Accurate observation of ground conditions at the tunnel face is essential for defining temporary support in drill-and-blast tunneling.

Traditional geological face mapping relies on geologist observations which present limits in term of precision and consistency. Georeferenced 3D models generated directly at the face offer a faster, safer, and more accurate alternative, improving documentation quality and design decisions.

On the TELT C008 project, Implenia and partners standardized tunnel face mapping by photogrammetry. This large-scale 3D geological model enables early detection of anomalies, supports back-analysis, and helps refine support strategies. It also provides measurable records for verifying support installation, profiling shotcrete thickness, and improving quality control.

Photogrammetry simplifies the production of as-built documentation, unfolded views, stereograms, and accurate overbreak evaluations. It further contributes to a digital twin integrating all construction stages within a BIM-compatible framework.

The paper presents implementation methods, challenges, and the evolving role of photogrammetry as a multi-purpose digital tool that strengthens safety, accuracy, collaboration, and long-term asset management.

--- 

Artificial intelligence (AI) is profoundly transforming engineering professions, and the underground works sector is no exception. This trend is all the more significant in this field, where the technical challenges, human risks and environmental constraints are particularly high. 

By integrating AI, project managers and contractors can optimise planning, improve safety and increase productivity on underground sites. 

One of the main advantages of AI is its ability to perform predictive analyses. By exploiting geotechnical, historical and real-time data, the algorithms are able to anticipate ground movements, detect structural anomalies or predict wear and tear on equipment. This capability helps to avoid accidents, reduce unscheduled stoppages and optimise digging methods according to ground conditions. 

We present four use cases where AI is being used operationally on construction sites. These are: 

Detecting the clogging of a TBM; 

Predicting settlement as work progresses; 

Using a ‘Ground Machine Interaction’ indicator for tunnel boring machines; 

Optimising the advance speed of an earth pressure tunnel boring machine as a function of geology. 

These four use cases illustrate how AI is currently being deployed on real construction sites, and help us to understand its advantages and limitations. 

Finally, an analysis of the regulatory and legal issues associated with the use of AI and data is proposed.

--- 

Digitalization in tunnelling is accelerating—but often in directions that overlook the real levers for improvement. While advanced solutions like AI-powered monitoring and digital twins are being developed for complex, infrequent edge cases, the day-to-day efficiency losses are frequently underprioritized. In practice, project teams struggle not only with slow or fragmented information flows, but with limited access, poorly structured interfaces, and non-intuitive data representations. 

This contribution calls for a redirection: toward pragmatic, human-centered IoT platforms that focus on simplifying the operational reality—not adding layers of complexity. Interfaces must be clean, logic-driven, and role-specific. Transitions between digital systems and physical processes should be seamless and standardized, not improvised. AI bots and context-aware logic can deliver real value—only when embedded within accessible, well-designed tools. 

Bridging OT, IT, and IoT environments in tunnelling requires more than integration; it demands design thinking. The goal isn’t just data collection—it’s clarity, speed, and relevance. By prioritizing user experience and modular standards, digitalization can finally become a tool that accelerates the site instead of slowing it down. 

 

--- 

While Building Information Modelling is becoming increasingly widespread in underground infrastructure projects, the use of Digital Twins for tunnel operation remains at an early and largely experimental stage. Their ability to enhance asset management, support incident analysis and maintenance strategies, and ensure seamless digital continuity has yet to be demonstrated under real operating conditions. In this context, a collaboration has been initiated between the Centre d’Études des Tunnels (CETU) and the Société Française du Tunnel Routier du Fréjus (SFTRF) to design, test, and evaluate a digital twin applied to an operational road tunnel. The Fréjus Road Tunnel, consisting of two tubes, one in service since 1980 and a second recently opened in 2025, provides an ideal case study. The coexistence of legacy and modern structures, combined with technical systems, offers a realistic environment for experimenting with DT concepts. 

The objective of the project is to develop a functional digital twin integrating both static and dynamic data. The first phase focuses on BIM modelling of a representative section, with particular emphasis on equipment, especially the ventilation system, whose digital behaviour aligns closely with existing DT applications in the building and industrial sectors. Subsequent stages will gradually extend the scope to additional equipment and, later, to the civil engineering structures. 

