a liner route, methanol is the practical alternative; if it is a cut this year, certified biofuel blends are the only option.

About 13 GJ/m³, roughly 3x the tank volume of fuel oil · liquid at -33 °C or about 10 bar at ambient · smell at a few ppm, immediately dangerous to life or health at 300 ppm · roughly 180–190 Mt made a year, about 20 Mt traded by sea · no commercial bunkering, pilot transfers only · N2O warms about 270 times as much as CO2 · conventional ammonia $22–32/GJ, green $38–65/GJ

Fortescue Green Pioneer, the first ammonia dual-fuel vessel, which ran trials in Singapore in 2024; the Global Centre for Maritime Decarbonisation's ammonia bunkering safety studies and pilot transfers in Singapore and Western Australia; MAN B&W ME-LGIA and WinGD X-DF-A engines; NEOM's 1.2 Mt/yr green ammonia plant in Saudi Arabia; the IMO interim guidelines for ammonia-fueled ships, approved in December 2024.

Ammonia is the cheapest way to ship hydrogen and still an expensive marine fuel. Conventional ammonia at $400–600 a tonne works out at $22–32/GJ and certified green ammonia at $700–1,200 a tonne at $38–65/GJ, against about $14 for VLSFO, so the fuel bill runs two to five times conventional. On the shore side the terminals mostly exist, so the marginal cost of adding bunkering is a barge, a set of procedures and a port permit rather than a new $100–300M import terminal. What is not marginal is the exclusion zone: the hazardous zone around an LNG bunker manifold is tens of meters, while a worst-case ammonia release can carry toxic concentrations hundreds of meters downwind, which in practice means a dedicated berth and no simultaneous cargo work, and that is a throughput cost a busy terminal notices. On the ship side ammonia dual-fuel adds roughly $20–30M to a large newbuild, above the methanol premium, because of the pressure tanks, the release capture system and exhaust treatment for N2O and ammonia slip. Almost none of that is being bought on the open market. The orders that exist are gas and ammonia carriers on long charters to producers and bulkers chartered to miners with their own targets, which is the one structure where the company buying the fuel also holds the emissions target.

7 minGCMD Ammonia Bunkering Pilot (with English subtitles)Global Centre for Maritime Decarbonisation · 1k+ views

6 minGreen Pioneer: The World's First Dual-Fuel Ammonia ShipRE:TV · 5k+ views

Study Investigating the Safety of Ammonia as Fuel on Ships (European Maritime Safety Agency) · Comparison of ammonia with methanol, liquefied natural gas and conventional marine transportation fuels through life cycle cost and emissions analysis (Energy Conversion and Management)

Marine biofuel is mostly fatty acid methyl ester, made by transesterifying used cooking oil and other waste fats, plus a smaller volume of hydrotreated vegetable oil, which is chemically a paraffin and behaves like distillate diesel. It is sold as a blend named for its bio share, so B24 is 24% bio and 76% conventional fuel; FAME holds about 37 MJ/kg against 40.5 for VLSFO, so a blend gives slightly less energy per tonne. Nothing on the ship changes, which is the whole point: no engine modification, no new tank, no class approval on most engines, and therefore the only fuel on this sheet that can cut a trading ship's emissions this year. B24 is a carriage rule rather than an engine limit, because under the IMO's MSC-MEPC.2/Circ.17 a blend holding 25% or more of a MARPOL Annex II substance has to be carried as Annex II cargo, so bunker barges stop at 24% and stay under Annex I. Singapore now supplies up to B30 routinely and higher blends up to B100 with flag approval, and ISO 8217:2024 added grades for distillate and residual fuels blended with FAME and paraffinic diesel. The cut is proportional to the bio share and to the certified pathway: B24 from used cooking oil cuts roughly 20% well-to-wake and B100 FAME or HVO cuts 65–90%. Something like 20–30 ports supply it in volume, led by Singapore and Rotterdam.

Because the cut scales with the bio share, a small blend buys a lot of compliance. Moving a ship from the 91.16 gCO2e/MJ FuelEU baseline to the 2025 limit of 89.34 takes only about 2.4% of its energy from a fuel at roughly 15 gCO2e/MJ, since 1.82 divided by 76.16 is 0.024, and that is a derived figure rather than a published one. The weakness is feedstock: waste lipids are capped, road diesel and aviation fuel bid for the same molecules with mandates behind them, and shipping burns roughly 250–300 Mt of fuel a year against a global waste-oil supply measured in single-digit millions of tonnes. FAME also brings real handling problems, since it oxidizes in storage over about six months, grows microbes wherever water collects in a tank, gels at low temperature, attacks some elastomers, and acts as a solvent, so the first high-blend stem lifts old sludge off the tank walls and blocks filters. And the emissions cut lives entirely in the certificate, which EU investigations into mislabeled waste-based imports have shown can be wrong.

If you need a measured cut this year with no capital and no yard time, start here, because nothing else on this sheet can do it. If the ship trades in and out of the EU, buy the blend for the legs the regulation counts, since FuelEU is calculated on energy used on board and a few percent of bio covers the 2025 and 2030 limits, with pooling letting one ship's surplus serve a fleet. If the plan is a full pathway to 2050, do not build it on biofuel, because the feedstock is not there and the price rises as the mandated sectors take more of it. Before the first B30 or higher stem, clean the tanks, change the filters, get the engine maker's letter and confirm the flag approval, and expect a filter problem on the first voyage regardless. Read the certified intensity on the bunker delivery note rather than the blend name, because a B24 backed by a bad certificate cuts nothing.

B24 cuts about 20% well-to-wake, B100 FAME or HVO 65–90% · 25% bio content triggers MARPOL Annex II carriage, which is why barges stop at B24 · Singapore supplies to B30 routinely and to B100 with approval · FAME 37 MJ/kg against 40.5 for VLSFO · neat biofuel 1.5–2.5x fuel oil per gigajoule · roughly 20–30 supply ports · about 2.4% bio energy meets the 2025 FuelEU limit (derived)

Singapore and Rotterdam as the two main supply points; the MPA biofuel bunkering framework and Singapore's supplementary standard WA 2:2022; ISO 8217:2024, which added the blended grades; MSC-MEPC.2/Circ.17, the carriage rule behind B24; GoodFuels, FincoEnergies and TotalEnergies among the suppliers.

This is all operating cost and no capital, which changes who gets to decide. A charterer buying the bunkers can start next month without asking the owner for anything, the opposite of every engine decision on this sheet. Neat FAME or HVO runs 1.5–2.5 times fuel oil per gigajoule, so a B24 blend typically costs $100–200 a tonne over VLSFO, and on a ship burning 25,000 tonnes a year that is $2.5–5M every year against a scrubber that costs $2–6M once. The useful comparison is cost per tonne of CO2 avoided: a tonne of fuel oil burns to about 3.1 tonnes of CO2, B24 cuts roughly 20% of that, so 0.62 tonnes are avoided for a $150 premium, or about $240 a tonne. That figure is derived here, and it sits well above the EU ETS allowance price of €70–80 and far below the FuelEU penalty of €2,400 per tonne of fuel-oil-equivalent deficit, which is why FuelEU rather than the ETS is what actually moves owners to buy blends. On the supply side, blenders buy waste oils in a market where road-fuel and aviation mandates set the price, so marine demand is the marginal unmandated buyer, paying the top of the range in a tight year and picking up discounts in a loose one.

12 minIntroduction to Biofuel as a fuel for shippingLloyd's Register · 100k+ views

Biofuel Bunkering (Maritime and Port Authority of Singapore) · Final Report - Safe Bunkering of Biofuels (European Maritime Safety Agency)

A ship's resistance splits into skin friction over the wetted surface, wave-making at the free surface, viscous pressure drag set by the hull's shape, and a small air-drag term. On a slow full-form bulker or tanker, friction is 60–80% of the total; on a fast container ship, wave-making takes a much larger share. Hull form optimization is the design work that reshapes the forebody waterlines, the bulbous bow, the midship section and the aft body ahead of the propeller to cut those components at one chosen speed and draft. Current practice runs hundreds of parametric hull variants through RANS CFD and then confirms the best few in a towing tank at model scale. A bulbous bow generates a wave out of phase with the bow wave and partly cancels it, which is why it helps only near the speed and draft it was tuned for. The biggest lever is the design speed itself: a hull drawn for 19 knots is a different hull from one drawn for 23, and the IMO's design index, EEDI, now pushes buyers toward the lower number.

This is the cheapest efficiency measure per percent saved, because the cost is design hours rather than hardware, and a better hull draws no parasitic power, needs no maintenance and keeps working for the ship's whole 25-year life. Typical gains are 2–8% against an already competent baseline, and 10–20% when the comparison is a hull drawn for a service speed the ship never runs. The weakness is that it reaches newbuilds only, and it is optimized for one speed and draft pair. The 2009 slowdown is the standing example: a fleet designed for 24–25 knots ended up trading at 17–19, where the bulbous bow adds resistance instead of canceling it, and bow replacements recovered 2–5% on ships that had lost it. The design point also has to be picked three or four years before anyone knows the ship's real trading pattern, and yards quote a standard hull unless the buyer pays for the tank program and the extra engineering.

Settle the operating profile before the lines plan. If the ship will spend most of its life at 15–17 knots, specify the hull and the bulb for that, and accept that it will be slow when freight rates spike. If the charter market for the segment rewards speed, buy the faster hull and expect to lose a few percent whenever the market makes you slow down. Ask the yard which tank tested the hull, at which drafts, and whether the CFD was validated against those tests, because a CFD-only optimization at model Reynolds number routinely overstates full-scale gains. For a ship already trading, this entry does not apply: use energy-saving devices, coatings and voyage optimization instead, and consider a bow replacement only if the ship's service speed has dropped several knots below its design point and it has ten or more years left.

Friction 60–80% of resistance on full-form hulls · 2–8% from hull form work against a good baseline · 10–20% from designing for the real service speed · bulbous bow replacement 2–5% · EEDI Phase 3 requires 30% below the 2008 baseline for most types, up to 50% for the largest container ships · 5% on a 15,000 TEU ship burning 150 t/day is roughly $1M a year at $550/t, derived here from those figures

Maersk's Triple-E class, drawn around a 19-knot design speed with a twin-skeg aft body instead of the 23–25 knots the previous generation was built for; Ulstein's X-BOW on offshore and expedition vessels; bulbous bow replacements carried out across container fleets after service speeds fell; the KRISO Container Ship hull, the open benchmark most published ship CFD is validated against; MARIN, HSVA and SSPA, the commercial towing tanks that run most merchant model programs.

The money goes into engineering hours, CFD licenses and a tank program, which together run a few hundred thousand dollars on a large newbuild against a $150–200M ship. That is small enough that the real cost is schedule: a bespoke hull adds months to the design phase and yards booked into 2028–2029 charge for the slot as well as the steel. The owner pays, and on a time charter the charterer collects the fuel saving, which is why standard yard designs still win most orders in tramp segments. The payback arithmetic is easy to check. A 15,000 TEU ship burns roughly 150 tonnes a day at sea, so 5% is 7.5 tonnes a day; at $550 a tonne that is about $4,100 a day, and over 250 sea days about $1M a year, which is derived here rather than quoted. Against that, EEDI compliance is not optional, so part of this work is a cost of being allowed to build the ship at all rather than a discretionary investment.

4 minWhat is the BULBOUS BOW for?Casual Navigation · 1m+ views

8 minThe Limits of Bulbous BowsDMS | Marine Consultant · 100k+ views

Basic principles of ship propulsion (MAN Energy Solutions) · Hull Lines Reliability-Based Optimisation Design for Minimum EEDI (Brodogradnja)

Air lubrication blows air out through the flat bottom of a hull so that a layer of bubbles or a continuous air carpet sits between the steel and the water, cutting skin friction over the area it covers. Blowers or low-pressure compressors feed release units set into the bottom plating, and the air is dragged aft by the flow to form the layer. The physics is straightforward: friction is 60–80% of a full-form ship's resistance, and air is roughly 50 times less viscous than seawater, so displacing water from part of the wetted surface removes part of the friction. The catch is the compressor work, which is set by the hydrostatic pressure at the release depth, about 1 bar gauge at 10 meters of draft, so a deeper-loaded ship costs more to lubricate and gets less back. Two families exist: air lubrication systems that release a constant stream of bubbles onto a flat bottom, and air cavity ships, which recess shaped cavities into the bottom and hold a standing pocket of air in each. The first needs almost no structural change and is what nearly every retrofit uses.

It is one of the few measures that can be fitted to a ship already trading, it works in any weather and on any route, and unlike wind it does not depend on where the ship goes. The savings claim to watch is the gross figure. Vendors quote 5–10%, measured as the reduction in propulsion power, and the blowers consume roughly a third of that, so the net saving on the ships it suits is usually 4–8%. Mitsubishi's sea trials on the 92,000 dwt bulker Soyo showed the shape of the problem clearly: 8.1% at a 6.6 meter ballast draft, 4.4% at 8.8 meters, and on the maiden voyage 5% in ballast against 3% loaded. The other limit is hull shape. The system needs a wide flat bottom for the air to stay under, so it suits LNG carriers, large container ships, cruise ships and car carriers, and it does very little on a fine V-sectioned or hard-chined hull where the air escapes up the sides within a few meters of the release point.

