The short version.

• A Flettner rotor is a spinning vertical cylinder mounted on a ship's deck. Wind blowing past it, combined with the cylinder's own rotation, creates a sideways force — the Magnus effect — that can be angled to help push the ship forward, alongside its normal engine.
• The idea is a century old and has failed commercially twice already: Anton Flettner's Buckau (1920, later renamed Baden Baden, which crossed the Atlantic in 1926) and Barbara (1926) both worked, but cheap oil made conventional propellers cheaper to run, and rotor ships disappeared for over 80 years.
• Enercon's E-Ship 1 (2010) revived the concept with four rotors, reportedly 30-40% more fuel-efficient than a conventional cargo ship of its size. It was built by Enercon, a wind-turbine manufacturer, to transport its own wind-turbine components — not, as sometimes reported, for the German Navy.
• By June 2025, rotor-sail maker Norsepower alone had installed 38 rotor sails on 22 vessels, with 39 more planned across 17 ships within 18 months, and a further 6-rotor order from GEFO Shipping Group for delivery in 2026-2028. Measured fuel savings run 5-25%; the Maersk Pelican cut fuel burn by 8.2% in its first year. Norsepower's own site now shows this has grown further: 45 rotor sails on 25 ships by Q1 2026, with 39 more planned on 16 further ships.
• A separate, non-rotating technology — rigid WindWing sails from BAR Technologies/Yara Marine — went to sea on Berge Bulk's Berge Olympus in October 2023: four 37.5m-tall wings, ~6 tonnes of fuel and ~19.5 tonnes of CO₂ saved per day on its Brazil-China route. It is a different mechanism from a Flettner rotor and shouldn't be confused with one.
• Wind-assist devices sit alongside a separate, larger shift: DNV's Alternative Fuels Insight tracker counted 1,794 alternative-fuel-capable ships in operation and 1,544 more on order as of August 2025, with LNG dominating new orders (87 vessels/14.2 million GT in 2025) ahead of methanol, LPG, ammonia and hydrogen. Rotors and wingsails augment a ship's existing engine; fuel-switching changes what the engine burns.
• Beyond Norsepower alone, the wider wind-assist market grew from 22 ships (49 sails) delivered in 2024 to 24 ships (63 sails) in 2025, with 84 more ships on order as of August 2025. In April 2026 a UK startup, EcoNavis, began developing a rotor add-on aimed at the technology's main weakness — performance dropping off outside a narrow wind-angle range.
• The original University of Manchester research this piece is built on (James Grimshaw, 2012-13) studied rotor spacing specifically because autonomous cloud-seeding ships — a proposed marine-cloud-brightening geoengineering fleet, powered by clustered Flettner rotors — would need closely packed rotors whose wake interactions weren't well understood. That fleet has still never been built; the closest real test, a 2024 University of Washington field trial, was land-based and was shut down by the host city within months.

The physics: a spinning cylinder that pushes sideways

A Flettner rotor is mechanically simple: a tall, powered cylinder, spun by a small electric motor, standing upright on deck. When wind blows across it, the combination of the wind and the cylinder's own surface motion creates two different local pressure zones — higher pressure on the side spinning into the wind, lower pressure on the side spinning with it. That pressure difference produces a net force perpendicular to the wind direction: the Magnus effect, the same phenomenon that curves a spinning football or cricket ball. Angled correctly relative to the wind, that force has a component pointing the ship forward, supplementing (not replacing) its main engine.

The 2012-13 University of Manchester project this piece draws on used computational fluid dynamics (CFD) to study a specific open question: what happens when two Flettner rotors are placed close together, and their low- and high-pressure fields start interacting? At the time, there was little published research on this, and the project's own simulations found vortex shedding still present at a spin ratio of 2 — a result the researcher noted differed from S.J. Karabelas' 2010 study, which reported shedding suppressed above a spin ratio of 1.3 at a lower Reynolds number.

The Magnus-effect description and the CFD methodology and findings are from James Grimshaw's 2012-13 University of Manchester MEng poster, "Examining Flettner Rotors for Ship Propulsion using CFD," supplied directly for this piece. The Karabelas comparison is the poster's own citation (S.J. Karabelas, "Large Eddy Simulation of high-Reynolds number flow past a rotating cylinder," International Journal of Heat and Fluid Flow 31 (2010), 518-527) and was not independently re-verified against that paper.