This initiative forms part of the CETU’s broader digital transformation strategy and aims to deliver concrete feedback for the tunnelling community. It also seeks to identify the technical and organisational barriers specific to underground assets and propose operational solutions to overcome them. 

Predicting the impact of tunnel boring machine excavation on deep foundations is still a current topic in urban tunnelling. Such predictions usually necessitate a numerical analysis of the interaction between the TBM, soil and foundation due to the complexity of the mechanisms involved. This paper presents a numerical model recently developed by ENTPE and CETU to address this issue. While this 3D numerical model is based on well-established methods for modelling TBM–soil interaction, the originality of this work lies in the way the modelling choices and boundary conditions were defined, calibrated, and validated. These choices were intended to replicate the action of the TBM in the field and were informed by full-scale experimental observations from the TULIP research project. The modelling procedure was developed based on a rigorous phenomenological analysis of the observed responses of the soil and piles, including surface settlements, subsurface displacements, and pile behaviour (displacements and forces).
The developed model satisfactorily reproduces the soil displacements and pile–soil interaction mechanisms observed at the TULIP experimental site under both quasi-stationary excavation conditions and transient accidental events (pressure drops in the excavation chamber, increase of the grouting pressure). The study also emphasises the importance of certain modelling choices, such as using a nonlinear elastic–plastic soil model (HSM type) and implementing an elastic–plastic pile–soil interface model with normal and tangential stiffness and a sliding threshold calibrated from qs values. It also highlights the need to explicitly consider confined water–soil fluid along the shield. --- Urban centers in major cities worldwide are progressively densifying, due to the enormous population growth over the last years. These changes strain existing infrastructure, requiring expansion or construction of new underground transportation systems, for example in Paris and Toulouse. , tunnel excavation near adjacent structures and in particular piled structures outcomes could be significantly limited or controlled by predicting foundations behaviour during and following tunnel construction as well as operations phases. In particular, it would help to limit financial provision for constructions phases and improve environmental performance of the infrastructure project. In this context, the Gustave Eiffel University has set up a research project called E-PILOT (Study of the impact on Piles During the passage of a Tunnel Boring Machine), funded by the ANR, which brings together 10 partners (the Gustave Eiffel University, the University of Lille (LGCgE), the Cerema, the CETU, the ENTPE, GDS, Itech, Terrasol, SGP, Tisséo Ingénierie) to deepen the understanding of the mechanisms of interaction tunnel boring machine-soil-foundations. The research work proposed by the consortium will help to manage impact on existing structures during tunnelling and tunnel operation phase in the framework of actual important infrastructure projects in a metropole area. --- 

The calculation of settlements induced by tunnels excavated with a TBM  is carried out using two types of methods:

empirical methods, with all the limitations of this type of "black box" method, based on an assumed "ground loss" factor ;

numerical methods, adapted to various geotechnical configurations, representing the soil-tunneling machine interaction by a deconfinement rate adapted to the excavation and confinement method.

Analytical methods exist, but they generally start from an assumed deformation of the excavated surface (mean convergence and ovalization), which limits their use since this deformation is unknown and refers to the limitations of empirical methods.

A physically rigorous analytical method would nevertheless have the advantage of being able to quickly scan a large number of configurations and would allow for automatic, real-time back-analysis of surface settlement measurements. The proposed method meets this objective by combining, for any stratigraphy:

a calculation of the average convergence and ovality for a tunnel excavated in a homogeneous soil with Young's modulus Eo, Poisson's ratio α, and anisotropy factor K0, and explicitly taking into account the confining pressure,

a rigorous calculation of the displacements induced by this deformation of the excavated surface throughout the entire massif, using a "mirror tunnel" technique,

an extension to non-homogeneous stratigraphies by considering a virtual cover height H*, defined as the sum of the thicknesses Hi weighted by the modulus ratio Ei/E0, Eo being the modulus of the layer in which the tunnel is excavated.

This method results in surface settlements that are very close to the results obtained by finite element calculations, except in the cases of very shallow tunnels, or of soils covered by a layer significantly stiffer than the excavated soil.

Apart from these cases, the deviations, of the order of +/- 10%, are within the required accuracy for settlements, and negligible compared to the precision with which the moduli of soil layers are generally determined.