Take it seriously if the ship has a broad flat bottom, burns more than about 100 tonnes a day, and is going into drydock anyway. If the hull is fine-formed, or the ship burns under about 40 tonnes a day, the arithmetic does not work at $1.5–3.5M installed and you should spend the money on coatings and energy-saving devices first. Always ask for the net figure measured with the blower load subtracted, at the drafts the ship actually trades at, and preferably from a sea trial rather than a CFD study. If the ship spends most of its life deeply loaded, discount the vendor number, because both the compressor work and the escaping air get worse with draft. On a newbuild the release units and ducting can be designed into the bottom structure, which is cheaper and cleaner than cutting a trading ship's shell plating.

Friction 60–80% of a full-form ship's resistance · gross saving 5–10% claimed, 3–13% in route-based modeling · blower power roughly a third of the gross saving · net 4–8% on suitable hulls · Soyo sea trials 8.1% at 6.6 m draft against 4.4% at 8.8 m · about 1 bar of compressor head at 10 m draft · $1.5–3.5M installed · payback around 2–3 years on a ship burning 150 t/day and 12–20 years at 30 t/day, derived here from those figures

Silverstream Technologies' system, fitted across cruise, LNG carrier and large container ship fleets; Mitsubishi Heavy Industries' MALS, trialed on the module carrier Yamatai and the bulker Soyo; Samsung Heavy Industries' SAVER Air; Armada Technologies and Marine Performance Systems on the fluidic-release side.

Installed cost is $1.5–3.5M depending on ship size and whether it goes in at a newbuild or in a drydock, and the recurring cost is the blower power, which is why the net number matters more than any other figure here. The payback is entirely a function of how much fuel the ship burns. A large container ship burning 150 tonnes a day saves 7.5 tonnes at 5% net, about $4,100 a day at $550 a tonne and roughly $1M a year over 250 sea days, so a $2.5M system pays back in two to three years. The same system on a handysize bulker burning 30 tonnes a day saves 1.5 tonnes, about $825 a day and $165k a year, and pays back in twelve to twenty years, which is why nobody fits one. That arithmetic is derived here rather than quoted, and it explains the order book: the installed base is concentrated in cruise ships, LNG carriers and 15,000 TEU-class boxships. EU ETS and FuelEU improve it, since a tonne of fuel oil not burned also avoids about 3.1 tonnes of CO2 and roughly €220 of allowances at €70 an allowance.

8 minIt Sounds Crazy, But It Actually Works!Casual Navigation · 100k+ views

3 minAir lubrication for shipping - Introducing the Silverstream® SystemSilverstream Technologies · 100k+ views

Air lubrication (Wärtsilä Encyclopedia of Marine and Energy Technology) · Current state and prospects on applications of ship drag reduction using air lubrication (Chinese Journal of Ship Research)

A propeller leaves its slipstream rotating, and that swirl is delivered power that never became thrust, typically several percent of it on a single-screw merchant ship. Energy-saving devices are fixed appendages placed around the propeller to take some of it back. They fall into three positions: ahead of the disc, where ducts and pre-swirl stators straighten and pre-rotate the inflow against the propeller's direction; at the hub, where propeller boss cap fins break up the hub vortex; and behind it, where rudder bulbs, thrust fins and twisted rudder profiles recover the rotation that is left. Ducts do a second job on full-form hulls, evening out a wake field that is much slower directly behind the hull than at the edges of the disc, which raises propeller efficiency and cuts cavitation and stern vibration at the same time. None of them has moving parts or draws power, and a typical unit is a few tonnes of fabricated steel welded on during a scheduled drydock.

Per percent of fuel saved this is the cheapest capital on a ship: $100–500k fitted, 2–6% typical, no maintenance beyond inspection, and no parasitic load. It also reaches ships already trading, since the work is a few extra days at a docking the ship was going to take anyway. The honest weakness is measurement. Most quoted savings come from model tank tests, and the model-scale boundary layer is thick relative to the hull in a way the full-scale one is not, so a pre-swirl device usually looks better at model scale than in service; the ITTC publishes a scaling procedure precisely because this gap is systematic rather than random. The second weakness is that these devices only pay where there is swirl and wake non-uniformity to work on, meaning slow single-screw ships with high block coefficients. On a fast fine-form or twin-screw hull the appendage mostly adds drag, and a badly matched duct can move the cavitation pattern and start eroding the propeller or shaking the stern.

If the ship is a single-screw bulker or tanker with a block coefficient above roughly 0.8 running at 12–15 knots, fit a pre-swirl duct at the next docking and treat it as close to a default. If it is a container ship, boss cap fins plus a rudder bulb are the usual pairing, with smaller gains because the wake is already more uniform. If the ship is twin-screw, fast, or fine-formed, skip the appendages and spend the money on coatings and voyage optimization instead. Ask the supplier for full-scale measured results on a sister ship, at a stated draft and speed, and ask who carries the risk if the saving does not appear, because several suppliers will guarantee a number and most will not. Fit it in the same docking as the coating renewal so the yard time is paid for once.

2–6% typical saving, up to about 8% on full-form slow ships · $100–500k installed · fitted in 1–3 extra days at a scheduled drydock · works best above roughly 0.8 block coefficient at 12–15 knots · over 3,000 sets of boss cap fins and over 1,000 Mewis Ducts delivered · about $194k a year on a Capesize burning 40 t/day at 4%, derived here from those figures

Becker Marine Systems' Mewis Duct, with over 1,000 units delivered to bulkers and tankers; MOL and Nakashima's Propeller Boss Cap Fins, over 3,000 sets since 1987; Schneekluth's wake equalizing duct; Wärtsilä's EnergoFlow and EnergoProFin; Kawasaki's RBS-F rudder bulb with fins; the gate rudder first fitted to the Japanese coastal container ship Shigenobu in 2017.

The bill is one purchase order and a few days of yard time: $100–500k for the device, fabrication and welding, against a drydock the ship was taking anyway. Payback is short wherever the device suits the hull. A Capesize bulker burning 40 tonnes a day saves 1.6 tonnes at 4%, about $880 a day at $550 a tonne and roughly $194k a year over 220 sea days, so a $300k duct pays back in under two years; that arithmetic is derived here rather than quoted. On a time charter the owner buys the device and the charterer buys the bunkers, which normally kills a retrofit, and this one survives it because a better EEXI and CII rating also affects what a broker can fix the ship for and what a buyer will pay at resale. Suppliers compete on guaranteed savings backed by CFD plus a tank test, so the commercial risk sits partly with them, and the margin is in engineering the device to a specific hull rather than in the steel.

7 minBecker Mewis Duct® Twisted installation (retrofit)Becker Marine Systems · 100k+ views

4 minPBCF (Propeller Boss Cap Fins) -Best-selling Energy-Saving Device-MOL Official Channel (English) · 5k+ views

EETs for Ships – Guide and High Level Assessment Tool (IMO GreenVoyage2050) · Feasibility Study on Effect of Structural Flexibility of Asymmetric Pre-Swirl Stator on Propulsion Performance for KRISO Container Ship (KCS) (Brodogradnja)

A hull starts fouling within days of leaving drydock: a bacterial film, then slime, then weed, barnacles and tubeworms if the ship sits still in warm water. Roughness costs power because it thickens the turbulent boundary layer, and a slime film alone adds roughly 10% to the shaft power needed for the same speed, with heavy calcareous growth reaching 30–60%. A fouled ship in normal service is usually somewhere in the 10–30% band, which is more fuel than almost any retrofit on this sheet gives back. Three coating families are used against it: self-polishing copolymers, whose binder hydrolyzes and wears away to keep exposing fresh copper and booster biocide; fouling-release silicones and fluoropolymers, which carry no biocide and rely on a surface too slippery for organisms to key onto, so growth sheds once the ship is moving; and hard inert epoxies that assume regular mechanical cleaning. The underwater area of a large ship is 15,000–25,000 square meters, so the coating is a real purchase, and the propeller matters out of proportion to its area because roughness there acts on a much faster flow.

This is the cheapest fuel saving available and the only one that reaches every ship trading today, since a coating renewal happens at a docking the ship is required to take anyway. Cleaning is cheaper still: an in-water hull clean on a large ship runs roughly $10–50k and a propeller polish $5–15k, and polishing alone is worth 1–3%. The weaknesses are operational rather than technical. Fouling-release coatings need speed and short port stays to work, so a ship that waits at anchorage for two weeks fouls anyway, and the soft silicone is easily damaged by tugs, ice and aggressive cleaning. The harder constraint now is biosecurity: cleaning releases both the organisms and paint particles carrying copper into the harbor, so several jurisdictions restrict or ban in-water cleaning without capture and filtration, and a ship can arrive somewhere needing a clean it is not allowed to have.

Treat coating and cleaning as the first thing you buy, before energy-saving devices and long before air lubrication or wind. Match the coating to the operating profile: fouling-release for liner ships that run at steady speed with short calls, self-polishing copolymer for tramp trades with long idle periods, anchorage waiting or slow speeds. Measure rather than assume, using hull performance monitoring to ISO 19030, so the cleaning decision comes from a rising power-for-speed trend rather than a calendar. Check the biosecurity rules at the ports in the actual rotation before scheduling anything, because New Zealand, California and Australia all impose biofouling requirements and many ports allow in-water cleaning only with a capture system. If the ship is going to sit idle for weeks, plan a clean before it trades again rather than after the fuel bill shows the problem.

Slime film about 10% extra shaft power, heavy calcareous fouling 30–60%, 10–30% typical in service · underwater area 15,000–25,000 m² on a large ship · premium coating scheme $200–500k above a basic one · in-water hull clean $10–50k, propeller polish $5–15k · propeller polishing alone 1–3% · California's rules apply to vessels of 300 gross registered tons and up · about $145k a year from 3% on a ship burning 40 t/day, derived here from those figures

AkzoNobel's Intersleek fouling-release range, Jotun's SeaQuantum copolymers and its Hull Skating Solutions robot that lives aboard and grooms the hull between calls, Hempel and Chugoku Marine Paints; capture-and-filter cleaning contractors such as ECOsubsea and Fleet Cleaner; New Zealand's Craft Risk Management Standard for biofouling, which requires a clean hull on arrival; California's Marine Invasive Species Program; the IMO's 2023 biofouling guidelines, which are recommendatory rather than mandatory; ISO 19030 for measuring hull and propeller performance.

A premium coating scheme costs $200–500k more than a basic one and lasts a five-year docking cycle, so about $40–100k a year, against a saving of 3–8% on the ship's whole fuel bill. On a ship burning 40 tonnes a day, 3% is 1.2 tonnes, about $660 a day at $550 a tonne and roughly $145k a year over 220 sea days, which is derived here rather than quoted and is several times the annual cost of the coating. Cleaning is even better value when the hull is actually fouled, since $30k of diver time can recover 5–15% for the rest of the docking interval. The split is the usual one: the owner pays for the drydock and the paint while the charterer buys the bunkers, and BIMCO's hull fouling clause exists to settle who pays when a long idle period at the charterer's orders leaves the ship fouled. Coating suppliers sell against guaranteed performance rather than price per liter, and the margin sits with the four large marine coatings houses rather than with the applicators. The regulatory direction adds value on top: a cleaner hull moves a CII rating in the same year it is done, which no fuel-switching project can match.

9 minShip Painting Preparation in Dry Dock | Washing & Sand Blasting Explained | Ep#235Josemon's Clicks · 10k+ views

2 minRemora Hull Cleaning | Underwater Robotic SystemsRemora Technology · 10k+ views

Marine Invasive Species Program (California State Lands Commission) · Enhancing Ship Energy Efficiency and Preventing Pollution through Effective Biofouling Control Measures as a Future Direction for a Sustainable Shipping (TransNav)

Four devices are being fitted to merchant ships, and they all do the same job: produce thrust from the apparent wind so the main engine can be throttled back. Rotor sails are spinning cylinders 18–35 meters tall and 3–5 meters across, driven by a small electric motor drawing tens of kilowatts, which develop side force by the Magnus effect and resolve part of it forward. Suction wings are fixed aerofoil sections with a fan pulling air through perforations to hold the boundary layer attached, which gets a high lift coefficient out of a much smaller structure than a conventional sail. Rigid wing sails are hinged aerofoils up to about 37.5 meters tall that fold to the deck for port entry and air-draft limits. Kites are parafoils flown on a tether 200–300 meters up, where the wind is stronger and steadier and where the pull is nearly horizontal, with no tall structure on deck at all. All four work best with the apparent wind on the beam or quarter, and none of them produce useful thrust dead upwind or directly astern at merchant ship speeds, which is why the route matters more than the hardware.