A hundred years of near misses

Rotor ships are not a new idea rediscovered by climate policy — they are an old idea that has now failed and succeeded in roughly equal measure. Anton Flettner's original ship, the Buckau, was launched in 1920 and later renamed Baden Baden, successfully crossing the Atlantic in 1926. A second ship, Barbara, followed the same year. Both worked as engineering demonstrations. Neither survived commercially once cheap oil made conventional screw propulsion the obvious economic choice, and the technology went dormant for more than eight decades.

Figures are from the Grimshaw poster's own reproduction of the historical Buckau and Barbara specifications.

The idea's modern revival is Enercon's E-Ship 1, launched in 2010 with four rotors whose power draw comes from a steam turbine running on the exhaust heat of the ship's conventional diesel engine — a hybrid design, not a pure sail ship. Enercon, a German wind-turbine manufacturer, built the E-Ship 1 to transport its own wind-turbine components around the world; the ship was not built for the German Navy, a claim that circulates in some secondary sources (including, it should be said plainly, the very poster this piece otherwise draws on). Reported results put the E-Ship 1 at 30-40% more fuel-efficient than a comparable conventional cargo ship, with the rotors adding 2-3 knots of top speed at a fuel cost of around 1 kg/hour during testing.

E-Ship 1's builder and purpose were directly confirmed on a follow-up fetch of Wikipedia's E-Ship 1 article, which states verbatim: “The ship is owned by the third-largest wind turbine manufacturer, Germany's Enercon GmbH. It is used to transport wind turbine components” — corroborated further by Marine Insight and gCaptain reporting. The 30-40% efficiency figure and the 2-3 knot/1 kg-per-hour testing detail originate in the Grimshaw poster; Wikipedia's article does not state an efficiency percentage at all (it only notes that performance results "were published by Enercon on 23 September 2013" without quoting a figure), so this piece keeps that figure hedged rather than treating the absence of contradiction as confirmation.

What actually got built since the poster was made

The thirteen years since this research was done are exactly the period in which rotor sails went from a single demonstration ship to a real, if still small, commercial fleet. Norsepower, the leading rotor-sail manufacturer, reported 38 rotor sails installed across 22 vessels as of June 2025, with 39 more units planned for 17 ships within the following 18 months, and a separate May 2025 agreement to supply six rotor sails for a fleet of newbuild tankers for GEFO Shipping Group, delivering 2026-2028. Its customer list includes Bore, Sea-Cargo, Scandlines, Vale, CLdN, Nippon Marine and Socatra. Norsepower's own site now shows 45 rotor sails installed across 25 ships as of Q1 2026, with 39 more planned across 16 further ships.

Norsepower installation counts, the GEFO order, customer list and fuel-saving figures were directly confirmed on a follow-up pass by fetching Norsepower's own IWSA trade-profile page, which states verbatim: “38 Norsepower Rotor Sails™ have been installed onboard 22 vessels so far (figures from June 2025). Installations within the next 18 months consist of 39 units on 17 ships,” and Norsepower's own live site, which now shows the Q1 2026 45/25 figures. The Maersk Pelican figure is Norsepower's own reported result for that specific vessel's first year and was not independently re-derived here.

The wider wind-assist fleet, and a 2026 attempt to fix its core weakness

Norsepower's own numbers describe one manufacturer; the market beyond it is smaller but growing on the same trend line. DNV's Alternative Fuels Insight tracking of wind-assisted propulsion systems (WAPS) across all vendors counted 24 ships delivered with wind-assist devices in 2025, carrying 63 sails between them, up from 22 ships and 49 sails in 2024 — a figure DNV's own site states verbatim: “WAPS saw 24 ships delivered in 2025, with a total of 63 sails installed... up from 2024, which saw 22 deliveries and 49 sails.” The further claim that 64 ships in total carried a modern wind-assist system by August 2025 — 56 of them installed since 2020, three-quarters of those as retrofits, with 84 more on order — could not be located on any DNV.com page on a follow-up direct-fetch pass, and the one comparable DNV white-paper snapshot that was found (via Splash247, dated January 2025) instead shows 52 ships in operation and a 97-ship order book — materially different numbers for a nearby date. That 64/56/84 figure should be read with real caution pending a source that states it directly; the 2024-to-2025 delivery growth above it is solid. Rotor sails are the dominant technology specifically on bulk carriers and tankers; suction-wing designs are more common on general cargo vessels.