 --- The Meuse/Haute-Marne Underground Research Laboratory has been operated by Andra since the early 2000s to investigate the feasibility of a deep geological disposal for radioactive waste in the Callovo-Oxfordian claystone (Cigéo project). The laboratory consists of a network of drifts excavated along the directions of the principal horizontal in-situ stresses. Observations have shown that drift excavation induces the development of an asymmetric fractured zone around the openings, leading to anisotropic convergence depending on the drift orientation.
To investigate the interaction between the rock mass and the support system, several experimental drifts have been dedicated to testing different support systems. Monitoring of these drifts includes convergence measurements, as well as stress and strain measurements within the supporting structures. The analysis of in situ measurements, complemented by numerical simulations, highlights the role of compressible wedges installed in the shotcrete support on the mechanical response to both the temporary support and the final lining.
The implementation of various experimental configurations in a numerical model, in which the fractured zone induced by excavation is explicitly represented, makes it possible to assess the long-term response of the various support structures.

For the city, the subsoil represents fundamental challenges for its development and evolution. It is the foundation for surface developments, a place that houses public transportation and technical networks. It is also a source of geothermal energy that is still insufficiently exploited in urban areas. In the context of climate change, the effects of which will be particularly marked over the coming decades in major metropolises, the subsoil can also become a new space to be developed as an alternative to urban sprawl. The S-PASS scientific research project, part of the “Subsurface: a common good” research program funded under the France 2030 plan, is built around these issues, on the geographical perimeter of the Greater Paris metropolis. The project covers the first 200 m of the subsoil, and aims to (1) gain a better understanding of the geological formations, their associated variability, and their geomechanical properties, and to test new geophysical methods in urban areas; (2) build a 3D digital model coupling the 3D geological model of the Cenozoic geological formations, with existing underground public transport infrastructures. The merging of these two models will enable the creation of a digital twin prototype of the Parisian urban underground; (3) to place these underground spaces in the public imagination and in its perception of future developments, to analyse the environmental footprint of underground use in the urban development model in comparison with surface developments ; and finally (4) to consider innovative methodologies for increasing circular economy applications, i.e. reclaiming excavated soil from underground works and using low enthalpy geothermal solutions. This research project, with a budget of 3 million euros and a duration of 7 years, started in 2023. It brings together eleven academic and institutional partners. This article describes the stakes, content and prospects of this project. --- 

This feasibility study investigates the integration of underground heat production and seasonal thermal energy storage into the urban planning framework of Zurich’s inner city. Rather than treating energy infrastructure as a purely technical necessity, the study positions subsurface energy systems as a strategic tool for sustainable urban development, spatial efficiency, and social resilience.
The analysis explores how existing underground spaces can be reused, or new underground facilities developed, to support the expansion of Zurich’s district heating and cooling network while preserving valuable surface space for social, cultural, and community-oriented uses. Particular emphasis is placed on seasonal thermal energy storage (STES), which offers significant decarbonisation potential and enables synergies between heating and cooling demand.
From an urban planning perspective, the study critically assesses the planned conversion of the Selnau substation—currently a vibrant cultural and civic hub—into a surface-based energy plant. It proposes instead the relocation of energy production to centrally located underground sites, enabling continued public use of surface space and reinforcing social sustainability in the surrounding neighbourhoods.

Key planning scenarios include a centralised cavern system beneath Zurich’s University District and the adaptive reuse of existing underground infrastructure such as the decommissioned Letten Tunnel. These scenarios are evaluated in terms of spatial integration, compatibility with current and future land uses, and their contribution to long-term urban resilience and climate goals.
The study concludes that underground energy infrastructure should be proactively integrated into broader urban strategies. Coordinated planning across disciplines—bringing together energy providers, urban planners, engineers, and local communities—is essential to ensure technical feasibility, spatial quality, and social acceptance. Treating the subsurface as a shared urban resource can support the transition towards a climate-resilient, liveable, and space-efficient inner city.

 --- Reuse and Repurpose is a concept of growing importance towards keeping existing urban areas liveable and achieving sustainability. In existing cities there is a growing potential and intent to reuse and repurpose existing facilities; however considerations need to be given to inherent limitations of the existing facilities given their original use and purpose as well as their functional, safety and spatial features during their renewed use. Primary sources of underground spaces that might be candidates for reuse and repurpose are abandoned mines; old military, transportation and logistics facilities; previous parking and storage spaces, and unused civilian shelters. Cities that embarked on reviving these facilities, providing them with a new needed purpose, improved on resiliency of their communities.

