In the global energy transition process, offshore wind is rapidly emerging as a crucial energy source. This shift has led to an increasing demand for submarine cables to connect offshore wind farms to onshore substations, with cable installation costs representing 9% of the total installation expenses. Cables are buried in the seabed, primarily through cable ploughing, adopting a special plough to create narrow trenches up to 3 m deep in the seabed. Accurate prediction of the vessel tow force is essential for efficient cable installation design, with the tow force being influenced by soil type, burial depth, and target velocity. Currently, available analytical approaches for sands rely on empirical correction factors, leading to inaccuracy in application across different scenarios. Advanced numerical methods could be in principle adopted to study the complex hydromechanical system response, but at the cost of significant computational resources and with the limitations related to field soil characterization. To address these limitations, artificial intelligence (AI)-based modeling has emerged as a promising alternative in many engineering fields, where wide data sets are available, as in cable ploughing applications due to the abundance of operational data. This study provides a comparison of the predictive capabilities of three literature analytical models against new field data from three different cable ploughing projects in sands, highlighting their limitations and proving the potential of the support vector machine (SVM) regression models in enhancing the prediction of the tow force. Remarkably, the developed SVM model only requires input parameters that are available from standard offshore geotechnical investigations, like cone penetration test results and soil granulometry. The study is based on 113 km of cable ploughing data, emphasizing the applicability of adaptable predictive models in the field of offshore cable installation.

Offshore Cable Ploughing in Sands: Comparison of Literature Models and Machine Learning–Based Models with Field Data

Marveggio, Pietro;Gritti, Davide;Mariani, Stefano;Della Vecchia, Gabriele
2026-01-01

Abstract

In the global energy transition process, offshore wind is rapidly emerging as a crucial energy source. This shift has led to an increasing demand for submarine cables to connect offshore wind farms to onshore substations, with cable installation costs representing 9% of the total installation expenses. Cables are buried in the seabed, primarily through cable ploughing, adopting a special plough to create narrow trenches up to 3 m deep in the seabed. Accurate prediction of the vessel tow force is essential for efficient cable installation design, with the tow force being influenced by soil type, burial depth, and target velocity. Currently, available analytical approaches for sands rely on empirical correction factors, leading to inaccuracy in application across different scenarios. Advanced numerical methods could be in principle adopted to study the complex hydromechanical system response, but at the cost of significant computational resources and with the limitations related to field soil characterization. To address these limitations, artificial intelligence (AI)-based modeling has emerged as a promising alternative in many engineering fields, where wide data sets are available, as in cable ploughing applications due to the abundance of operational data. This study provides a comparison of the predictive capabilities of three literature analytical models against new field data from three different cable ploughing projects in sands, highlighting their limitations and proving the potential of the support vector machine (SVM) regression models in enhancing the prediction of the tow force. Remarkably, the developed SVM model only requires input parameters that are available from standard offshore geotechnical investigations, like cone penetration test results and soil granulometry. The study is based on 113 km of cable ploughing data, emphasizing the applicability of adaptable predictive models in the field of offshore cable installation.
File in questo prodotto:
File Dimensione Formato  
marveggio-et-al-2026-offshore-cable-ploughing-in-sands-comparison-of-literature-models-and-machine-learning-based.pdf

Accesso riservato

: Publisher’s version
Dimensione 5.5 MB
Formato Adobe PDF
5.5 MB Adobe PDF   Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/1323569
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
  • OpenAlex ND
social impact