The fuel is free, the systems retrofit at a drydock, and they need no bunkering infrastructure and no engine change, which makes them one of the few deep cuts available to a tramp owner who cannot commit to an alternative fuel. Measured savings are real but wide: route-based modeling of ships in service found 1–12% per rotor, sea trials on the ro-ro Estraden gave 2.6% with one rotor and 6.1% with two, the four-rotor E-Ship 1 has been reported at about 15%, and a single rotor on the passenger ship Viking Grace returned 1.7%. The number depends on the wind rose along the actual trade, so a 15% result from a North Atlantic run can fall to a third of that on an intra-Asia route, and a saving quoted without naming the voyages it came from is not usable. The physical cost is deck space and air draft: rotors and wings conflict with crane swing on a bulker, with manifolds and vapor piping on a tanker, and with container stacks on a boxship, where the space lost is cargo revenue rather than convenience. Structure also has to be added into the deck to take the heeling moment, and heel and leeway themselves eat into the net gain.

Look at wind propulsion when the ship is a bulker, tanker or car carrier with clear deck, on a trade with a known and windy rotation. Get the supplier to model your own AIS track against a wind atlas rather than accepting a headline percentage, and ask which ship, which voyages and which season produced it. If the ship is a container vessel on a fixed liner route, the deck is already sold and the answer is usually no, unless the rigs go on a section of deck that cannot take boxes anyway. Kites suit ships with no spare deck and no air-draft margin; suction wings suit smaller vessels where a rotor's mass and foundation would be disproportionate. If the trade is short-sea with frequent port calls and low sea time, the payback stretches out and the money is better spent on hull coating and energy-saving devices first.

Rotor sails 18–35 m tall, 3–5 m across, motor draw tens of kW · rigid wings up to about 37.5 m · kites flown at 200–300 m · 1–12% per rotor in route-based modeling, 5–20% for a full set · Estraden 2.6% with one rotor and 6.1% with two · Viking Grace 1.7% · E-Ship 1 about 15% on four rotors · $1–3.5M per set installed · about $412k a year at 10% on a bulker burning 30 t/day, derived here from those figures

Norsepower's rotor sails on the ro-ro Estraden, the ferry Viking Grace and the LR2 tanker Maersk Pelican; Enercon's four-rotor E-Ship 1; BAR Technologies' WindWings on Cargill's Kamsarmax bulker Pyxis Ocean, reported at roughly 3 tonnes of fuel a day; bound4blue's eSAIL and Econowind's VentoFoil suction wings; Airseas' Seawing and SkySails' towing kites; Wallenius and Alfa Laval's Oceanbird wing.

A set of rigs runs $1–3.5M installed including deck reinforcement, foundations and class approval, and the marginal running cost is the few tens of kilowatts the rotors or suction fans draw. Payback follows the fuel bill and the wind together: a Kamsarmax burning 30 tonnes a day saves 3 tonnes at 10%, about $1,650 a day at $550 a tonne and roughly $412k a year over 250 sea days, so a $2M set pays back in about five years before any carbon cost, which is derived here rather than quoted. Carbon pricing shortens that materially inside Europe, because a tonne of fuel oil not burned also avoids about 3.1 tonnes of CO2 and roughly €220 of EU ETS allowances at €70 each, and FuelEU applies a wind reward factor that cuts a ship's reported intensity by up to 5% once installed wind power reaches about 15% of propulsion power. The commercial obstacle is the same one everywhere on this sheet: on a time charter the owner buys the rigs and the charterer buys the bunkers, so most installations so far are on owner-operated tonnage or under long charters to cargo owners with their own targets. Deck space is a second cost that does not show up in the capital number, since a rig on a container ship displaces slots that were sold. Fewer than one merchant ship in a thousand carries wind propulsion today, so unit costs still reflect small production runs rather than a mature supply chain.

4 minRotor Sail Explainedmarineinsight · 100k+ views

20 minThe Wind-Powered Cargo ShipsEngineering Secrets · 10k+ views

Rotors and bubbles: Route-based assessment of innovative technologies to reduce ship fuel consumption and emissions (International Council on Clean Transportation) · Impact of control strategies for wind-assisted ships on energy consumption (Brodogradnja)

This bundles two levers that need no hardware: how fast the ship runs, and how the voyage is planned around weather, trim and arrival time. Speed is by far the larger one, because resistance on a displacement hull rises roughly with the square of speed and the propulsion power to overcome it rises roughly with the cube. Work the arithmetic through and it is the most useful calculation on this sheet. Cut speed 10% and power goes to 0.9³, which is 0.729, so the ship burns about 27% less fuel per day; but the voyage now takes 1/0.9, about 11% longer, so fuel per tonne-mile falls by 1 − 0.729/0.9, about 19% rather than 27%. Those two figures are derived here from the cube law rather than quoted from a study, and the gap between them is the thing most speed-saving claims leave out. The rest of voyage optimization is smaller and additive: weather routing is worth roughly 2–4%, trim optimization 1–3%, and slowing the approach leg to hit a berth window rather than waiting at anchor is worth more again on the leg it applies to. The cube law does flatter very deep cuts, because a two-stroke's specific fuel consumption is roughly flat between about 50% and 85% of rated power and worsens below about 40%.

Nothing else on this sheet cuts 20% or more with no capital and no yard time, and it works on every ship trading today rather than on the small share that can be retrofitted. It is also the fastest lever on a CII rating, since a speed change shows up in this year's carbon intensity number while a fuel switch takes a newbuild cycle. The weaknesses are commercial rather than technical. Holding the same service frequency at a lower speed takes more ships, cargo spends longer at sea, and on a time charter the charterer directs the speed while most charterparties still oblige the master to proceed with utmost dispatch, so the owner cannot simply slow down. Very deep cuts also run into engine and machinery limits, with turbocharger and scavenge fouling on prolonged low load and periodic load-ups required to clear it.

Slow down when freight rates are weak, when bunkers are expensive relative to the value of an extra round trip, or when a CII rating has to move inside the year. Speed up when rates spike, because one extra voyage can be worth more than the fuel it costs; that trade-off, not the engineering, is what actually sets fleet speeds. Before touching speed, take the free measures: weather routing, trim optimization against the ship's own loading condition, and a berth window agreed with the port so the ship is not burning fuel to wait. If the ship is on time charter, get the speed and consumption warranty and the emissions clauses right at fixture, since the party that saves the fuel and the party that loses the time are usually different companies. Treat a limit on engine power as a separate instrument: it caps the top of the range for compliance and does nothing about how the ship is run day to day.

Power roughly proportional to the cube of speed · 10% slower gives 0.9³ = 0.729, so about 27% less fuel a day · the voyage takes about 11% longer, so fuel per tonne-mile falls about 19%, both derived here · weather routing 2–4% · trim optimization 1–3% · specific fuel consumption roughly flat from 50% to 85% of rated power · a weekly service on a 35-day round trip needs 5 ships, and 6 once the round trip stretches to 39 days · an extra ship costs $30–60k a day in hire and crew

The container fleet's shift from roughly 25 knots before 2008 to 17–19 knots afterward, and Maersk's super-slow steaming program from 2009; engine power limitation, fitted across most of the existing fleet to comply with the IMO's EEXI rather than by changing anything physical; the IMO's CII, which rates ships on grams of CO2 per deadweight-mile and responds to speed within a single reporting year; weather-routing services from StormGeo, Wärtsilä and DTN; the Blue Visby Solution for coordinating arrival times across competing ships.

Fuel is 50–70% of voyage cost at $550 a tonne, so speed is where an operator's money actually is, and the decision is commercial rather than technical. Two costs sit against the fuel saving. The first is ships: a weekly service on a 35-day round trip takes 5 vessels, and slowing 10% stretches the round trip to about 39 days and takes 6, so the operator adds a ship at $30–60k a day in hire and crew to save the fuel. The second is cargo capital, which is smaller than people assume: a $20,000 container of goods costs about $5.50 a day at a 10% cost of capital, so four extra days is roughly $22 a box, against a fuel saving on the order of $50 a box for the same ship. Both of those are derived here rather than quoted. The party that pays is decided by the charter: on a voyage charter the owner buys the bunkers and captures the saving, while on a time charter the charterer buys them and the owner carries the schedule risk. Slower steaming also absorbs fleet capacity, which supports freight rates, and that side effect is a large part of why liner operators use it deliberately in weak markets.

6 minWhat does "Slow Steaming" mean?Casual Navigation · 1m+ views

5 minSlow steamingMaersk · 10k+ views

Slow steaming (Wärtsilä Encyclopedia of Marine and Energy Technology) · Regulating speed: a short-term measure to reduce maritime GHG emissions (CE Delft)

An automated container terminal takes the drivers out of the yard. Rail-mounted automated stacking cranes work blocks of containers laid perpendicular to the quay, automated guided vehicles or automated straddle carriers shuttle boxes between the quay and the ends of those blocks, and a terminal operating system decides where every container sits and in what order it is dug out. The ship-to-shore cranes are usually still driven, increasingly by an operator at a desk ashore who handles the fine positioning while the software does the long travel moves. The industry classes a terminal as semi-automated when only the stacking yard runs itself, fully automated when horizontal transfer does too, and completely automated when the quay cranes are included. As of 2025 about 76 major container terminals worldwide were fully or partly automated, roughly 8.3% of the total, and in the United States six terminals out of 84, or 7.1%. The first was ECT Delta in Rotterdam in 1993, so this is a 30-year-old technology that has still not reached one terminal in ten.

The three things automation reliably delivers are density, safety, and consistency. Taking people out of the stack allows one-over-four and one-over-five stacking, which raises yard capacity from about 700 TEU per hectare with straddle carriers to 1,250–1,400 TEU per hectare with automated stacking cranes, and a yard nobody walks in has far fewer injuries. Labor per quay crane falls roughly 50–60%. What automation has not reliably delivered is a higher peak rate: operators expected 25–50% lower operating cost and up to 30% more productivity, and those expectations have gone largely unmet, because an automated machine runs a fixed cycle time and cannot be pushed the way a good gang can be pushed on a good day. The other weakness is that converting a working terminal costs throughput while it happens, and the conversion period is longer than building a conventional terminal from scratch.

Automate a greenfield terminal if labor costs $100,000 a year per worker or more, if the site cannot be expanded, or if the terminal is a transshipment hub that has to work around the clock. If labor runs $20,000–40,000 a year, as it does in Korea and Taiwan, the labor saving will not carry the capital and you should buy electrified rubber-tired gantries instead. Do not plan a conversion of a working terminal without agreement from the union first, because most port labor contracts specify which functions union labor performs and how many people a terminal must employ, and the technology is rarely the thing that stops the project. If the constraint you actually have is truck queues rather than yard capacity, buy gate automation and an appointment system, which cost a fraction of the equipment and can be installed without touching the yard. Whatever you decide, size the electrical supply for the finished terminal at the start, since every piece of automated equipment is electric.

About 76 automated terminals worldwide in 2025, roughly 8.3% of the total · six of 84 US terminals, 7.1% · yard density about 700 TEU/ha with straddle carriers against 1,250–1,400 TEU/ha with automated stacking cranes · labor per quay crane down 50–60% · expected 25–50% operating cost cut and up to 30% productivity gain, largely unmet · yard productivity around 80,000 TEU per man-year automated against 17,000 conventional · global average 23.5 container moves per gross crane hour.

ECT Delta in Rotterdam, automated from 1993 and the first of its kind; Hamburg CTA, Rotterdam World Gateway and APM Terminals Maasvlakte II; Long Beach Container Terminal and TraPac in Los Angeles, converted in 2016 and 2014; Virginia International Gateway, the first US automated terminal in 2007; Yangshan Phase IV in Shanghai, which ranked first in the World Bank's Container Port Performance Index 2023; Patrick's automated straddle carrier terminal in Brisbane; PSA's Tuas terminal in Singapore.

Automation is a labor-cost arbitrage, and the arithmetic only works where labor is expensive. PEMA's benchmarks put yard operations at roughly 80,000 TEU per man-year at Hamburg CTA and Rotterdam Euromax, where labor runs about $100,000 per man-year, against 17,000 TEU per man-year at conventional Antwerp and 7,000 at Korean and Taiwanese terminals where labor costs $20,000–40,000. The capital side is unforgiving in a different way: an automated stacking crane costs roughly $1.1M more than an electrified rubber-tired gantry, the equipment cannot be bought in increments because it only works as an integrated system, and delivery runs 15–24 months before the civil works are counted. That combination makes the decision irreversible in a way a fleet of straddle carriers is not, which is most of why terminal operators have been slow. The saving that does show up consistently is variance rather than average: an automated terminal delivers roughly the same number of moves at 3 a.m. as at noon, which is worth real money to a liner operator buying schedule reliability. And the binding constraint in North America and much of Europe is the labor agreement, not the equipment, which is why the International Longshoremen's Association shut East and Gulf Coast ports in October 2024 and made automation the central issue of the contract settled in early 2025.