Underground structures are generally designed and verified in relation to two categories of limit states: ultimate limit states (ULS) and serviceability limit states (SLS). Verification against SLS is particularly relevant for shallow structures, where excavation may affect neighbouring structures. For ULS, verifications focus on structural (STR) and geotechnical (GEO) states. STR states involve verifying that the lining is not overloaded, while GEO states involve protecting against ground failure.
Some AFTES recommendations describe the approach for justifying concrete lining and the consistency of this approach with the Eurocodes. This method is suitable for structures whose stability relies on rigid support or lining.
In the case of deep tunnels, stability often depends more on the rock mass itself. Assessing the safety margin therefore becomes more complex, requiring a good understanding of the behaviour of the ground and geotechnical uncertainties. There are three main situations:
• The instability mechanism is identifiable, and this allows justification by limit equilibrium.
• The mechanism is uncertain, but the geotechnical model is sufficiently robust for numerical simulation, using a c-ϕ reduction approach and deformation limitation.
• Neither the mechanism nor the parameters are understood justification is then based on an observational approach and real-time monitoring.
The article develops some thoughts on these approaches, emphasising the need to propose a method for assessing the safety margin when stability depends mainly on the contribution of the rock mass. --- 

The Agnettes station (AGN) on section 2 of the Grand Paris Express Line 15 West (L15) is located adjacent to the eponymous RATP Line 13 (L13) station. The project involves a main station box constructed with diaphragm walls, approximately 35 m deep, and an interconnection corridor, constructed with secant piles, providing connectivity with L13. The L15 tunnel runs a few metres below the toe of L13 diaphragm walls, while the closest point of the station endwall is about 13 m away.

The vulnerability study for L13 and other neighbouring structures with respect to the L15 construction was carried out using a global 3D ISS model, incorporating L15 station and tunnel as well as the L13 structure. Each diaphragm wall panel, slab plot and each waterproof joint of the existing L13 cut-and-cover trench and station were explicitly represented. The main structural elements of the L15 station and tunnel were also modelled, simulating all construction phases.

The article will detail the modelling methodology, with particular attention to the joint modelling of the station, tunnel and neighbouring structures within a single model, enabling accurate consideration of the cumulative effects of all planned works on existing structures. The results obtained will be analysed, among others, in terms of deformation and L13 waterproof joint relative displacements.

 --- Various types of primary linings have been constructed and monitored at Andra’s Underground Research Laboratory (URL CMHM), operated by the French National Agency for Radioactive Waste Management (Andra). These structures form part of a scientific and technological demonstration programme, which aims to strengthen the knowledge required for the design and construction of the future facilities of the Cigéo project. Such design is particularly important in view of the operational lifetime and the time-dependent behaviour of the host rock (Callovo-Oxfordian claystone, at a depth of approximately 500 m). This paper focuses on selected elements of the knowledge gained regarding the behaviour of shotcrete primary linings. At the material scale, several characterisation campaigns were carried out, based on specimens taken from dedicated test panels, cores extracted from in situ support systems or multilayer wall panels. In particular, the paper presents shear and permeability tests performed at the interface between two layers of shotcrete applied with a 48 h time interval and under various confinement levels, showing good mechanical bonding at the interface. At the support system scale, selected results related to the construction procedures, instrumentation, adequacy of integrating compressible layers and support loading are discussed. The paper highlights the importance of proper shotcrete application and modelling, as well as its contribution to the structural design of the final structure. --- The Mont-Cenis base tunnel, a key component of the Lyon–Turin railway link, crosses highly tectonised formations, including the Houiller (coal bearing) facies near the Briançonnais Front thrust fault. These zones exhibit significant squeezing behaviour, with diametral convergences up to 2 meters and convergence rates reaching 50 mm/day under overburden exceeding 600 meters. Based on over a decade of monitoring data from the Saint-Martin-la-Porte access tunnel, a robust geomechanical model was developed to calibrate both instantaneous and time-dependent rock mass behaviour. The S2IP design consortium adopted an innovative approach that integrates time-dependent effects at each excavation step, using a viscoplastic Norton law and a cohesion degradation model. This methodology was applied to the R23 cross-passage, located in the Moderate-deformability Houiller facies, and modeled in 3D using FLAC3D. The simulation enabled assessment of excavation phasing, support performance, and long-term lining loads over a 120-year horizon. Results were validated against analytical convergence-confinement methods, confirming stress levels and guiding design adjustments. The proposed approach improves understanding of rock mass/support/lining interaction and supports optimized design in complex geological settings.