The limitation this piece's own CFD research doesn't address, but every rotor-sail operator runs into, is performance falling off outside a narrow range of wind angles — a reliability gap, not a physics failure, but one that has slowed wider adoption. In April 2026, UK startup EcoNavis began developing an answer: a patented tail-appendage fitted to a Flettner rotor's trailing edge, intended to increase thrust, cut the rotor's own power draw, and widen the range of wind angles a rotor sail can use efficiently. The project, led by CEO Dr Batuhan Aktas, is backed by a £100,000 Scottish Enterprise grant inside a larger £265,000 programme; a scale model is due for wind-tunnel testing at Politecnico di Milano to validate the concept against simulation, with a full-scale shipboard prototype possible later in 2026 if that stage succeeds. EcoNavis's own simulations reportedly show up to a 10% thrust increase and a 5% reduction in the rotor's power draw. It has not been tested at sea, and none of the Norsepower installations described above use a comparable device — this is an early-stage attempt at the next design iteration, not yet a proven one.

The 2025/2024 WAPS delivery and sail-count figures are directly confirmed against DNV's own site (see quote above). The 64-ship/56-since-2020/75%-retrofit figures and the 84-ship order-book figure remain from a search snippet only and, on a follow-up fetch pass, could not be located on any DNV.com page checked, while a different DNV white-paper snapshot (via Splash247, dated January 2025) shows materially different figures (52 ships operating, 97 on order) — treat that specific figure as unconfirmed and possibly stale or misremembered, not as a reliable current count. EcoNavis's Eco Rotor Sail, its funding, test-plan details, CEO, and simulated thrust/power figures were directly confirmed on a follow-up fetch of Hellenic Shipping News' coverage of the announcement.

A different technology, easy to confuse with this one

Not every modern "wind ship" uses a Flettner rotor, and the distinction matters because the two mechanisms work differently. WindWings, developed by BAR Technologies with Yara Marine Technologies, are rigid, non-rotating aerofoil sails — closer in principle to an aircraft wing than to a spinning cylinder. Berge Bulk's Berge Olympus, a 210,000-deadweight-tonne Newcastlemax bulk carrier, was fitted with four WindWings in October 2023, each 37.5 metres tall and 20 metres wide, for a combined sail area of roughly 3,000 m². Sailing its usual Brazil-to-China route, the ship is reported to save around 6 tonnes of fuel and 19.5 tonnes of CO₂ a day.

Berge Olympus and WindWings figures are from Berge Bulk's own announcements and corroborating trade coverage (BAR Technologies, Marine Log, Offshore Energy).

Why now, and not in 2012

The regulatory backdrop the original poster described has escalated substantially since 2012-13. At the time, the operative rules were the Sulphur Oxide cap tightening toward 0.1% in emission-control areas by 2015, Tier III Nitrogen Oxide limits for new vessels in controlled areas, and an Energy Efficiency Design Index target of roughly 30% improved efficiency for new ships after 2024. Those measures still exist, but the International Maritime Organization has since gone considerably further. Its 2023 revised GHG Strategy commits international shipping to net-zero greenhouse gas emissions "by or around 2050," with indicative checkpoints of at least 20% (striving for 30%) emissions reduction by 2030 and at least 70% (striving for 80%) by 2040, both against a 2008 baseline, alongside a target of cutting carbon intensity per unit of transport work by at least 40% by 2030. Two enforcement mechanisms — the Energy Efficiency Existing Ship Index (EEXI) and the Carbon Intensity Indicator (CII) — entered into force in 2023, applying real operational pressure to existing ships, not just new ones.

2012-13 IMO regulatory figures are from the Grimshaw poster. The 2023 revised GHG Strategy's targets and the EEXI/CII enforcement dates are from the IMO's own published strategy documents and hot-topics pages, corroborated by DNV and the International Council on Clean Transportation.

