BRENNEN

Industries / Transport and vehicles / Shipbuilding and repair

Sea trials

Before an owner accepts a new ship, the yard takes it to sea to show that it reaches the contract speed at the contract power, turns, stops and is quiet enough, while the owner and the classification society watch and record.

6

Milestones achieved, from 1895 to 2025.

2

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1895

    Trials find cavitation

    Charles Parsons launched Turbinia on the Tyne in 1894 to show his steam turbine at sea. With one propeller her trials disappointed, and Parsons traced the problem to cavitation: bubbles of vapor that form on a propeller turning very fast and rob it of thrust. Oosterveld of MARIN writes that Parsons first saw the need for a test facility at lowered air pressure and built his first cavitation tunnel in 1895. Refitted with three shafts carrying nine propellers, Turbinia reached 34.5 knots against 29.6 knots before, by the figures of the Tyne Built Ships record, and in 1897 she turned up unannounced at the naval review at Spithead.

    Lessontrials can expose physics no one designed for, and the test rigs built to study it become design tools.

    tynebuiltships.co.ukvliz.be

    Credit

    Charles Parsons

  2. 1969

    QE2 fails her trials

    Blades broke off the high-pressure turbines of Cunard's new liner Queen Elizabeth 2 during her sea trials, and as of 29 January 1969 Cunard had still not accepted the ship. Sir Arnold Lindley, appointed by the Minister of Technology, reported in March 1969 that the blades on both rotors had cracked from resonant vibration excited by steam from the nozzles, "a common design fault," with "no reflection whatever on the quality of workmanship," and that blades of a new form were needed before trials could resume. He added that "more work should be done in turbine development and instrumentation under running conditions."

    Lessonan owner's right to refuse a ship that fails its trials is what gives trials their force; machinery built once, as Lindley noted, gets little chance to be proved before it sails.

    api.parliament.ukapi.parliament.uk

    Credit

    Sir Arnold Lindley; the parties he named were Cunard, Upper Clyde Shipbuilders, John Brown Engineering (Clydebank), Lloyd's Register and the Board of Trade

  3. 2002

    Pass marks for maneuvering

    IMO's Standards for Ship Manoeuvrability, adopted on 4 December 2002 for ships built from 1 January 2004, set pass marks that trials check for ships 100 m and longer and for all chemical and gas carriers. In a full turn the advance (distance traveled in the original direction) may not exceed 4.5 ship lengths and the tactical diameter 5 lengths; zig-zag tests (swinging the rudder from side to side) limit how far the ship overshoots each turn; a full-astern stop must take no more than 15 ship lengths. Model tests and computer predictions may show compliance at the design stage, but full-scale trials are to confirm them.

    wwwcdn.imo.org

    Credit

    IMO Maritime Safety Committee (resolution MSC.137(76))

  4. 2014

    Noise checked at trials

    IMO's Code on noise levels on board ships became mandatory on 1 July 2014 for new ships of 1,600 gross tonnage and above (gross tonnage measures a ship's enclosed volume). Compliance rests on "the satisfactory conclusion of sea trials that result in issuance of a Noise Survey Report," measured with Class 1 sound level meters. Limits include 55 dB(A) in cabins on ships of 10,000 gross tonnage and above (60 dB(A) on smaller ships) and 65 dB(A) on the bridge (dB(A): decibels weighted to match human hearing).

    wwwcdn.imo.org

    Credit

    IMO Maritime Safety Committee (resolution MSC.337(91), adopted 30 November 2012)

  5. 2015

    One method for speed trials

    Trial speeds are corrected to the calm, deep-water conditions written in the contract, and MARIN says some of the corrections had not been revised in 30 or 40 years when shipowners, after what it calls "adverse experience," started the Sea Trial Analysis joint industry project (STA-JIP). STA analyzed trial results from 30 ships and, with 37 member companies, wrote a new method with new corrections for wind and waves (MARIN dates the industry standard 2005 in one publication and 2006 in another). The ITTC adopted it in 2012 and ISO built it into ISO 15016:2015, which IMO named the preferred standard for the trial speed used in a new ship's EEDI (Energy Efficiency Design Index, IMO's carbon rating for new designs).

    marin.nlmarin.nlmagazine.marin.nl

    Credit

    MARIN and the STA-JIP members, including A.P. Moller-Maersk and Shell; ITTC; ISO

  6. 2025

    Waves measured, not guessed

    After a four-year revision by a working group of 40 experts, ISO 15016:2025 tightened the wind corrections and allows higher wind speeds when wind is measured by lidar or 3D ultrasonic sensors. Unless waves stay below a set height, the wave spectrum must now be measured by a wave buoy, a ship-borne wave radar or a drone. A 1965 shallow-water correction by Lackenby was replaced by a method from Hoyte Raven, based on CFD, model tests and full-scale trials of three ships at three or four water depths. IMO's environment committee (MEPC 83) accepted the standard for EEDI speed trials in April 2025, and the free STAIMO 3.0 program from the STA group follows it.

    magazine.marin.nlstaimo.com

    Credit

    ISO; MARIN and the STA group; IMO

Watch for

  1. Speed within 0.1 knot

    MARIN says its two decades of work on speed trials aimed at a standard that assesses ship speed within 0.1 knot and propulsion power within 2%, and that STA-2, a follow-up project started in April 2025, will add a protocol for measuring ships in service.