2 minThe Automation Effect At Ports Of LA And Long BeachCBS LA · 100k+ views

4 minLong Beach Container Terminal – Safer, greener and more productiveabbmarine · 100k+ views

Chapter 6.6 – Container Terminal Automation (Port Economics, Management and Policy) · Container Terminal Automation (PEMA)

A container terminal is a chain of machines and its throughput is set by the slowest link. At the quay, ship-to-shore cranes (portainers) span the ship's beam, reaching 18 containers wide for post-Panamax ships and up to 24 for the 24,000 TEU class, with spreaders that twist-lock onto the corner castings and lift more than 150 tonnes on the newest cranes when doing a tandem lift of four boxes at once. Between quay and stack, terminal tractors pull chassis or bomb carts, or straddle carriers pick the container up and drive over the stack with it. In the yard, rubber-tired gantries roll on tires and can be moved between blocks, while rail-mounted gantries are fixed but span wider, and reach stackers handle odd moves such as reefers and rail. Dry bulk terminals use a different set: grab or continuous unloaders discharging into a quay conveyor, stacker-reclaimers building and cutting the stockpile, and shiploaders on the export side.

The equipment is mature, financeable, and has a 25-year life, and the choice between types is mostly a trade of stacking density against flexibility. Straddle carriers do everything and need no second machine, at a yard density of 500–700 TEU per hectare; rubber-tired gantries stack one over four or five and reach about 1,000 TEU per hectare; rail-mounted gantries go above that but fix the yard layout in concrete. The weakness at the quay is that crane speed has stopped improving: the global average is 23.5 container moves per gross crane hour, and World Bank data shows that number barely varies with ship size, call size, or how many cranes are on the ship. What determines how long a ship sits alongside is crane intensity, the number of cranes actually worked, which runs about 2.6 on small calls and 4.7 on the largest. Diesel is the other weakness, since rubber-tired gantries can be half of a container terminal's diesel consumption.

Pick straddle carriers if the terminal is under roughly a million TEU a year, the layout may change, and land is not scarce, because one machine type covers quay transfer, stacking, and truck service. Pick rubber-tired gantries when yard density matters and you still need to move blocks around, and electrify them with a conductor rail or cable reel at the same time, since the fuel saving is the easiest money in the yard. Pick rail-mounted gantries when the layout is settled, the volume justifies it, and you may automate later, because an automated stacking crane is a rail-mounted gantry with the driver removed. For the quay, do not buy speed, buy reach and a second crane: adding a crane to the call cuts ship time far more reliably than a faster crane does. And size the berth for the ship you will get in 15 years, since the largest boxships need about 400 m of length and a 450 m berth.

Ship-to-shore cranes span 18–24 containers and lift over 150 tonnes on a tandem move · global average 23.5 container moves per gross crane hour · crane intensity 2.6–4.7 cranes per call, and it drives ship time more than crane speed · straddle carriers 500–700 TEU/ha, rubber-tired gantries about 1,000 TEU/ha, rail-mounted gantries above that · a rubber-tired gantry serves 8–9 trucks an hour, 30–40 container moves · largest container ships need roughly 400 m alongside and a 450 m berth · rubber-tired gantries can be half of a terminal's diesel use.

Portainers from ZPMC, which builds the large majority of the world's ship-to-shore cranes, plus Konecranes, Liebherr and Kalmar; Kalmar and Konecranes straddle carriers, the standard at Antwerp, Gothenburg and Brisbane; electrified rubber-tired gantries with conductor rails, first used in China in 2007 and now the majority of new and converted units; grab unloaders and stacker-reclaimers at the HES dry bulk terminal on the Maasvlakte in Rotterdam; the iron ore shiploaders at Port Hedland, the world's largest bulk export port.

Cranes are the terminal's balance sheet. A ship-to-shore crane is a multi-year, multi-million dollar order with a 25-year life, and terminals finance them the way an airline finances aircraft, so the decision is made against a concession term rather than a business cycle. Yard equipment is where the operating cost sits: an automated stacking crane costs roughly $1.1M more per unit than an electrified rubber-tired gantry, but two of them can replace three gantries and they need a fraction of the drivers, so the payback depends almost entirely on the local wage. Electrification is the one upgrade that pays back on fuel alone at most terminals, because rubber-tired gantries can be half of the diesel burned on site and grid power is cheaper per unit of work than a diesel genset. ZPMC's dominance in quay cranes has become a procurement issue as well as a price one, since US ports face pressure to buy elsewhere and the alternatives cost more and take longer. The margin in this business goes to the terminal operator rather than the equipment maker: crane builders compete on price into a lumpy order book, while the concession holder collects a per-move charge from the shipping line for 25 years.

15 minInside the Giant Cranes Loading & Unloading Massive Container Ships at US PortsNauctis · 500k+ views

9 minHow Do Port Cranes Work So Accurately? The Tech and Teamwork Behind Smooth Cargo HandlingHistory of Simple Things · 10k+ views

Chapter 6.5 – Container Terminal Design and Equipment (Port Economics, Management and Policy) · The Container Port Performance Index 2023: A Comparable Assessment of Performance Based on Vessel Time in Port (World Bank)

Shore power, also called cold ironing, shuts down a ship's auxiliary diesel generators while it is alongside and feeds the ship's switchboard from the local grid instead. High-voltage connections run at 6.6 kV or 11 kV under IEC/ISO/IEEE 80005-1, which covers anything from 1 MVA upward, and smaller craft use the low-voltage standard below that. The shore side needs a substation, a frequency converter where the grid runs at 50 Hz and the ship at 60 Hz, a cable management system on the quay, and a connection point sized for the ship; a container or ro-ro berth is typically served by a 3 MVA outlet, while the Port of Seattle built a 20 MW dual-voltage supply for a single cruise berth at Pier 66. The ship side needs a receiving switchboard, a transformer, protection and synchronization so the transfer happens without dropping the load, and on a cruise ship four cables. Emissions at the berth fall a long way: studies collected by the EPA put NOx reductions at 92–99% and CO2 at 25–49%, with the CO2 number set entirely by the local generation mix.

The local air quality benefit is large, immediate, and lands on the neighborhood next to the port, which is why regulators reached for this before industry did. It is also one of very few measures that cuts emissions from ships already trading, needing a retrofit rather than a newbuild. Two weaknesses stop it. First, the ship-side retrofit and the shore-side substation are separate purchases with separate payers, and the party that spends the money is not the party that benefits, so about 15% of container ships and 27% of cruise ships worldwide are equipped. Second, plugging in usually costs the operator more than burning gas oil: the break-even for cruise operators sits around $0.05–0.10/kWh, and real port rates run near $0.20/kWh plus demand charges of roughly $14.67 per kW of monthly peak. The grid connection is the hard part on the shore side, because the utility interconnection, not the cable reel, sets the schedule and most of the cost.

Fit shore power where the same ships call the same berth repeatedly and stay more than a few hours, which means liner container terminals, cruise berths, and ro-ro ramps. Start with the utility interconnection study rather than the equipment specification, since at San Diego the largest single cost was the utility's infrastructure, ahead of the terminal's electrical work and well ahead of the ship modifications. If the ship is a tanker, expect a capture-and-control system on the exhaust instead, because tankers run steam-driven cargo pumps and take inert gas from boiler exhaust, so the boilers keep running whether or not the generators stop. If the local grid is coal-heavy, sell this as an air quality measure and not a carbon measure, because the CO2 cut can be as low as 25%. And plan the tariff before the hardware, because a berth nobody plugs into reduces nothing.

High-voltage shore connection at 6.6 kV or 11 kV, 1 MVA and up · 3 MVA per outlet at a container or ro-ro berth, 20 MW for one cruise berth at Seattle Pier 66 · NOx down 92–99%, CO2 down 25–49% depending on the grid · roughly 15% of container ships and 27% of cruise ships equipped · operator break-even around $0.05–0.10/kWh against port rates near $0.20/kWh plus $14.67/kW demand charges · California's at-berth rule costs about $2.23 billion against $2.32 billion of health benefit.

California's At Berth Regulation, which required emission control for container, reefer and cruise calls from January 2023, ro-ro and Los Angeles and Long Beach tanker calls from January 2025, and all regulated tanker calls from January 2027; the EU's Alternative Fuels Infrastructure Regulation (EU) 2023/1804, in force since April 2024, which sets shore power targets at the larger TEN-T maritime ports, alongside FuelEU Maritime's requirement that container and passenger ships connect at berth from 2030; Berth 100 at the Port of Los Angeles West Basin, the first US container terminal to offer it in 2004; Juneau, Alaska, the first US cruise installation in 2001; Port Hueneme's six outlets across three berths.

Two costs, two payers, and no operating saving is the whole story. Port Hueneme spent $14M in 2014 on six outlets serving three berths and now needs to move its substation from 16.9 kV to 66 kV to keep going; Seattle's Pier 66 project ran about $30M for one berth and required a new 20 MW supply from the utility; the Brooklyn Cruise Terminal came to $19.3M assembled from a port authority, an EPA grant and a state development corporation. On the running cost, New York delivers power to vessels at $0.12/kWh against a delivered cost of $0.26/kWh, with the city covering the gap, and Port Hueneme reports that vessel operators see no saving from using it. As carbon abatement the numbers are poor: a study of Shenzhen put the marginal cost at about $2,300 per tonne of CO2, and California's rule is roughly break-even valued on health alone, $2.23 billion of cost against $2.32 billion of benefit. That is why mandates came before adoption rather than after it. The money to watch is the grid connection, which is a utility capital project on a utility timescale, and increasingly competes with truck charging and cranes for the same substation capacity.

13 minHOW TO CHANGE OVER SHIP POWER TO SHORE POWER - STEP BY STEP DURING DRYDOCKjoel sibulo - Life of a Seafarer · 10k+ views

Shore Power Technology Assessment at U.S. Ports (US EPA) · Ocean-Going Vessels At Berth Regulation (California Air Resources Board)

Just-in-time arrival means slowing down on the approach so the ship reaches the pilot station when the berth is actually free, instead of steaming at full speed and then waiting at anchor. The physics is trivial. Propulsion power rises roughly with the cube of speed, so fuel burned over a fixed distance rises roughly with the square, and a 10% speed cut on the approach saves about 19% of the fuel on that leg (that is arithmetic, not a measured result). A study for the IMO's Low Carbon Global Industry Alliance put full port-call optimization at about 14% of CO2 per voyage on average, optimizing only the last 24 hours at about 5.9%, and the last 12 hours at about 4.2%. What it requires is that the port tell the ship when the berth will be free, early enough for the ship to do something about it, which means a berth planning process and a standard message rather than a first-come, first-served queue. No new hardware goes on the ship.

This costs nothing to install, works on every ship already trading, and cuts anchorage crowding as well as fuel, which is a safety benefit in its own right. The weakness is entirely commercial. Under most voyage charters the owner must proceed with due or utmost dispatch and without deviation, and laytime starts when a valid Notice of Readiness is tendered, so arriving early starts the demurrage clock; demurrage per day is usually worth more than the fuel saved per day by slowing, so the rational move is to race and then wait. On a time charter the split is worse still, because the charterer buys the bunkers and the owner who slows down captures nothing. Terminals that allocate berths first-come, first-served reward exactly the behavior the measure is trying to stop. Virtual arrival clauses have been available since INTERTANKO published one in 2011 and BIMCO in 2013, and they are still barely used in oil trades and effectively absent from dry bulk.

Do it now if you run a liner service on a fixed rotation and own or long-term charter the ships, because you hold the fuel bill, the schedule, and the berth window at the same time. If you are fixing a voyage charter, get a virtual arrival or just-in-time clause into the fixture before the ship sails, because the clause is the whole project and the speed reduction is the easy part. If you run the port, the highest-value change is a credible notified arrival time issued about a week out and a berth allocation rule that does not reward early arrival, since the exporters and charterers have to trust the time enough to accept it in place of a race. Do not expect anything from a system that only pushes an ETA update six hours out; by then the ship has already burned the fuel. And if the cargo itself is not ready, fix the inland logistics first, because no amount of vessel data makes coal appear at the stockpile.

Full port-call optimization about 14% of CO2 per voyage, last 24 hours about 5.9%, last 12 hours about 4.2% · fuel over a fixed distance rises roughly with the square of speed, so a 10% cut saves about 19% on that leg (arithmetic, not a published figure) · Newcastle 2009: 11.1 days average between joining the anchorage queue and entering port · 2017–2019 under the Vessel Arrival System: 64% of arriving ships did not anchor and 3 days was the average wait for those that did · average voyage speeds down over 20% under that system · virtual arrival clauses published in 2011 and 2013 and still rarely used.

The IMO Low Carbon GIA's Just In Time Arrival Guide and the GreenVoyage2050 Just in Time portal; the Port of Newcastle Vessel Arrival System, running since 2009 after the Pasha Bulker grounded in 2007 with more than 70 ships at anchor on some days; the Hunter Valley Coal Chain Coordinator and the capacity framework the Australian competition regulator authorized in December 2009; INTERTANKO's 2011 virtual arrival clause and BIMCO's charter party clauses; the International Taskforce Port Call Optimization.

