Development and Validation of a Simplified 2D Digital Twin for a Large-Scale Marine Debris Cleanup System
Year: 2026
Type: conference-paper
Abstract: In [1], we proposed a 3D Digital Twin (DT) of our U-shaped plastic marine litter cleanup system, constructed from repurposed fishing nets. The system is characterised by a very large length-to-depth ratio of 𝒪(103). The first version of the DT was based on Finite Element Method (FEM), a model developed with OrcaFlex, which was further enriched with data from the Great Pacific Garbage Patch (GPGP), as well as detailed Computational Fluid Dynamics (CFD) simulations of the system. In this study, we introduce a faster 2D simplified model for faster decision-making. This model is based on piece-wise catenary curves and utilises the Newton-Raphson method to establish equilibrium between internal and external forces within the netted system. Similar to the OrcaFlex model, the 2D model was calibrated using experimentally acquired drag coefficients of a full-scale fishing net. The accuracy of the model was assessed in terms of Mean Absolute Error (MAE) for the hydrodynamic load on the system, considering offshore missions between September 2022 and September 2023. The results indicate that the 2D model performs as well as the 3D model while reducing computational time from hours to minutes. The points of comparison were defined as a function of the span (S) to system length (Ls) ratio (S/Ls). They are categorised into three groups of: i) narrow span (S/Ls ≤ 0.2), ii) medium span (0.2 < S/Ls ≤ 0.6) and iii) wide span (S/Ls > 0.6). Furthermore, our Towing Efficiency Model accurately predicts tow tension, propeller pitch settings, bow thruster usage, rudder angle and fuel consumption, ultimately enabling reliable estimation of fuel requirements. Validation across all cleanup extractions from 2021 to 2024 yielded an average error of 12%.
Institutions: Delft University of Technology, The Ocean Cleanup
Language: English
Subfield: Ocean Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7214664149