Compressive Wave Phenomena in Subsea Power Cables

Abstract

This thesis examines the occurrence and interpretation of compressive wave phenomena in subsea power cables during normal lay operations, with a focus on addressing a central uncertainty in the field. In current industry practice, the compressive behaviour predicted by numerical cable-lay simulations is frequently regarded as a limiting factor for installation, however, a significant research gap remains regarding whether these compressive waves correspond to actual physical cable responses or are merely artefacts of modelling assumptions. Accordingly, the primary aim of this research is to clarify this gap by evaluating whether the compressive waves observed in OrcaFlex simulations can be substantiated as phenomena that occur during normal lay operations.

To address this question, a comparative modelling, verification, and field-data assessment approach is adopted. First, the theoretical possibility of axial compression is demonstrated using a simplified Kirchhoff rod formulation, showing that time-dependent transverse excitation can induce compressive forces near the touchdown region. Subsequently, a Python-based finite element model is developed to capture both static and dynamic cable responses and is compared with OrcaFlex simulations. The validation strategy proceeds stepwise: starting with analytical catenary theory, followed by static verification across multiple cable types, and finally dynamic comparison using measured operational input data.

This thesis investigates the occurrence and interpretation of compressive wave phenomena in subsea power cables during installation operations. In industry practice, the compressive behaviour predicted by numerical cable-lay simulations is often treated as a limiting factor, although it remains unclear whether these compression waves reflect a physical cable response or are partly due to modelling assumptions. The objective of this research is therefore to assess whether the compressive waves observed in OrcaFlex simulations correspond to phenomena that can occur during normal lay operations.

To address this question, a comparative modelling, verification, and field-data assessment approach is adopted. First, the theoretical possibility of axial compression is demonstrated using a simplified Kirchhoff rod formulation, showing that time-dependent transverse excitation can induce compressive forces near the touchdown region. Subsequently, a Python-based finite element model is developed for both static and dynamic cable response and is compared with OrcaFlex simulations. The verification strategy proceeds stepwise: starting with analytical catenary theory, followed by static validation across multiple cable types, and finally dynamic comparison using measured operational input data.

The simulations are driven by manufacturer cable data and operational data obtained from DEME's cable-lay vessel Living Stone, including vessel motion data from Grafana and onboard sources, as well as ROV observations of the touchdown region. Within the simplified Python model, vessel motions are translated into chute boundary motions, allowing measured operational data to be applied to the cable model in a controlled manner. And the touchdown point is added as an extra boundary condition to verify the catenary shape.

However, an important limitation of this approach is that no sufficiently complete dynamic field dataset was available for the specific severe conditions in which compressive waves are observed in OrcaFlex. As a result, the Python model could only be verified for relatively mild or near-static conditions, and not for a measured case in which the suspected compressive-wave behaviour occurred. Consequently, the combined Python model and vessel data approach is not sufficient on its own to fully answer the main research question.

For that reason, an additional tension analysis is carried out using recorded cable-lay data. Time periods corresponding to normal lay operations are selected and screened based on the availability of reliable tensioner measurements. Particular attention is given to tension peaks, irregularities, and intervals in which negative tension values are observed, as these may indicate conditions relevant to compressive behaviour. These selected cases are subsequently analysed in OrcaFlex using the appropriate vessel motions and cable properties, while the corresponding ROV footage is reviewed in parallel to investigate whether any visually observable compressive behaviour can be identified.


The tensioner-selected cases did not show clear travelling compressive waves in the corresponding OrcaFlex simulations. Although several cases contained irregular or negative tensioner readings, these did not develop into sustained regions of axial compression near the touchdown point. This indicates that negative tensioner values should not be interpreted directly as evidence of compressive-wave behaviour in the suspended cable span, since the tensioner measurement represents a differential force across the onboard tensioner system rather than the true top tension of the free-hanging cable.

Because the project-data-based cases did not reproduce compressive waves, an additional parameter-variation study was conducted. The bending stiffness, touchdown point location, and imposed chute motion were varied to investigate which factors can trigger the phenomenon. The results show that changing the bending stiffness or touchdown point location mainly influences the local cable response but does not, in itself, generate travelling compressive waves. The imposed chute motion, and especially the resulting chute velocity, was found to be the dominant trigger. Compressive waves were only reproduced when the imposed boundary motion became sufficiently dynamic.


The thesis, therefore, combines theoretical analysis, numerical model development, validation, tensioner-based case selection, and parameter variation to investigate compression in subsea power cables from multiple perspectives. The results show that compressive waves were not observed in the analysed normal lay operations. However, the phenomenon cannot be entirely ruled out as physically possible under more severe dynamic excitation. The main conclusion is that the available project data were not dynamic enough to reproduce compressive waves. At the same time, the numerical parameter study indicates that sufficiently high chute velocity can trigger the phenomenon. Further validation requires complete operational datasets from more dynamically severe installation conditions.