Upon completion, Kwu Tung Station (KTU) on the East Rail Line (EAL) will become the MTR Corporation’s 100th station, marking a major milestone in Hong Kong’s rail development. As the first station built directly above an operational tunnel 20 meters underground, KTU is a landmark engineering achievement. Originally planned over thirty years ago, the station is now a key part of the Northern Metropolis, serving the future Kwu Tung North New Development Area (NDA), which will accommodate 130,000 residents.

KTU’s underground location enables efficient and sustainable transit, reducing travel time between Kwu Tung and Sheung Shui from 15 minutes to just 3 minutes when it opens in 2027. The station will also be an interchange and terminus for the future Northern Link (NOL), a 10.7 km underground railway with five new stations, scheduled for completion by 2034.

The project reflects long-term urban planning and is being delivered in three phases: the Kwu Tung tunnel was completed in 2006; the EAL station will open in 2027; and the NOL section is set for 2034. Early works in 2002 included provisions such as an underslab drainage system and reserved openings, facilitating efficient future construction.

KTU stands out as a large-scale project entirely within a brownfield environment, requiring precise engineering to avoid affecting existing rail operations. Unlike conventional methods, the tunnel box uses a raft foundation, subject to 200 kPa underslab buoyancy pressure due to groundwater, demanding constant monitoring and adjustment. This paper describes the solutions developed to overcome these engineering and logistical challenges.

 --- The Underground Crossing of Marseille (TSM) is a strategic component of the phase 2 of the New Provence Côte d’Azur Line (LNPCA), managed by SNCF Réseau. Currently in the concept design stage, the project proposes the construction of 8 km long twin bore tunnel beneath the city of Marseille. This tunnel is connected to the existing railway lines by cut-and-cover access structures, built in direct interface with the existing tracks. The tunnel comprises two tubes, each with an internal diameter of 7.80 meters, built using a pressurized tunnel boring machine. The alignment crosses the Oligocene geological formation in the Marseille area with a maximum cover of 66 meters beneath a densely built and sensitive area. The tunnel serves an underground high-speed train station, connected to the historic train station, the Saint-Charles metro station, and the planned extension of the T2 tramway. This station, measuring 406 meters long by 46 meters wide, can accommodate up to four double high-speed trains simultaneously on two central platforms. This infrastructure provides significant volumes on the upper levels, facilitating the development of underground parking. The station is managed by SNCF Réseau and SNCF Gares et Connexions. Junction tunnels at the ends of the station connect the four platform tracks to the two tunnel tracks. These four junction tunnels, up to 120 meters long, are constructed using traditional methods under low coverage and existing buildings. Their sections vary between 100 and 200 m² to adapt to the loading gauge and allow the tunnel boring machines to pass through to the station. The project includes significant fire safety challenges. --- 

Line 18 of the Grand Paris Express will connect Orly Airport to Versailles Chantiers by 2030. The infrastructure works on lot 3A was awarded to the Spie Batignolles - Ferrovial consortium and began in 2023. This lot is notable for its sparsely urbanized context, with the future 6.7 km line stretching across the municipalities of Guyancourt and Versailles.

This article presents the design approach, the consideration of local constraints, and the progress of the service shafts representative of this lot. Some of the guidelines are based directly on the local context, but also on the feedback from lot 1, while allowing for variations in the construction methods used by the contractor.

Most of the ancillary structures have an emerging building that integrates the technical premises due to the availability of land, allowing the dimensions of the shafts to be optimized. The methods used to construct the shafts and the connection gallery to the main TBM tunnel have been adapted to the geological and hydrogeotechnical context, as well as to the depth (in relation to the tunnel layout), enabling feedback on the various methods. Groundwater management was one of the major challenges and led to specific considerations that will be detailed in the article.