The fuel saving is real and it lands on the wrong party, which is why a measure with no capital cost has taken 15 years to go nowhere. Under an FOB coal sale the voyage charter is between the buyer and the shipowner while the exporter pays the demurrage, so three companies have to agree on terms fixed weeks or months before anyone knows the fuel price or the delay. Since a day of demurrage is normally worth more than a day of fuel, the owner who slows down is trading a certain revenue for an uncertain saving that a different company banks. Newcastle broke that loop by making berth availability predictable enough that the port's Notified Arrival Time is accepted as the Notice of Readiness, so a ship that slows down loses nothing in the queue. That took a port authority willing to lead, a competition-law clearance for the mines, railways and terminals to coordinate, and a grounded bulk carrier to force the issue. The lesson for anyone selling port-call software is that the product is a contract change with a data feed attached, and the buyer is the party that controls berth allocation rather than the party that burns the fuel.

4 minPort Call OptimisationPort of Rotterdam · 5k+ views

Just in Time Portal (GreenVoyage2050, IMO) · Reducing Anchorage in Ports: Changing Technologies, Opportunities and Challenges (Frontiers in Future Transportation)

Inland waterway transport carries cargo between a seaport and the interior on rivers and canals, using self-propelled motor vessels in Europe and pushed barge tows in North America. A newly built European dry cargo vessel carries about 2,664 tonnes and a new tanker about 4,022 tonnes, while a Mississippi tow lashes many barges to one towboat. European inland waterways moved roughly 116 billion tonne-km in 2023 and the Rhine alone carried 276.5 million tonnes, down 5.4% on 2022. In the United States about 90% of inland waterway cargo moves on the Mississippi and Ohio river system. The intermodal side is the terminal where the box or the bulk changes mode, and the economics of the whole chain are set there: land transport is between half and two-thirds of the total transport cost of a container, and intermodal only beats trucking above roughly 500 km.

A loaded inland vessel moves a tonne of cargo for a fraction of the fuel a truck uses, needs no new right of way, and takes trucks off roads that ports are usually being sued about. It also scales cheaply, since adding a barge costs far less per tonne of capacity than adding a lane. The weakness is water. Load factor is set by available draft, so a low-water period cuts how much each vessel can carry and raises the cost per tonne-km at exactly the moment shippers need capacity, as happened on the Rhine in autumn 2018 and summer 2022 with modal share losses that persisted for years afterward. The mode is also losing ground: inland waterways fell to 5.1% of EU-27 inland freight in 2022, the lowest since 2005, while road rose to 77.8%. Speed is the other structural problem, since a barge takes days where a truck takes hours, which rules it out for anything time-critical.

Use inland waterways when the cargo is heavy, low-value, and not urgent, and when the origin or destination sits within a short drayage of a quay. For containers, the barge leg makes sense on a corridor with enough volume to fill a scheduled service, which in practice means the Rhine-Scheldt delta, the Yangtze, and a handful of US corridors. If the hinterland leg is under 500 km, expect trucking to win on cost and time; between 500 and 750 km rail is usually the answer, and beyond that water or rail both work. Build the inland terminal before the volume exists only if a single large shipper will commit, because a terminal at low utilization is expensive and a barge service that is not scheduled will not attract cargo. And model the low-water case explicitly, since a corridor that only works at full draft is a corridor with a seasonal outage.

European inland waterway transport about 116 billion tonne-km in 2023 · Rhine traffic 276.5 million tonnes in 2023, down 5.4% · new European dry cargo vessels average 2,664 tonnes, new tankers 4,022 tonnes · inland waterways 5.1% of EU-27 inland freight in 2022 against road at 77.8% and rail at 17.1% · Netherlands 41.0% waterway share, Germany below 7.0% · about 90% of US inland waterway cargo on the Mississippi and Ohio system · land transport is half to two-thirds of a container's total transport cost · intermodal generally needs distances above 500 km.

The Rhine-Scheldt corridor, where barge carried 48.7% of Antwerp-Bruges' maritime throughput in 2023 and 34.0% of its container hinterland traffic; the Rhine gauge at Kaub, which sets the available draft for the whole middle Rhine and closed the corridor to loaded vessels in 2018 and 2022; the Mississippi and Ohio barge system feeding the Gulf grain export terminals; the Yangtze feeder network serving Shanghai; inland ports such as Duisburg, the largest inland port in Europe.

The unit economics are driven by load factor, which is driven by water level, which nobody controls. A vessel that can only load to two-thirds draft carries two-thirds of the revenue for close to the same crew and fuel, so freight rates on the Rhine spike during low water and shippers who can switch to rail or road do, and often do not come back. That ratchet is why the mode keeps losing modal share in Europe despite lower emissions and lower cost per tonne-km at full load. On the capital side, the vessels are owned by small operators and family firms in Europe and by a few large companies in the US, while the locks, dams, and dredging are public infrastructure with public budgets, so the mode's cost structure depends on a maintenance program the operators do not fund and cannot schedule. For a shipper, the number to work out is total landed cost including drayage at both ends and the terminal handling charge, because inland intermodal costs run around half of total transport cost once terminal costs are counted, and a cheap barge leg with two expensive truck legs attached is not cheap.

7 minInland Barges - Different Types & Advantages | Barge Series #1Pile Buck · 10k+ views

3 minA Day in the Life of a Towboat CrewIngram Barge Company · 100k+ views

Annual Report 2024: Inland Navigation in Europe, Market Observation (Central Commission for the Navigation of the Rhine) · 5.6 – Intermodal Transportation and Containerization (The Geography of Transport Systems)

An integrated bridge system puts the ship's navigation sensors and controls on one network and one set of workstations, so the officer of the watch reads radar, electronic chart, position, heading, depth, speed and autopilot from the same console instead of six separate boxes. The core is ECDIS, the electronic chart display and information system, which SOLAS made mandatory on a phased schedule between 2012 and 2018 and which now carries the passage plan, the safety contour and the depth and crossing alarms. Around it sit radar with automatic tracking, AIS (mandatory on cargo ships over 300 GT in international trade since the end of 2004), a GNSS receiver, a gyro or fiber-optic compass, an echo sounder, a voyage data recorder, and track control that steers the planned route. E-navigation is the IMO's program to harmonize all of that and to connect it to shore services: the Strategy Implementation Plan defines five solutions covering bridge design, standardized reporting, equipment reliability, graphical presentation of information received over communications links, and improved vessel traffic service communication. The chart data itself is moving from the old S-57 format to the IHO's S-100 model, which lets bathymetry, currents, ice and route information layer onto the same display.

One console with one interaction style reduces the number of ways a watchkeeper can misread the situation, and route monitoring against a planned track catches errors a paper chart never would. The equipment is also cheap relative to the ship, which makes it the easiest safety upgrade on this sheet to justify. The failure modes are real and well documented. Alarm flooding is the classic one, where an integrated system generates so many alerts that the crew silences them as routine, which is why IMO wrote a separate bridge alert management standard. ECDIS-assisted groundings recur in accident reports when the safety contour or the depth alarms were set wrong or switched off, so the display's authority exceeds its reliability. And nearly everything on the bridge ultimately depends on GNSS: jamming and spoofing are now routine across the Baltic, the eastern Mediterranean and the Black Sea, and a spoofed position propagates into the chart, the AIS transmission, and the autopilot at the same time.

Specify a full integrated bridge on any newbuild, since the incremental cost over separate units is small and the wiring is decided once. On an existing ship, retrofit ECDIS and radar to a common workstation at a scheduled drydock, and treat crew training as the larger half of the budget, because ECDIS type-specific familiarization is where most of the accident reports point. Set the safety contour, safety depth and alarm limits as a fleet standard rather than leaving them to each watchkeeper, and audit them, because a ship-by-ship setting is how the wrong contour survives a whole voyage. Keep an independent position check that does not use GNSS, whether radar fixes, a terrestrial system or an inertial unit, and expect to use it in the areas where jamming is now normal. If the fleet is small and the ships are old, buy standalone ECDIS and radar rather than an integrated suite, since the integration payoff needs a bridge team trained on it.

ECDIS carriage phased in under SOLAS between 2012 and 2018 · AIS mandatory on cargo ships over 300 GT in international trade since the end of 2004 · five e-navigation solutions in the IMO Strategy Implementation Plan · chart data moving from S-57 to the IHO S-100 model · GNSS jamming and spoofing now routine across the Baltic, Mediterranean and Black Sea · capital cost is small next to a hull, which is why this is the cheapest safety upgrade available.

ECDIS and integrated bridges from Furuno, Wärtsilä, Kongsberg, JRC and Raytheon Anschütz; the IHO S-100 Universal Hydrographic Data Model and the S-101 electronic navigational chart specification built on it; VDES, the VHF Data Exchange System extending AIS to two-way data; IMO's bridge alert management performance standard; the UK Marine Accident Investigation Branch's series of ECDIS-assisted grounding reports; Baltic Sea GNSS interference reporting by national maritime administrations.

This is the rare item on a ship's specification where the hardware is cheap and the training is expensive. An integrated bridge is a small fraction of a newbuild price, well under a percent on a large merchant ship, and the recurring cost sits in chart licenses, software maintenance, and type-specific ECDIS training for every officer who joins. Chart data is a subscription business: hydrographic offices sell electronic navigational charts through regional coordinating centers, and a worldwide folio for a tramp trader costs real money every year, which is why some owners buy regional coverage and update it per voyage. The vendors make their margin on the service contract rather than the box, since a bridge suite has a 15-year life and a software support agreement that runs the whole time. For a buyer, the number worth calculating is not the purchase price but the cost of a single grounding, because hull and machinery deductibles alone run into the millions and the cheapest control available is a correctly configured alarm setting.

6 minECDIS and Other Connected Navigational EquipmentCamote XPrez · 50k+ views

8 minNavigation Equipment and Resources Used Onboard in a Modern Ships (Part 1)Marine Knows · 50k+ views

E-navigation (International Maritime Organization) · GNSS Jamming and Spoofing Situational Awareness Maps (TransNav, International Journal on Marine Navigation and Safety of Sea Transportation)

A maritime autonomous surface ship (MASS) is one where software or a shore operator takes over functions a crew normally performs on board. The IMO's definition is deliberately functional: a ship counts as MASS only when autonomous or remote technology replaces or supports crew functions and the ship holds a MASS Safety Certificate, and more automation on its own does not qualify. In practice the equipment is a sensor stack (radar, AIS, cameras, lidar) feeding a situational awareness and collision avoidance system, an autopilot with track and speed control, an auto-berthing function, and a satellite or 5G link to a remote operations center ashore. The IMO adopted a non-mandatory MASS Code in May 2026, in force from 1 July 2026, which applies to cargo ships covered by SOLAS chapter I, generally over 500 GT on international voyages. The Code keeps the master responsible for the ship whether or not the master is on board, and requires the shore center to be assessed, certified and run under a safety management system. As of that adoption, the IMO's own position is that no commercially operating, internationally trading SOLAS cargo ship runs autonomously or under remote control.

The real argument for this is crew supply rather than crew cost. Japan's coastal fleet has more than half its crew over 50, the trade is unattractive, and roughly 400 inhabited offshore islands need services that are hard to staff, which is why the country funded the largest demonstration program in the world. Human error is cited in 70–80% of maritime accidents, so a system that never gets tired has a plausible safety case. Against that, the demonstrated vessels are small, short-route and heavily supervised: MEGURI2040's Stage 1 tested six vessels including the 749 GT coastal container ship SUZAKU, and Yara Birkeland carries 120 TEU on a route of about 14 km and still sails with people aboard. Removing the crew also removes the people who fix things, and a merchant ship at sea has no roadside assistance. Bandwidth and latency set the limits on what a shore operator can actually do, and the whole architecture inherits the GNSS jamming and cybersecurity problems of the systems it is built on.

Take remote support seriously today and full autonomy only on a specific short route. If you run ferries, harbor craft or offshore support vessels on a fixed track with reliable connectivity, automated crossing and auto-berthing are commercially available and worth buying, because they cut the workload on the part of the voyage where accidents happen. If you run deep-sea tonnage, buy the shore-support version instead: a fleet operations center doing route advice, engine diagnostics and condition monitoring reduces crew workload without touching the certificate. Do not plan a crewless newbuild against the current rulebook, because the mandatory MASS Code is not expected to be adopted before 1 July 2030 and enters force on 1 January 2032, and a flag state's approval today is a case-by-case alternative-design exercise. If your interest is uncrewed vessels as a survey or defense platform rather than as cargo carriers, see the `uncrewed-systems` sheet, which covers that class in its own right.

