Author: Tom Bache PhD, Data Scientist
At Indeximate, we use Distributed Acoustic Sensing (DAS) alongside proprietary compression and analytics to detect and predict the location and causes of these stresses.
DAS transforms a standard optical fibre within a subsea power cable into a continuous array of thousands of highly sensitive, real-time vibration sensors. An interrogator unit located at the landing station sends short pulses of laser light along the fibre, continuously analyses the small fraction of light that naturally scatters back.
When acoustic waves or physical vibrations (such as those from the cable moving across the seabed or from tidal-induced vibrations) strike the fibre, microscopic changes in strain occur within the glass core. These changes alter the backscattered light signal, allowing the system to detect disturbances and locate the issue across tens of kilometres of cable route. For cable operators, this enables a shift from reactive fault response to proactive monitoring, helping identify emerging risks before they develop into major failures.
In previous blogs, we discussed how we identify strain and fatigue, as well as issues affectingthe integrity of cable armour. In this blog, we focus on one potential cause of increased strain and damage: a type of abrasion known as skipping.
Unlike gradual abrasion that develops over long timescales, skipping occurs when tension issuddenly released, producing a short, high-intensity event that appears in the data almost like an impact.
For example, skipping can occur when the cable being pushed by strong currents becomes temporarily caught on a rock and is then suddenly released. Each event introduces additional strain on the cable and increases the chance of abrasive damage. If repeated frequently, thiscan have serious consequences for cable health. Although subsea cables are typically armoured with steel wires, repeated high-impact abrasion can ultimately lead to armour damage and failure.
Why skipping occurs
To understand skipping, it is helpful to consider static friction. Static friction is the resistive force that keeps a stationary object at rest. It acts equal and opposite to an applied force until a critical threshold is reached.
In the case of a cable resting against a rock example, ocean currents exert force on the cable.This force increases until it pt exceeds the frictional threshold, at which point the cable suddenly moves, effectively “skipping” across the rock.
Because the cable is under tension when this occurs, the sudden movement produces a rapid change in strain. This appears as a high impact event in DAS data, making skipping relatively easy to detect within the strain signal.
The prevalence of skipping often depends on weather conditions. It is particularly prevalent during high tidal flows as well as storms. One important consideration is that skipping is only expected to be observed when the cable is exposed. If buried, the seabed stops any largescale movement, meaning that if it is observed, the cable is highly likely to be exposed. Of course, not all cables can be buried, and skipping is particularly dangerous on rocky seabed in high tidal flows. See our blog on exposure for further details.
What does skipping look like in the DAS data?
Given skipping’s nature as a sudden release of tension, it appears as a sharp increase in strain over a short period of time. Since DAS effectively measures change in strain along the cable, we can identify these events by looking for rapid strain spikes within the dataset.
An example of this is shown in the figure below, illustrating the change of strain for a specific section of subsea cable. The red band highlights the detected skipping event.
When we run this skip-finding algorithm over months of data across the thousands of channels present on subsea cables, we get an indication of how often these skipping events occur, and where. We translate this into a daily average number of skips per 10m section of cable, as shown in the below figure. A 10km cable section is shown, where a number ofsections of cable are experiencing significant levels of skipping (more later on what “significant” means).
This leads to the key question: how much skipping leads to damage?
The honest answer is that it depends. Each cable installation and seabed is different, meaning the relationship between skipping frequency and damage risk can vary. However, by analysing historical data from cables where damage locations are already known, we can estimate the probability of damage associated with different levels of skipping.
Using a probabilistic approach is important because
By analysing the number of skipping events observed in DAS data alongside known damage locations, we can estimate how damage probability increases as skipping frequency rises.
The below figure shows how we do this in practice. The horizontal axis represents the average number of skips required to reach a certain damage probability (shown on the vertical axis), as calculated from the damage “truth” data and the number of skips we observe in the DAS data. The grey points show the probability of damage calculation (with error bars equal to the 95% confidence intervals), performed in bins according to the number of skips. We then fit a logistic regression curve to these points, as shown by the green line, which allows us to calculate, for some amount of skips, what the damage probability is.
How can we protect against skipping?
At Indeximate, we provide cable health monitoring through our Scattersphere platform. Scattersphere is a non-intrusive, hardware-agnostic system that replaces isolated alarm eventswith continuous analysis of cable behaviour.
By analysing long-term datasets, the platform highlights emerging risks before they develop into failures. Skipping is one such phenomenon that than be detected and tracked.
Scattersphere uses cloud-based analytics and proprietary data compression to process very large datasets efficiently, allowing continuous monitoring of cable behaviour across long time periods. This enables operators to understand how their cables respond to real-world conditions and to plan maintenance or intervention before damage occurs.
Subscribers can access a range of risk metrics through the platform, helping them prioritise maintenance activity and reduce the number of unplanned repairs. One such customer used our technology to identify the precise location of skipping on their cables, then performed an intervention to fix the problem. We’ll have to leave that as a cliffhanger for now though – keep your eyes peeled for details on this in a future blog.
If you are interested in learning more about skipping, or our services in general, please reach out here.