 --- The Léman 2030 project is a major initiative aimed at modernizing and increasing the capacity of the railway network between Lausanne and Geneva, in response to the significant growth in passenger traffic in the Lake Geneva region. A central component of this programme is the expansion of the Geneva-Cornavin station rail hub, which is facing increasing saturation. This article outlines the technical aspects of the geotechnical design and the tunnel of the “Airport Sector”. This sub‑project consists of creating a new 3‑km underground link between Geneva-Cornavin station and the existing surface tracks leading to the airport. To excavate the double‑track tunnel, a slurry pressure‑balance tunnel boring machine (TBM) with a diameter of 12.30 metres—unprecedented in the Geneva region—is planned for use in soft soils that are partially saturated with water. In order to minimize the impact on urban traffic, compensate for the lack of available installation space, and optimize material management, an aerial conveyor system is planned to transport the excavated materials from the TBM logistics area to a remote processing site located 3 km from the urban work zone, and to organize the reuse of part of the excavated material. From there, a dedicated rail platform is envisaged to evacuate the non‑reusable materials by train to disposal sites, thereby drastically reducing the number of trucks on the road and the associated nuisances. In terms of interfaces, the TBM launch and reception zones are located in dense urban areas where several infrastructure projects are underway or planned. Coordinating these projects, combined with works carried out adjacent to an operating railway line, represents a major challenge.

During the construction of tunnels and underground networks, the construction phase poses specific fire safety issues. We have carried out experimental model and numerical studies using the Fire Dynamics Simulator to answer questions raised by the Société des Grands Projets, particularly regarding: the potential heat release of the fire, smoke management, and the possibility of confining the fire zone using water curtain-type devices. The absence of operational ventilation during the construction phase and the complexity of the connections between galleries and structures with access to the outside air influence the oxygenation of the fire and the stratification of smoke. In the event of a fire in a gallery during the excavation phase, the smoke produced spreads to the upper part of the gallery in a stratified profile. A return layer remains in all simulated cases in the lower part downstream of the fire, supplying it with oxygen, even if only partially. This smoke-free layer forms naturally despite the absence of operational ventilation due to the buoyancy forces of the smoke, which is hotter than the incoming air. This air layer, with a thickness of around 40 to 50% of the total height of the gallery, can allow access for rescue and firefighting services in the event of a fire. However, its thickness is reduced in the case of sloping galleries. An external wind above the access shaft is also likely to impact this smoke stratification, or even destroy it completely by disrupting the plume of smoke pouring into the access shaft. Similarly, the use of a water curtain to confine the fire zone is counterproductive: it causes a loss of stratification without blocking the smoke produced or the layer of air returning to the fire zone. The smoke and incoming air are simply mixed and diluted by the spray. --- As part of the work on lines 15 South and 16-17 of the Grand Paris Express, the large scale of the project and the presence of interconnected tunnels, as well as the multitude of stakeholders involved, required a holistic approach coupled with a high degree of adaptability for the temporary ventilation system.
An audit of the temporary ventilation systems has been commissioned by the project owner. Following this, the owner appointed a design consultant to carry out an interdisciplinary study for lines 15 South and 16/17.
The methodology implemented was based on an assessment of the current situation (when it was possible), followed by the definition of a general principle. It has been associated with a verification phase through the sizing calculations and digital simulation of a likely specific case. The proposed methodology has been tested theoretically through anticipated phasing of evolution.
Finally, this methodology was implemented with the involvement of the project owner, including:
▪ the identification of provisions and technical means compatible with the constraints of the work,
▪ an organizational approach that was more or less collaborative depending on the contractual and organisational context of the project. --- Metropolitan Line C of Toulouse is 27 km-lined-major project of construction whose 21 km are underground. The tunnelling phases raise significant safety issues for workers, particularly due to the presence of blind tunnels extending up to 1,800 m length.
Thanks to first meetings between the project owner, construction companies and emergency services, they highlighted the necessity of clarification of the tunneling methods and contraints linked to thes activities in order to adapt the emergency responses. Indeed, these gaps represented a risk for the management of emergency situations in underground environments.
This paper presents feedback from an approach initiated well before the start of construction works, based on close cooperation between Tisséo Ingénierie and the Haute-Garonne Fire and Rescue Department. Thanks to this method, project specific accident scnarios could have been defined and intervention procedures could be adapted. They have also been able to acquire dedicated equipment and implement an extensive awareness and training program.
Beyond information sharing, this process resulted in genuine interdisciplinary acculturation between project stakeholders and emergency services, validated through both theoretical and operational exercises. The lessons learned highlight the importance of integrating such initiatives at an early stage while anticipating their impact on construction organisation and productivity. --- 