Non-mandatory MASS Code adopted May 2026, in force 1 July 2026, covering SOLAS cargo ships generally above 500 GT · mandatory Code expected to be adopted by 1 July 2030 and in force 1 January 2032 · no internationally trading SOLAS cargo ship currently operates autonomously or remotely · MEGURI2040 Stage 1 tested six vessels with five consortia from 2020 to 2022, including the 749 GT SUZAKU · Yara Birkeland carries 120 TEU on a route of about 14 km and still sails crewed · human error cited in 70–80% of maritime accidents · more than half of Japan's coastal crew are over 50.

The IMO MASS Code, resolution MSC.595(111), preceded by the 2019 interim trial guidelines and the 2021 regulatory scoping exercise; the Nippon Foundation's MEGURI2040 program, with Stage 1 running 2020–2022 and Stage 2 from 2023 to 2026 on collision avoidance and automatic berthing, and a land-based Fleet Operation Center at Makuhari; Yara Birkeland in Norway, running Herøya to Brevik since 2022; automatic crossing systems on Norwegian ferries; remote operations centers offered by Kongsberg, Wärtsilä and ABB.

Crew is a small share of a large merchant ship's total voyage cost, roughly 5–10% once fuel and capital are counted, so removing it does not transform the economics of a deep-sea trade, and the accommodation block, lifeboats and hotel systems are already built. The money is elsewhere. On a short domestic route, one crew covering two or three vessels from a shore center changes the staffing math for an operator who cannot recruit at all, which is the Japanese and Norwegian case and not a cost-cutting case. On the capital side, the sensor and autonomy package is a modest addition to a newbuild, medium rather than high, while the expensive parts are the shore center, its certification, and the redundancy needed to satisfy a flag state under an alternative-design assessment. Insurance is the quiet blocker: liability rests with a master who may be ashore, the Legal Committee is still working through which conventions need amending, and underwriters price novelty. Until the mandatory Code lands around 2032, the addressable market is domestic and short-sea tonnage under national rules, and the vendors selling into it are selling sensors and software rather than ships.

3 minSelf-driving electric container ship sets sail in Norway - BBC NewsBBC News · 100k+ views

10 minYara Birkeland Electric Container Ship Takes Maiden VoyageTech Inspection · 10k+ views

FAQ - Autonomous shipping (International Maritime Organization) · The Nippon Foundation MEGURI2040 Fully Autonomous Ship Program (The Nippon Foundation)

Condition monitoring instruments the machinery that can stop a ship and streams the readings ashore. The usual set on a large two-stroke is cylinder pressure, exhaust gas temperature per cylinder, turbocharger speed and vibration, main and crankpin bearing temperatures, shaft torque and thrust, and lube oil sampling for wear metals and water. Sampling runs from one second to one minute rather than the single noon report that used to be the whole dataset, and the link ashore is now usually low-earth-orbit broadband at roughly $250–2,000 a month instead of a metered VSAT channel. A performance layer sits on top: measured shaft power for a given speed and draft is compared against an expected value, following ISO 19030 for hull and propeller performance, which is how fouling becomes visible months before the next drydock. Most products sold as marine digital twins are this plus a 3D view and a physics model of one subsystem; genuinely coupled twins exist for hull structural fatigue and inside a few engine makers' own diagnostics, and the honest description of the rest is condition monitoring with a better interface.

The value is concentrated in a short list of failures that strand a ship: turbocharger failure, scavenge fire, liner scuffing and a wiped bearing. A main engine breakdown mid-ocean means a tow, a repair berth and off-hire at roughly $20,000–60,000 a day for a large ship, so one avoided event pays for the system several times over. The second payback is regulatory rather than mechanical, because class societies will approve a condition-based maintenance scheme that lets an owner skip opening machinery at survey when the data supports it, which removes work instead of adding a dashboard. The weaknesses are ordinary: sensors drift and few crews recalibrate them, alarm volume outruns the attention available on a 20-person ship, and most predictive claims rest on models trained on too few real failures to predict anything, so what works in practice is trending against a known-good baseline. The performance half is worth 1–3% of fuel on its own, which is real but small beside a single avoided breakdown, and the platform costs money every month whether or not a failure was ever coming.

If the ship has a large two-stroke and long ocean legs, start with main-engine and turbocharger monitoring plus lube oil analysis, because those are the failures with the biggest bill behind them; budget $30,000–150,000 per ship for the sensors and gateways and $1,000–5,000 a month for the software and the link. If the ships are on time charter and the charterer buys the bunkers, do not pay for the fuel-performance half of the product, since the saving goes to someone else; buy the reliability half, which the owner keeps. If a vendor is selling a digital twin, ask which physics is actually modeled, on which subsystem, and what it predicts that a trend line against a baseline does not. And if the system has to pay for itself on paper, go after the class condition-based maintenance notation, which is the only part of this that reliably takes cost out rather than adding it.

Sensor and platform retrofit roughly $30,000–150,000 per ship · $1,000–5,000 per ship per month for software and connectivity · off-hire $20,000–60,000 a day for a large ship · sampling at 1 second to 1 minute against one noon report a day · performance monitoring typically worth 1–3% of fuel · low-earth-orbit broadband at roughly $250–2,000 a month per ship

MAN Energy Solutions PrimeServ Assist and CEON; WinGD's WiDE engine diagnostics; Kongsberg Vessel Insight; ABB Ability Tekomar XPERT; Wärtsilä Expert Insight; ISO 19030 for hull and propeller performance measurement; condition-based maintenance notations from DNV, Lloyd's Register and ABS.

This is one of the few digital items on a ship where the owner both spends the money and keeps the benefit, because maintenance and off-hire sit on the owner's side of a time charter while fuel does not. The capital is small against the asset: $30,000–150,000 of sensors and gateways on a ship worth $50–200M, plus $1,000–5,000 a month for the software and the link. Connectivity used to be the blocker, since a metered VSAT channel made high-rate telemetry expensive per megabyte; low-earth-orbit plans at roughly $250–2,000 a month for tens of megabits removed that constraint and are most of why continuous monitoring spread across the fleet after 2023. The vendors make their margin on subscription rather than hardware, which is why nearly every engine maker and integrator now sells a platform and why the feature lists look alike. On the buyer's side the return is lumpy: one avoided main-engine failure at $20,000–60,000 a day of off-hire covers years of subscription, and a condition-based maintenance scheme takes recurring survey work out, while the 1–3% fuel saving from performance monitoring is real but usually credited to whoever pays for the bunkers.

7 minPredictive Maintenance ExplainedRealPars · 100k+ views

16 minDigital Twin in action: Advanced real time diagnostics for WinGD enginesSAFETY4SEA · under 1k views

A performance degradation evaluation method for a turbocharger in a diesel engine (Royal Society Open Science) · Research and application of digital twin technology in waterway transportation (Chinese Journal of Ship Research)

Cyber risk became a shipboard compliance item through the safety management system rather than through a separate rule. IMO Resolution MSC.428(98) requires cyber risks to be addressed in the company's safety management system, checked at the first annual verification of the Document of Compliance after 1 January 2021, with the practical detail left to the guidelines in MSC-FAL.1/Circ.3. The exposure on a ship is operational technology, not the office network: ECDIS, radar, AIS, the voyage data recorder, GNSS receivers, the engine alarm and monitoring system, and propulsion, steering and cargo control. Much of that equipment runs old embedded Windows on a flat network with no segmentation, and takes chart and software updates from a USB stick or a service engineer's laptop, which are the two ingress paths that show up in real incidents. The buses between the boxes make it worse, because NMEA 0183 and NMEA 2000 carry position and heading with no authentication at all, so anything on the wire is believed. IACS unified requirements E26 and E27 push a fix into newbuilds by requiring cyber resilience of the ship and of individual equipment for ships contracted from 1 July 2024, which means the existing fleet is covered only by the management-system rule.

The rules are cheap to satisfy on paper and hard to satisfy in fact, because an audit checks that procedures exist while the vulnerability is in twenty-year-old equipment nobody will replace before scrapping. The incidents that have actually cost money were IT, not OT: Maersk's NotPetya infection in 2017 ran to roughly $250–300M and took about 45,000 PCs and 4,000 servers to rebuild, and the Port of Nagoya ransomware attack in 2023 stopped container handling for about two days. Nothing public has yet grounded or sunk a ship through compromised shipboard control systems, so the honest risk ranking puts shoreside ransomware first and OT compromise second. The routine operational problem is GNSS: jamming and spoofing are now everyday conditions in the eastern Mediterranean, the Black Sea, the Gulf and the Strait of Hormuz, the Baltic near Kaliningrad, and around the Korean peninsula, and the bridge sees a position that jumps, an AIS track that puts the ship inland, and an autopilot disagreeing with the gyro. That is not an attack on any one ship and there is no patch for it, which is why the mitigations are procedural rather than technical.

Do the safety-management-system work first, because it is the audit item and it forces an inventory of what is actually on the ship's networks. Then segment: put the bridge and engine networks behind a boundary from crew wifi and the business network, since a flat network turns one infected laptop into a bridge problem. Control the two ingress paths that matter, which are removable media and vendor remote access, before buying any detection product, because a monitoring service on an unmapped network reports noise. If the ship trades in a spoofing region, train the bridge team to cross-check GNSS against radar parallel indexing and visual bearings and to log the event, since multi-constellation receivers and inertial aiding reduce the problem without removing it. And if you are specifying a newbuild, ask for E26 and E27 compliance in the shipbuilding contract, because retrofitting segmentation and secure update paths into a delivered ship costs several times what designing them in does. None of this is expensive by shipping standards: a risk assessment runs $10,000–30,000 per ship, segmentation hardware and engineering $20,000–60,000, and a managed monitoring service $5,000–20,000 per ship per year, so the decision is about sequencing rather than budget.

MSC.428(98) applies from the first Document of Compliance verification after 1 January 2021 · IACS UR E26 and E27 for ships contracted from 1 July 2024 · cyber risk assessment typically $10,000–30,000 per ship · segmentation and hardware $20,000–60,000 per ship · managed monitoring $5,000–20,000 per ship per year · Maersk's 2017 NotPetya loss roughly $250–300M and about 45,000 PCs rebuilt

IMO Resolution MSC.428(98) and the guidelines in MSC-FAL.1/Circ.3; IACS unified requirements E26 and E27; the industry Guidelines on Cyber Security Onboard Ships published by BIMCO, ICS and other associations; the US Coast Guard's 2025 cybersecurity rule for US-flagged vessels and regulated facilities; the 2017 NotPetya infection of Maersk and the 2023 Port of Nagoya ransomware shutdown.

The whole program is small money against the asset, which is exactly why it is treated as compliance rather than as an investment case. A risk assessment runs $10,000–30,000 per ship, segmentation hardware and the engineering behind it $20,000–60,000, and a managed monitoring service $5,000–20,000 per ship per year, against a ship worth $50–200M. Nobody buys it on expected loss, because the loss distribution is one large shoreside event every few years rather than a steady drip, and the 2017 Maersk figure of $250–300M is the only public number large enough to anchor on. The demand actually comes from audits and vetting: flag and class check the management system, tanker charterers check it through TMSA, dry bulk charterers through RightShip, and hull and war-risk policies now carry cyber exclusions that make the coverage question explicit. Vendors sell managed detection and response by the ship-year, which suits a fleet operator with no security staff and produces a recurring revenue line that scales with fleet size rather than with risk.

2 minSpoofing on the High SeasTexas Engineering · 500k+ views

19 minMaritime Transportation System Cybersecurity: An Overview (1/3)NCyTE Center · 5k+ views

Maritime cyber risk (IMO) · Raising Awareness on Cyber Security of ECDIS (TransNav, the International Journal on Marine Navigation and Safety of Sea Transportation)

The IMO gives every large ship an annual carbon intensity rating, and what it rates is how the ship was operated, not how it was designed. The Carbon Intensity Indicator applies to ships of 5,000 gross tons and above, and for most cargo ships it is the Annual Efficiency Ratio: total CO2 emitted over the calendar year divided by deadweight capacity times distance sailed, in grams of CO2 per deadweight tonne-nautical mile. Both inputs were already being collected, since fuel burn is reported under the IMO Data Collection System and distance comes from the ship's own log, which is why the rule was cheap to introduce. The attained figure is compared with a required CII, which is the 2019 reference line for that ship type and size cut by a Z factor of 5% in 2023 rising 2 points a year to 11% in 2026, and the comparison produces a letter from A to E with C sitting on the line. EEXI is the separate design-side rule and works nothing like it: it applies to ships of 400 gross tons and above, is calculated once from installed power, reference speed and capacity, and has to come out below the EEDI reference line by a factor set per ship type and size, 20% for most bulk carriers and tankers and up to 50% for the largest container ships. Both sit under MARPOL Annex VI amendments in force since November 2022, and both serve the IMO's target of cutting the fleet's carbon intensity 40% by 2030 against 2008. A wider IMO net-zero framework, built on a fuel intensity standard with remedial units priced at $100 and $380 per tonne of CO2e, was approved in draft in April 2025 and is meant to supersede this argument, but its adoption was deferred and it is not yet in force.