    Achieved whenan independent study, with two trial teams or analysts working on the same ships, publishes agreement within 0.1 knot and 2%.

    magazine.marin.nlmarin.nl

    Claimed by

    MARIN, June 2025

  2. Autonomy on builder's trials

    Jiangxi Jiangxin Shipbuilding took the 740 TEU battery-electric container ship Ning Yuan Dian Kun on sea trials off Shanghai from 6 to 13 February 2026, planning to test its battery supply, propulsion and autonomous navigation systems; the yard bills it as the largest ship of its kind.

    Achieved whenthe owner or the classification society publishes what the autonomous navigation system did on trials, or a class notation is issued for it.

    maritime-executive.com

    Claimed by

    Jiangxi Jiangxin Shipbuilding, as reported by The Maritime Executive, February 2026

Open

  1. 01

    Maneuvering trials on a model

    a radio-controlled model, about 2 m long, of the KCS benchmark container ship that runs the IMO turning-circle and 10°/10° and 20°/20° zig-zag tests by itself on a pond or pool, logs position and heading, and computes advance, tactical diameter and overshoot angles in the format of a trial report.

    The brief
    Credit · unclaimed
  2. 02

    A cheap noise survey kit

    a phone or single-board computer with a calibrated external microphone that follows the IMO Noise Code's measuring procedure (A- and C-weighted levels, readings held at least 15 s, rounded to the nearest decibel) and writes a draft noise survey table, tested against a Class 1 sound level meter in the passenger spaces of ferries.

    The brief
    Credit · unclaimed

Open 01 · the brief

Maneuvering trials on a model

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

a radio-controlled model, about 2 m long, of the KCS benchmark container ship that runs the IMO turning-circle and 10°/10° and 20°/20° zig-zag tests by itself on a pond or pool, logs position and heading, and computes advance, tactical diameter and overshoot angles in the format of a trial report.

Why it is still open

IMO's standard lets model tests show compliance before a ship is built, as long as full-scale trials then confirm it, and free-running model tests of KCS (the model steering itself rather than being held by a carriage) have been run at large basins such as SVA Potsdam and BSHC for the SIMMAN workshops. Research groups have built self-steering models that run maneuvering tests on a lake, such as Moreira and Guedes Soares at the Technical University of Lisbon, who reported lake trials without comparing them to benchmark data. Searches in November 2026 found no open, low-cost kit that anyone can build and check against the published KCS data.

wwwcdn.imo.orgsimman2008.dkjmr.unican.es

Built with what exists

KCS hull geometry from the SIMMAN site, a 3D printer or foam mold, ArduPilot Rover firmware (which supports boats) with Lua scripts for the rudder sequences, a GNSS receiver and motion sensor, and Python for analysis.

  1. First step

    download the KCS geometry and test program from the SIMMAN 2008 site and print one hull section at 1:100 scale.

  2. Done looks like

    20 runs of each test with the spread between runs reported, and overshoot angles and turning dimensions compared with the SIMMAN free-running results at matching scale and speed.

Open 02 · the brief

A cheap noise survey kit

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

a phone or single-board computer with a calibrated external microphone that follows the IMO Noise Code's measuring procedure (A- and C-weighted levels, readings held at least 15 s, rounded to the nearest decibel) and writes a draft noise survey table, tested against a Class 1 sound level meter in the passenger spaces of ferries.

Why it is still open

The Noise Code requires integrating sound level meters to IEC 61672-1 class 1 and calibrators checked every two years by an accredited laboratory, so surveys need specialist equipment. A search in November 2026 found published comparisons of phone sound-meter apps with Class 1 meters, but only on land, none on board a ship.

wwwcdn.imo.org

Built with what exists

a smartphone or Raspberry Pi, a calibrated USB measurement microphone, an acoustic calibrator, open-source sound-level code in Python, and a rented Class 1 meter for the comparison.

  1. First step

    rent a Class 1 meter and take 10 paired readings in the public areas of one ferry crossing, with the crew's permission.

  2. Done looks like

    at least 50 paired readings in 10 or more spaces on three ferries, with the mean difference and 95% limits of agreement against the Class 1 meter published; a target is agreement within 2 dB(A).