Coupled effects of settling and cable snaking on backfill properties and heat transport in underground HVDC cable systems

The expansion of the German power transmission network represents one of the most significant challenges of the energy transition. The shift in the preferred expansion method towards underground cables has further complicated this process. The primary issues arise from a lack of experience in constructing underground cable trenches and insufficient knowledge about the operational impacts on the surrounding soil. The first part of this study presents the in-situ compaction achieved in a cable-laying test. A sampling and laboratory programme developed for this purpose is introduced, aiming to enable a standardized investigation and qualitative assessment of bedding and backfill materials as well as installation quality. The collected data are subsequently used to calculate the long-term settling behaviour. The second part of the study introduces an application for calculating thermally-hydraulically coupled explicit multiphase flow, which is experimentally validated. The application is based on the generalised Darcy's law and uses saturation-based relative permeability and matrix potential according to van Genuchten for multiphase flow. Additionally, phase miscibility is accounted for using Henry's and Raoult's laws as well as mass diffusion coefficients. The third part of the study examines cable deformation due to thermal expansion (snaking) and its impact on heat transport within the duct and the surrounding soil. Using the application developed in Part Two, four potential cable positions were simulated in comparison to the frequently assumed centred position, for two different cable spacings and conductor temperatures. The results from Part One indicate that compaction trends are consistent regardless of the bedding material, showing increased or decreased values at specific cross-sectional positions in the trench. Notably, the gusset zones are particularly affected, with higher compaction observed in the outer gusset relative to the trench centre compared to the inner gusset. Calculated settling for non-cohesive soils was expectedly negligible, whereas for cohesive soils, a reduction in porosity of up to 1.6 % in the gusset zone and 2.6 % in the cable zone was determined, with settling of up to 8 mm near the duct. The application from Part Two was successfully validated for temperature calculations. However, the calculation of diffusion processes did not yield satisfactory results for hydraulic processes due to missing thermal diffusion coefficients. In Part Three, it was demonstrated that assuming a centred cable position within the duct leads to an underestimation of both the intensity and spatial extent of soil heating. Furthermore, thermal gradients in the insulation and the resulting ageing processes are neglected, potentially leading to an overestimation of the cables' service life. A combined model incorporating insights from Parts One through Three showed that for specific soils long-term settlement within the trench improves thermal properties, compensating for the disadvantages of poorly compacted gusset zones. Therefore, it is recommended to account for settling during route planning. Moreover, when considering preventive measures against snaking, the cable position at the bottom of the duct should be modelled preferably instead of assuming a centred position. Future research needs to include studying the settling behaviour of shallow backfills with cohesive and non-cohesive materials under self-weight. Finally, the development of a 3D coupled thermo-hydro-mechanical (THM) model is recommended, integrating thermal-hydraulic transport processes, trench settlement behaviour, and snaking formation.