The strength of CII is that it measures what a ship actually emitted rather than what a naval architect calculated, and it produced a commercial signal within about two years: charterers write CII clauses, lenders report their portfolios against the same data under the Poseidon Principles, and brokers quote the letter. The weakness is in the denominator. Deadweight capacity times distance is not transport work, so a ship that sails a long ballast leg with empty holds improves its rating while carrying nothing, and a ship that waits three weeks at anchor off a congested port burns auxiliary and boiler fuel with no distance to divide it by, which can cost it a band for reasons the crew did not choose. Ships in short-sea trades spend more of the year alongside per mile sailed and are penalized for it, and a partly loaded ship gets no credit for the cargo it did carry, because cargo never enters the formula. Enforcement is weak as well: three consecutive D ratings, or one E, require a corrective action plan in the ship's SEEMP, and there is no fine, no detention and no trading ban behind it.

Treat the rating as a commercial constraint rather than a technical one. If a ship is rated D and sliding, do the cheap operational work first: hull cleaning and propeller polishing usually recover 3–8% of fuel, and after that speed does the heavy lifting, because fuel per mile falls with roughly the square of speed, so going from 14 to 12.5 knots takes about 20% off fuel per mile and moves the rating directly. If you own the ship but the charterer sets the speed and decides how long it waits, the rating is not yours to fix, so settle it in the charterparty before the fixture instead of arguing about the letter a year later. If the ship trades short-sea or into congested ports, expect the trade rather than the ship to set the rating, and do not spend capital chasing a band. And if you are buying secondhand, check the attained CII against the line as it will stand in 2026 and after, not against the letter printed on today's Statement of Compliance.

CII applies at 5,000 GT and EEXI at 400 GT · AER in grams CO2 per deadweight tonne-mile · Z factor 5% in 2023 rising to 11% in 2026 against a 2019 reference line · ratings A to E, three consecutive D or one E triggers a corrective action plan · EEXI required 20% below the EEDI reference line for most bulkers and tankers, up to 50% for the largest boxships · IMO target 40% lower fleet carbon intensity by 2030 against 2008 · draft IMO net-zero framework prices remedial units at $100 and $380 per tonne CO2e

MARPOL Annex VI regulations 25, 27 and 28, in force since 1 November 2022 with EEXI and CII certification required from 1 January 2023; the IMO Data Collection System, which supplies the fuel figures; BIMCO's CII Operations Clause for Time Charter Parties 2022; the Poseidon Principles, under which lenders report their shipping portfolios against a carbon intensity trajectory.

No fine is attached to a bad CII rating, so the cost arrives commercially, through charter clauses, lender reporting and what a broker can get for the ship. EEXI compliance was the cheap part: most existing bulk carriers and tankers met it with an engine or shaft power limiter, which is engineering, a control device and a class approval for roughly $50,000–200,000 per ship, and many of those ships were already sailing below the limited power anyway. CII costs money in voyage days instead, since cutting speed 10% adds about 11% to the days at sea for the same distance, and on a $20,000 a day time charter the charterer pays that while the owner keeps the rating. Hull cleaning and propeller polishing run $20,000–60,000 in the water or at a drydock. The bigger number sits in the rule meant to supersede this one: the draft IMO net-zero framework approved at MEPC 83 in April 2025 would price remedial units at $100 and $380 per tonne of CO2e, and since a tonne of fuel oil burns to about 3.11 tonnes of CO2, the upper tier works out at roughly $1,180 per tonne of fuel, about twice the $500–600 a tonne fuel oil costs today. An extraordinary IMO session in October 2025 deferred adoption by a year, so as of mid-2026 nothing there is in force, but that is the instrument owners are pricing into a 2030s newbuild rather than the CII letter.

8 minAn Ocean Scientist Explains... Maritime Shipping's Carbon Intensity Indicator (CII)Sofar · 1k+ views

6 minNavigating EEDI, EEXI, CII & SEEMP: An Overview For Better Understanding!Oral exam Mmd prep · 10k+ views

EEXI and CII - ship carbon intensity and rating system (IMO) · A review of law and policy on decarbonization of shipping (Frontiers in Marine Science)

Two EU rules now put a price on the carbon in a ship's fuel, and they work differently. The EU Emissions Trading System brings shipping into the same allowance market as power stations and heavy industry: ships of 5,000 gross tons and above have surrendered allowances since 2024, phased in at 40% of verified 2024 emissions, 70% of 2025 and 100% from 2026, with methane and nitrous oxide added to CO2 from 2026. FuelEU Maritime trades nothing and instead caps the well-to-wake greenhouse gas intensity of the energy a ship uses on board, starting 2% below a 91.16 gCO2e/MJ baseline in 2025 and tightening to 6% in 2030, 14.5% in 2035 and 80% in 2050. Both cover the same geography: all emissions on voyages between EEA ports and at berth in an EEA port, and half the emissions on a voyage into or out of the EEA. Both name the ISM company as the responsible party, with a contractual right to recover the cost from whoever operates the ship commercially. The two stack, so a ship burning conventional fuel oil into Europe pays both.

The two rules do different jobs and together they leave few gaps: the ETS prices every tonne of CO2 at whatever an allowance is worth on the day, and FuelEU forces the intensity of the fuel down whether or not the allowance price is high enough to do it on its own. Pooling and banking are the part of FuelEU that changes behavior fastest, because a ship with a surplus can bank it into the following year or pool it with other ships, so one LNG or methanol vessel can carry several conventional ones and the first alternative-fuel ship in a fleet is worth more than its own fuel saving. The weakness is that the rules are regional. Only half of an intercontinental voyage is priced, and that half is measured from the last port of call, so a call at a transshipment hub just outside the EEA shortens the priced leg; the Commission anticipated it by designating container ports within 300 nautical miles whose transshipment share exceeds 65% as neighboring transshipment ports, with East Port Said and Tanger Med on the list, and a call there does not count as a port of call. The same routing move still works through hubs that are not on the list, and several third countries object to the EU pricing emissions burned in international waters, which is the argument a global IMO measure is meant to settle.

If you operate ships into Europe, work out the bill per voyage before arguing about fuel. At €70 an allowance and full phase-in the ETS adds about €220 to every in-scope tonne of fuel oil, and the FuelEU penalty of €2,400 per tonne of fuel-oil-equivalent deficit works out at roughly €48 more per tonne burned in 2025, so budget about €270 per in-scope tonne against a bunker price near €500, or about €40 per loaded TEU on an Asia to North Europe leg. If the fleet has one alternative-fuel ship, pool it rather than treating the surplus as that ship's own benefit, because the pool spreads it across the ships that would otherwise pay the penalty. If you are fixing a time charter, use BIMCO's ETS and FuelEU clauses rather than drafting your own, since the allocation question is already settled in them and the remaining argument is only about who keeps the accounts. And do not plan a routing change around the 50% rule without checking the neighboring transshipment port list first, because the two obvious hubs are already on it.

ETS phase-in 40% of 2024 emissions, 70% of 2025, 100% from 2026 · 50% of emissions counted on voyages into or out of the EEA, 100% intra-EEA and at berth · about €220 per tonne of fuel oil at €70 an allowance · FuelEU limit 2% below 91.16 gCO2e/MJ in 2025, 6% in 2030, 14.5% in 2035, 80% in 2050 · FuelEU penalty €2,400 per tonne of fuel-oil-equivalent deficit, about €48 per tonne of fuel burned in 2025 · roughly €40 per loaded TEU on an Asia to North Europe leg at full phase-in, derived here rather than quoted

Directive (EU) 2023/959, which amended the EU ETS Directive to cover maritime transport, and Regulation (EU) 2023/1805, the FuelEU Maritime Regulation; the Commission's list of neighboring container transshipment ports, which names East Port Said and Tanger Med; BIMCO's Emission Trading Scheme Allowances Clause 2022 and its FuelEU Maritime Clause for Time Charter Parties 2024; the ETS surcharges container lines published from January 2024.

The FuelEU headline number is €2,400 and the bill per tonne of fuel is much smaller, because the €2,400 is charged per tonne of fuel-oil-equivalent deficit rather than per tonne burned: divide the intensity gap by the ship's actual intensity and the answer is close to the reduction factor, so the penalty per tonne of fuel oil is roughly €2,400 times that factor, or €48 in 2025, €144 in 2030 and €348 in 2035. The ETS is the bigger bill today. Take a 15,000 TEU ship from Shanghai to Rotterdam, about 10,800 nautical miles at 19 knots, so roughly 24 days and 4,000 tonnes of fuel at 170 tonnes a day; the voyage enters the EEA from outside, so half counts, and 2,000 tonnes at 3.11 tonnes of CO2 per tonne of fuel is about 6,200 tonnes of CO2, or €435,000 at €70 an allowance, plus about €96,000 of FuelEU exposure. Spread over roughly 13,500 loaded boxes that is about €40 per TEU, and closer to €70 if the lightly loaded return leg is charged to the same cargo; those numbers are worked here rather than quoted, and they are what the container lines turned into published surcharges. Under a time charter the charterer pays through the BIMCO clause and passes it to the cargo, so the party writing the check is rarely the party that chose the ship. What the price does not do is close the gap to green fuel: a tonne of fuel oil holds about 40 GJ, so replacing it with green methanol at $50–80/GJ against $14/GJ costs a premium of roughly $1,450–2,650, and about $290 of combined carbon cost covers only 10–20% of that. The practical effect is to move certified biofuel blends and LNG ahead of fuel oil on cost, without making e-fuels competitive.

3 minEU ETS for shipping explained | Feb 2024DNV Maritime · 50k+ views

4 minFuelEU Maritime Explained in 3 MinutesMaritime library · 1k+ views

Reducing emissions from the shipping sector (European Commission) · Carbon Pricing for International Shipping and Border Carbon Adjustment Mechanisms: A Case for Regulatory Cooperation (European Journal of Risk Regulation)

Two different organizations decide whether a ship may trade, and they do different jobs. A classification society publishes its own technical rules for hull structure, propulsion, steering and power generation, approves the design plans against them, puts surveyors in the shipyard and at the steel mill and engine works, attends sea trials, and then keeps the ship in class through a five-year cycle of annual surveys, one intermediate survey, and a class renewal or special survey. The flag state is the country the ship is registered in, and it is the legal authority that enforces SOLAS, MARPOL, the ISM Code and STCW on that ship anywhere in the world. Most flags do not employ enough surveyors to do that themselves, so they delegate statutory survey and certification to class societies acting as recognized organizations under IMO Resolution A.739(18), which is why the same surveyor often signs both the class certificate and the flag's safety construction certificate. More than 90% of cargo-carrying tonnage is classed by a member of the International Association of Classification Societies, which has 11 members; Equasis data for 2022 put IACS coverage at 80% of ships by number and 97% by tonnage. Port state control is the third layer, where the country whose port the ship visits inspects it and can detain it until the deficiencies are fixed, organized as nine regional agreements plus the US Coast Guard operating on its own.

The structure works because class societies hold no commercial stake in ship design, building, ownership, operation, insurance or chartering, and because the rules are empirical, built up over 150 years of classing ships and then harmonized across IACS as Unified Requirements and Common Structural Rules. The weakness is that the shipowner pays the class society and can move the ship to another society, so the party being judged picks and pays the judge. Class is also narrower than it sounds: the certificate attests only that the ship complied with that society's rules at the time of survey, a surveyor is aboard roughly once in twelve months, and the survey is a sampling exercise rather than a full examination of the structure. Flag states compete for registrations on price, turnaround and crewing rules, and a registry with no technical staff of its own delegates everything to a recognized organization. Port state control is the backstop that catches the result, and the difference shows up in the numbers: about 2% of inspections end in detention, with IACS-classed ships detained at 2% against 4% for ships classed elsewhere or with no class record.

If you are ordering or financing a ship, specify an IACS member society and a flag on the Paris MoU white list, because banks, P&I clubs, charterers and oil-major vetting inspectors all price on both, and a flag on the Paris MoU Grey or Black list draws more inspections for the life of the ship. If the installation is novel (an ammonia fuel system, a large battery room, a rotor sail on a tanker deck), get an Approval in Principle from the class society before you sign the shipbuilding contract, then allow 6–18 months and real engineering cost for the SOLAS alternative-design risk assessment that follows, because that assessment, not the hardware, is usually what sets the delivery date. If the trade calls regularly at European or US ports, check the flag's Paris MoU standing and whether it qualifies for the US Coast Guard's Qualship 21 program, since a poor ship risk profile means more frequent and more detailed inspections. Do not change class late in a newbuild or shortly before a special survey; the IACS transfer-of-class procedure exists because that used to be a way to leave outstanding conditions of class behind, and both societies and port state control now look at the history. If the only reason to pick a register is that it is cheap, be clear about what comes with it, since the flag decides which conventions apply to the ship, who investigates a casualty, and which administration you deal with when a certificate is suspended.

