It is not easy to develop a method that can predict how much underwater noise a certain type of ship design will generate. This is shown in a new report from Lighthouse and The Transport Administration. However, the study also produced an unexpected result that needs to be followed up.

The emission that cannot be seen—underwater noise—is a growing environmental problem and is classified as a pollutant by both the UN and the EU. In recent years, the issue has received increasing attention as more and more reports show how noise from shipping disturbs marine life. So what if it were possible to predict how much underwater noise a ship will generate before it is built? In the innovation project Prediction Methods for Radiated Ship Noise – PUB, carried out within The Transport Administration's industry program Sustainable Shipping, which is run by Lighthouse, researchers investigated how well today’s existing methods perform in this regard.

“We used three different methods. One at model scale, where we used a two-meter-long ship in a test facility; one using numerical simulations at both model and full scale; and one where we used an existing vessel—a coastal tanker—and measured its actual underwater noise at sea outside Vinga. We compared how well the numerical analyses and model-scale experiments matched the noise emissions from the real vessel,” says Carl Andersson, a researcher at IVL and one of the authors of the report.

It turned out to be far from straightforward to achieve agreement between the results from the different methods.

“When it comes to the model-scale experiments, it was difficult to recreate the conditions experienced by a ship at sea. Some phenomena simply cannot be reliably scaled down. This applies in particular to the cavitation dynamics around the propeller, which do not behave the same way at model scale as they do at full scale.”

In the numerical models, the analyses primarily focus on the noise generated by the ship’s propeller (which has so far been believed to produce most of the noise), rather than the noise generated by the engines.

“At the speeds at which we measured the underwater noise from the existing vessel, there was not much cavitation from the propeller. Instead, noise from the engine itself dominated, which made the comparison between the numerical models and reality completely skewed. That the propeller did not generate more noise was surprising,” says Carl Andersson.

In a follow-up project that is already under way, measurements will be carried out on several existing vessels, along with numerical analyses and model-scale tests of additional ships. Instead of conducting onboard measurements, the researchers will use a measurement system placed outside Galterö, next to the shipping lane leading into Gothenburg. This allows measurements to be taken without personnel being present on site.

“We will also study specific noise-reduction measures. The goal is that even if we cannot predict a noise level with absolute accuracy, we should at least be able to correctly predict the effect of a noise-reducing measure,” says Carl Andersson.

The report Prediction Methods for Radiated Ship Noise – PUB was authored by Rickard Bensow and Qais Shehadeh Khraisat (Chalmers University of Technology), Torbjörn Johansson and Carl Andersson (IVL), Mathias Andersson, Markus Linné (FOI), Jan Hallander and Mattias Liefvendahl (RISE), Anders Kämpeskog (Kongsberg), Martin Persson (Kongsberg Maritime Sweden AB), and Peter Sigray (KTH Royal Institute of Technology).
In collaboration with: Kongsberg Maritime Sweden AB and Terntank AB