Sustainability of underground infrastructure – Part 3: TunCO, an open-source digital toolbox for accounting and optimising decarbonisation in tunnelling

• Open-source digital toolbox to streamline embodied carbon assessment of TBM tunnelling. • Whole lifecycle adaptation of TBM production, transport, excavation, and recycling. • Interface for carbon accounting, automated 3D modelling, and design optimisation. • Implementation of international databases, benchmarks, guidelines, and standard. • Evaluation and visualisation of carbon flow and reduction strategies in stages. As the infrastructure sector advances towards net-zero emissions, reducing embodied carbon in underground construction is essential, particularly for Tunnel Boring Machine (TBM) tunnelling. This paper introduces TunCO 2 , a digitalised framework that advances the practical integration of lifecycle carbon assessment with geotechnical design optimisation for TBM tunnelling. The framework extends prior carbon assessment models by incorporating the whole lifecycle of mechanised excavation as a quantifiable carbon source as TBM-specific adaptations in construction works. The proposed approach couples parametric modelling, carbon assessment, and geo-structural performance evaluation into a unified and repeatable methodology, enabling the explicit quantification of carbon–stability trade-offs in early design. Theoretical contributions include (i) the formalisation of a standardised carbon evaluation workflow directly compatible with geotechnical design practices, (ii) the introduction of a TBM lifecycle carbon module within parametric modelling, and (iii) a generalisable optimisation framework linking carbon intensity metrics to structural safety constraints. TunCO 2 ′s implementation featuring automated 3D modelling of segmental linings, integrated carbon factor databases, and instant scenario visualisation, demonstrates how these concepts can support low-carbon decisions in early project setting. The carbon database interface, segmental lining modelling module, and exemplary carbon–stability optimisation scripts are released open source to encourage replication and further practical refinement.