11 IACS member societies · over 90% of cargo-carrying tonnage in IACS class, 97% by tonnage and 80% by ship count in Equasis 2022 data · five-year survey cycle of annual, intermediate and renewal surveys · surveyor aboard roughly once in 12 months · about 2% of port state control inspections end in detention, 2% for IACS-classed ships against 4% for others · 9 regional port state control agreements plus the US Coast Guard · Approval in Principle plus a SOLAS alternative-design case adds 6–18 months

The 11 IACS members are ABS, Bureau Veritas, China Classification Society, Croatian Register of Shipping, DNV, Indian Register of Shipping, Korean Register, Lloyd's Register, ClassNK, Polish Register of Shipping and RINA; Panama, Liberia and the Marshall Islands are the three largest registries, and the Liberian and Marshall Islands registers are run under contract by private companies headquartered in the United States. The Paris MoU White, Grey and Black lists are recalculated each year from three years of inspection and detention data for every flag with at least 30 inspections in the period, and the nine regional agreements plus the US Coast Guard's Qualship 21 program are what a substandard ship actually runs into.

The direct fees are small. Class fees, statutory certification and registration together are a minor line against a $150–200M ship, which is exactly why shopping on them is the wrong instinct: the cost of a cheap flag arrives later as inspection frequency, detention risk, a worse P&I quote and a failed charterer vetting. Open registries are a business in their own right, run for fee income on tonnage, and owners use them because a national flag usually requires national officers and crew and brings national tax and labor law with it while an open register does not. Registries compete on price, turnaround and crewing rules rather than on standards, which is why the largest ones are operated by private companies under government contract. Class societies earn survey fees on a five-year cycle, and their fastest-growing line is now advisory work on new fuels, because every ammonia, methanol or battery installation needs an Approval in Principle and then an alternative-design risk assessment before a yard will build it. That is where this entry connects to the money on the rest of the sheet: a novel installation cannot be ordered until class will approve it, so the 6–18 months of approval work sits ahead of a yard queue that is already booked years out, and the technology being ready is not the same thing as the ship being orderable. The other place it shows up is resale, since a buyer in 2035 will read the survey history, the conditions of class, and any transfer of class or flag, and price the ship accordingly.

6 minClassification Societies in the Maritime WorldMeij Webdesign · 10k+ views

9 minTHE DIFFERENCE BETWEEN CLASS AND FLAGDöhle Yachts · 1k+ views

5 minPort State Control Inspections What to ExpectWest P&I · 5k+ views

Classification societies – what, why and how? (IACS) · Port State Control (International Maritime Organization)

Ships are hired under four contract types, and the only thing an investor needs to memorize is who buys the bunkers under each. On a voyage charter the owner carries a stated cargo between named ports for a freight rate per tonne and pays the bunkers, port charges and canal dues out of it. On a time charter the charterer hires the ship by the day, decides where it goes, and buys the bunkers, while the owner pays crew, insurance, stores and maintenance. On a bareboat or demise charter the charterer takes the ship with no crew and pays for everything including fuel, which makes it a finance lease with a hull attached. A contract of affreightment commits the owner to move a quantity of cargo over a period in agreed shipments, with the owner picking the ships and paying voyage costs, so it prices like a string of voyage charters. The standard forms are GENCON 2022, NYPE 2015 and BARECON 2017, all published by BIMCO, and rates are benchmarked against the Baltic Exchange indices: the Baltic Dry Index has been published daily since 1 November 1999 and is a weighted basket of the Capesize, Panamax and Supramax time-charter averages at 40%, 30% and 30%, multiplied by 0.10.

Each contract puts the fuel bill somewhere different, and that placement decides who is willing to pay for efficiency. An owner trading on voyage charters or a contract of affreightment buys the fuel and keeps every gallon saved, which is why owner-operators fit coatings and energy-saving devices without being pushed; the same owner also carries the bunker price risk unless the charterparty has an escalation clause. A time charter gives the owner steady daily revenue that a bank will lend against, and hands the fuel saving to the charterer, so the party spending $15–25M on a dual-fuel engine is not the party whose bill falls. The other property of this market is how far rates swing, and the mechanism is structural: demand comes from commodity and container flows that barely respond to freight price in the short term, while supply is fixed for two to three years because that is the build time and for 25 years because that is the ship's life. Once the fleet is near full utilization, a few percent of extra demand can multiply the rate, and a few percent of surplus tonnage can push it under the cost of running the ship. The size of the swing is easy to underrate: the Baltic Dry Index reached 11,793 in May 2008 and 663 by that December, its all-time low is 290 in February 2016, and a Capesize time-charter average has run from under $5,000 a day to over $80,000 a day inside five years against an operating cost of roughly $6,000–8,000 a day. Container freight behaves the same way, with Drewry's World Container Index composite at roughly $1,400 per 40-foot container in 2019, over $10,000 in September 2021, and $4,339 in August 2026.

If you sell a technology whose value is saved fuel, find out which contract the target ship trades on before you build a business case, because on a time charter you are selling to the owner a saving that lands with the charterer. If you are the owner and want that saving, trade the ship on voyage charters or a contract of affreightment, or write an efficiency adjustment into the hire rate on a long time charter. If you are a charterer with a decarbonization target, take a long time charter or a bareboat charter, since that is the only way to specify the ship you actually want and to hold it long enough to pay back the specification. If you are underwriting an investment, do not extrapolate today's spot rate: use a cycle average, because owners who model the peak forward are the ones who order at the top of the market and take delivery into the trough two or three years later. Hedge the exposure you cannot avoid with forward freight agreements, which are cash-settled against the Baltic indices and cleared, and compare voyage fixtures with daily hire using time charter equivalent, which is voyage revenue minus voyage costs divided by voyage days.

Baltic Dry Index is 40% Capesize, 30% Panamax and 30% Supramax time-charter averages, multiplied by 0.10 · published daily since 1 November 1999 · BDI 11,793 in May 2008 and 663 that December, all-time low 290 in February 2016 · Capesize hire has run from under $5,000/day to over $80,000/day within five years · Capesize operating cost roughly $6,000–8,000/day before finance · Drewry World Container Index composite $4,339 per 40-foot container in August 2026, against roughly $1,400 in 2019 and over $10,000 in September 2021 · newbuild lead time 2–3 years

BIMCO's GENCON 2022 voyage charter, NYPE 2015 time charter and BARECON 2017 bareboat form; the Baltic Dry Index and its Capesize, Panamax, Supramax and Handysize components, plus the Baltic Dirty and Clean Tanker Indices; Drewry's World Container Index, quoted at $4,339 per 40-foot container in August 2026 with Shanghai to New York at $8,706; forward freight agreements cleared against the Baltic indices.

Freight revenue splits three ways: voyage costs, operating costs, and capital. Bunkers are usually the largest single voyage cost, port and canal dues come next, and the broker takes a commission of a few percent on freight or hire. Operating cost for a Capesize bulker runs roughly $6,000–8,000 a day for crew, insurance, stores, maintenance and a drydock accrual, before any finance, which sets the floor: when the time-charter rate falls under that number, owners lay ships up or slow steam rather than trade at a loss. Above that floor there is no reliable ceiling to plan against, and a Capesize has gone from under $5,000 a day to over $80,000 a day inside five years, so a large part of a shipowner's return across a cycle comes from buying and selling ships rather than from operating them. That volatility is what breaks investment timing in this industry: owners have the most cash and the most confidence at the top, which is exactly when yards quote their highest prices and their longest lead times, and the ship then delivers two to three years later into a weaker market with a high capital cost attached to it. For anyone selling into shipping, the practical consequence is that the decision to spend depends less on the payback calculation than on where the cycle is and on which counterparty holds the fuel bill.

5 minShip Chartering // Voyage Charter // Time Charter // Bareboat charter.Captain's Adventures · 5k+ views

8 minShip Chartering Explained: Voyage Charter, Time Charter, Bareboat & Contract of Affreightment (CoA)Bazina Shipping | Chartering Training & Advisory · 1k+ views

2 minWhat's up... with the Baltic Dry Index?Marketplace APM · 1k+ views

NYPE 2015 (BIMCO) · Guide to Market Benchmarks (Baltic Exchange Information Services)

Merchant shipbuilding capacity is measured in compensated gross tonnage, which weights a ship's size by how much construction effort it takes, and almost all of it sits in three countries. China holds roughly half the global order book by CGT and has taken share steadily for a decade. South Korea moved upmarket rather than compete on volume, and LNG carriers were 53% of Korean yards' workload by CGT in February 2025 with containerships another 22%, which is a concentrated position now under pressure because Chinese yards win LNG orders too. Japan builds mostly standard bulk carriers and tankers for its own owners and has fallen well back from the position it held in the 1970s. Berths are largely sold into 2028 and 2029, and prices have followed the backlog: SEA Europe estimates that newbuild prices rose about 46% from the start of 2021 while yard input costs rose about 22%. Half of that ordering is now alternative-fuel capable, with 820 vessels of 62.2m GT ordered with alternative-fuel capability in 2024, which is 50% of the tonnage ordered that year, and Chinese yards hold 47% of that alternative-fuel order book by CGT against Korea's 42%.

A full order book is good for yards and bad for buyers, and the mechanism is simply that a berth is a fixed number of ship-months. Yards with three or four years of forward cover can price hard and can also decline work, and a repeat conventional design uses less engineering and less berth time than a first-of-class alternative-fuel ship, so the green premium widens exactly when the book is full. Yards carry their own risk, though: roughly 70% of the value of a shipbuilding contract sits in the supply chain, which leaves a thin gross margin on a fixed-price contract whose steel, main engine and cargo equipment will be bought two or three years after signing. The binding constraint on capacity is now labor rather than slipways, since shipbuilding is struggling to recruit worldwide, and SEA Europe's reading is that another boom on the scale of previous peaks may not be possible. For a buyer the practical weakness is the queue: 82 builders constructed an alternative-fuel-capable vessel in 2024, so the set of yards that has actually done your ship type on your fuel is small, and everything else in the decarbonization argument sits behind it.

If the ship has to be in service before 2030, book the berth before you finalize the specification, because the slot is scarcer than the engine and a decision taken today is steel in about three years. If you want an alternative fuel, pay for a yard that has already delivered that fuel on that ship type, since a first-of-class at a yard that has not costs schedule as well as money, and schedule is what you are short of. If price is what matters and the trade allows it, order a repeat design at a Chinese yard, which is where the cost advantage and most of the available capacity are. If the containment system is the hard part, as it is for LNG and to a lesser extent ammonia, expect to pay a Korean premium and wait longer. And if you are ordering because freight rates are strong, be explicit with yourself that you are paying near the top of the price index and will take delivery into whatever market exists two or three years later, which is the trap that catches owners in every cycle.

China roughly half the global order book by CGT · Chinese yards 47% of the alternative-fuel-capable order book by CGT, Korea 42% · LNG carriers 53% of Korean yard workload by CGT in February 2025, containerships 22% · newbuild prices up about 46% since the start of 2021 against about 22% input cost inflation · berths largely sold into 2028–2029 · alternative fuel on 50% of tonnage ordered in 2024, 820 ships and 62.2m GT · about 70% of a shipbuilding contract's value sits in the supply chain

China State Shipbuilding Corporation and its yards including Hudong-Zhonghua and Jiangnan, plus private groups such as Yangzijiang and New Times; HD Hyundai, Hanwha Ocean and Samsung Heavy Industries in Korea, which between them carry most of the world's LNG containment work; Imabari Shipbuilding and Japan Marine United in Japan, building mainly for Japanese owners.

About 70% of a shipbuilding contract's value goes to the supply chain, mostly steel, the main engine, generators, cargo equipment and coatings, so the yard's own margin comes from design, steel processing, assembly and outfitting labor and is thin by industrial standards. That is why backlog matters more than volume to a yard's profit: prices up 46% since the start of 2021 against 22% input inflation is margin expansion, and it lasts only as long as the queue does. Buyers pay in installments against construction milestones, with a refund guarantee from the yard's bank covering the pre-delivery payments, and the split of those installments moves toward the yard when the order book is full. The green premium itself is modest against the base price: a 15,000 TEU ship runs about $180–200M conventional and $15–25M more dual-fuel, which is roughly 8–13% on top (that percentage is arithmetic on those two ranges, not a published figure). What decides fleet renewal is the queue rather than the premium. With the fleet turning over at roughly 3% a year and berths sold into 2028–2029, an owner who decides today to order an alternative-fuel ship is committing capital now for tonnage that starts trading at the end of the decade, so yard capacity is a harder constraint on shipping's emissions path than either the engine or the fuel.

11 minHow China Came to Dominate Global ShipbuildingBloomberg Originals · 1m+ views

7 minHow the World’s Largest Shipyard Is Challenging China’s Dominance | WSJThe Wall Street Journal · 1m+ views

19 minWhy Only Three Countries Bother Building Ships AnymoreMicro · 1m+ views

Market Intelligence Report, June 2025 (SEA Europe) · Green Technology Tracker: Record Investments in Alternative Fuel (Clarksons Research)