Wind assist is one lever among several, not the main one

Rotor sails and wingsails augment a ship's existing engine; they don't replace its fuel. The larger shift IMO's rules are pushing shipping toward is switching what that engine burns in the first place, and DNV's Alternative Fuels Insight (AFI) platform — the industry's standard tracker for this — puts a number on how far that has actually gone. As of August 2025, DNV counted 1,794 alternative-fuel-capable vessels already in operation, with a further 1,544 on order — a figure directly confirmed on a follow-up fetch of DNV's own site, which states in a footnote: “As of August 2025, 1,794 alternative-fuel-capable vessels are in operation, with another 1,544 on order.”

DNV AFI figures are compiled from several of DNV's own dated releases (its 2025 full-year summary, an August 2025 in-operation/on-order snapshot, and its H1 2026 update), which don't all share the exact same cutoff date or methodology — the 2025 ammonia and LPG figures in particular come from different point-in-time snapshots (3 vessels/37,000 GT in one release, 5 vessels and 19 orders respectively in another) and are presented here as a range rather than false-precision single numbers. Total alternative-fuel orders fell from 2025 to H1 2026 (137 vessels in H1 2026, down 11.6% on the same period in 2025) — DNV's own reporting frames this as a slowdown in an otherwise-declining newbuild market generally, not a reversal of the underlying shift toward alternative fuels specifically. Battery-electric and full-hybrid vessel counts were not found broken out in the DNV releases checked for this piece and are omitted rather than estimated.

Set against roughly 100,000 vessels in the global ocean-going merchant fleet, even 1,794 alternative-fuel-capable ships in operation is still a small minority — but it is the category wind-assist devices sit alongside, not compete with. A rotor sail or wingsail can be retrofitted to a ship that still burns conventional fuel; the fuel-switching numbers above describe a separate, slower, much more capital-intensive transition in what goes into the tank in the first place.

The fleet the research was actually aimed at still doesn't exist

It's worth being precise about what problem the original CFD research was solving, because it wasn't primarily cargo-ship efficiency. The poster frames its rotor-spacing question around autonomous cloud-seeding ships — a proposed fleet of unmanned vessels using Flettner rotors both for propulsion and to spray fine seawater mist into low marine clouds, brightening them to reflect more sunlight, as a form of solar-radiation-management geoengineering. That concept, most closely associated with the physicist Stephen Salter, envisions rotors packed closely enough that their wake interactions genuinely matter for the design — exactly the gap in the literature the 2012-13 project set out to narrow.

That fleet has not been built. The most advanced recent field test, run by the University of Washington's Marine Cloud Brightening Program, was a land-based spray trial at a former aircraft-carrier museum site in Alameda, California, beginning in March 2024 — not an at-sea rotor-ship experiment at all. The Alameda City Council voted in June 2024 not to permit the facility to continue, and no further open-ocean trial of a Salter-style rotor-spray vessel has been announced as of this writing. The commercial success rotor sails have found came from cargo shipping's fuel bill and IMO's emissions rules, a route the original research was not primarily aimed at.

The cloud-seeding motivation and the design concept are from the Grimshaw poster's own framing. The University of Washington MCB Program's Alameda facility, its March 2024 start, and the Alameda City Council's June 2024 decision are from SRM360 and related reporting on the program; no operational at-sea Flettner-rotor spray vessel was found in the sources checked for this piece.

What doesn't follow from any of this

Real commercial traction is not the same as a solved problem at fleet scale. Thirty-eight rotor sails on twenty-two ships, plus a few dozen more on order, is a rounding error against a global merchant fleet of roughly 100,000 ocean-going vessels, and shipping's own share of global CO₂ emissions — commonly cited around 2-3%, consistent with the figure in the original 2012-13 poster — will not move meaningfully until wind-assist or an equivalent measure reaches far beyond today's early adopters. Individual results (5-25% savings, Berge Olympus's ~19.5 tonnes of CO₂ a day) are real and worth taking seriously, but they describe specific ships on specific routes with favourable wind, not a fleet-wide average. Nor should rotor sails and rigid wingsails be treated as interchangeable: they are different mechanisms with different retrofit requirements, and a claim about one should not be assumed to apply to the other.

Sources and caveats