Author: Abdulbasit Muhammed-Jamiu, Data Analyst Intern
Introduction
For offshore renewable energy in particular, wave conditions influence asset performance, structural loading, operational planning and worker safety. Reliable wave data supports everything from optimising wave energy converter efficiency to determining safe weather windows for vessel transfers and scheduled maintenance.
Traditionally, wave properties are measured using dedicated instruments such as buoys and radar systems. While these technologies are highly effective, they are limited by spatial coverage, cost, and infrastructure requirements.
In this blog, we explore how Distributed Acoustic Sensing (DAS) offers a complementary approach. By using fibre-optic cables already installed on the seabed, DAS can be used not only for asset health monitoring, but also to estimate wave height, speed, and direction. At Indeximate, we are developing algorithms to extract these properties directly from subsea power cables as part of the Horizon Europe-funded Future Ocean Renewable Energy Systems Technologies (FOREST) project, which runs until autumn 2028.
Current methods
Before discussing our approach, we provide a brief overview of current methods of estimating wave properties, particularly focusing on the in-water methods. In this area, people traditionally rely on a couple of instruments, namely wave buoys and wave radars. These tools have proved highly effective, and provide extremely useful data to the interested communities. These methods do come with trade-offs though meaning new methods of detecting wave properties could provide additional value. Wave buoys, for example, are limited spatially. They return a point-source measurement which, given the costs associated with deploying and maintaining them, is often the only such measurement within several hundred kilometers. If a wave buoys breaks, therefore, it can lead to blind spots in wave understanding across large areas of the sea, potentially impacting on a wide variety of issues. Wave radars may offer a larger spatial coverage locally in some circumstances but, as with wave buoys, failures can be highly impactful. Having additional methods available to measure wave properties is, therefore, important. ‘Free’ methods (without additional infrastructure in the water) such as DAS, which utilises the cables that are already present on the seabed and would already be in place for other purposes, may be particularly promising in this area.
Our approach
At Indeximate, we utilise DAS to monitor the health of the subsea cables themselves. DAS has predominantly been used for event detection in the past, such as fault finding (including for subsea cables) and perimeter monitoring (e.g. footstep detection), however, more and more use cases are being found all the time. At its core, DAS turns a fibre-optic cable into thousands of virtual sensors, each separated by a short distance along the cable. These sensors can be measured thousands of times per second, and the resulting data can then be used to make meaningful insights in a range of areas. On top of cable health monitoringwhich is our primary product, this data can also be used for orthogonal products, including investigating wave properties such as heights, periods and directions. While DAS provides concentrated spatial measurements along a cable, significant data processing is required to estimate wave heights at different locations. We discuss some of these steps in this blog.
The method we use in extracting wave height using DAS is fascinating. Firstly, DAS data comes by default with numerous background entities, such as noise from ship vessels, mammals passing by, temperature etc. To ensure the accuracy of the wave periods, the waves are isolated from said background noise. By using data processing techniques such as Frequency Fourier Transform (FFT), we convert time signals from the time domain into the frequency domain. The next step is comparing our periodicity estimates to that of the wave rider buoy to guarantee the selected period range, used in estimating our wave height, directly correlates with the actual wave periods. This ensures only the correct frequencies related to wave periods are used. The filtered signals are windowed into 30-minute segments, this mimics the wave rider buoy methods that can be used for calibration. Using the selected period range, the concatenated data are bandpass filtered to isolate wave related periods. This gives us the filtered signals that we use to calculate our wave height.
The conventional method for calculating the significant wave height is by averaging over the highest one-third of the waves. Since we have access to extensive wave-related data across different sections of the cable, we are able to calculate this value over a greater distance than a wave buoy. To illustrate the resulting wave heights, Figure 3 plots the estimated wave heights in a 2D image with distance along the cable on the y-axis and time on the x-axis. A clear tidal periodicity of the waves that passed across the full length of our cable is present. Additionally, we can then extract individual channels and compare them against the wave buoy. Figure 4 shows an example of this plotted against the significant wave height from a nearby wave buoy. Clearly the two lines track each other, perfectly demonstrating that this process works. In this example, if the wave buoy failed, the measured wave height from the cable could be used to estimate the wave height before repairs could be conducted.
Future Work: Wave As A Service
Wave Speed and direction are additional properties that can be estimated using DAS. Although this is a current work in progress, that will be developed further through the FOREST project, we have achieved some important milestones so far. By constructing f-k diagrams, we can extract a variety of additional wave properties, which opens up different avenues through which we can approach the development of our output (wave height, wave speed, wave direction). These could include examples such as reducing unwanted background noise from our data or using the different axes of our resulting 2D FFT’ing array to get wave speeds.
These are all fascinating concepts and I hope you tune in for updates as we develop further through FOREST.
—————————————————————————————————————
The FOREST Project (forestproject.eu) is funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Climate, Infrastructure and Environment Executive Agency (CINEA). Neither the European Union nor the granting authority can be held responsible for them.