The dust generated during tunnel excavation can severely pollute the construction environment inside the tunnel, thereby affecting the health of workers. The dust distribution in the tunnel face area is relatively concentrated and difficult to eliminate. This paper takes the Fangyushan tunnel as an example to study the distribution characteristics of dust in the excavation face area of the tunnel. And establish corresponding physical models to study the influence of different parameters on the dust reduction efficiency of the water mist device. The results show that the dust concentration near the tunnel face is highest during the blasting stage, and the workers' operation area highly overlaps with the high-concentration dust distribution area in space and time. The air humidity difference inside the tunnel has a significant impact on the distribution of dust concentration. When the pressure at the nozzle is constant, the dedusting efficiency generally decreases with the increase in the nozzle diameter. However, when the diameter of the nozzle is less than 0.3 mm, the diffusion area of the water mist is insufficient to cover the tunnel cross-section, resulting in a decrease in dust removal efficiency. When the nozzle diameter is fixed, as the nozzle pressure increases, the dust reduction efficiency shows an increase, and the growth rate changes from a sharp increase to a slow increase, and gradually converges. When the fine water mist is within a certain range from the nozzle, its dust reduction efficiency has little relationship with the distance from the nozzle. While outside this range, the dust reduction efficiency will sharply decrease with increasing distance.

Key words: Tunnel excavation, Mining method, Dust distribution, Water mist system, Model experiment.

Underground construction projects are characterised by structural uncertainty arising from subsurface conditions, technical complexity and significant operational constraints. In this context, disputes are not isolated incidents but are often the result of a gradual accumulation of deferred decisions, contractual ambiguities and/or inadequate governance. Conflict thus emerges as the symptom of a project that has lost its capacity for cooperation.
Too often, the contract is conceived as a defensive instrument, mobilised only at a late stage, once dialogue has already broken down. Yet in projects where uncertainty is inherent, this approach reveals its limitations and tends to fuel claims rather than to prevent them.
By contrast, more mature practices seek to make the contract a genuine project management and steering tool. The development and active use of risk matrices, the implementation of real-time dispute resolution mechanisms (Dispute Boards), project mediation, early warning procedures, and collaborative approaches all help to organise uncertainty, share responsibilities and address tensions before they become entrenched.
In conclusion, the article advocates a view of Contract Management as a fully-fledged governance discipline. In underground projects, contractual maturity does not lie in legal sophistication, but in the ability of stakeholders to identify risks, preserve dialogue and maintain cooperation in the face of the unexpected — an essential condition for performance and long-term trust. --- In the Line 15 West section of the Grand Paris Express, delivered under a design and build scheme, the management of the client’s requirements serves as a governance lever to ensure compliance, infrastructure–systems integration, and the effective control of technical and contractual risks across the project lifecycle. This section (approximately 15 km of tunnels, five underground stations, and seventeen service structures; €2.7 billion, 2023–2031) relies on more than 5,600 requirements spanning functional, technical, site, operations, maintenance, and FMD (RAMS) dimensions, thereby requiring disciplined traceability and continuous monitoring. The contractual framework specifies expectations for traceability and compliance demonstration, while leaving the consortium the latitude to define the operational arrangement. In the absence of established standards and off-the-shelf solutions comparable to those in industries (e.g., aerospace and nuclear) for managing cross-disciplinary requirements in an integrated infrastructure and systems context, the consortium implemented a phase-adaptive requirements management arrangement grounded in clarified responsibilities and validation criteria, structured management of deviations and changes, and a progressive coverage strategy aligned with design and approval milestones. The approach organises evidence production by distinguishing “structuring” requirements that drive architectural choices and system integration from those that can be demonstrated at later stages, thereby limiting rework and securing key trade-offs. Supported by a dedicated coordination team involving more than 300 contributors, this initiative has structured a requirements-oriented governance model, transforming the requirements baseline into a transversal steering instrument serving compliance, integration, and risk control within an integrated design and build schedule. --- The law governing public project ownership and its relationship with private Design and Engineer in charge services (the “MOP law”), now incorporated into the Public Procurement Code, provides for a lump-sum remuneration for each Design and Engineer in charge assignment, applied to given Owner’s requirements. This framework is not suited to underground works, which involve a high level of uncertainty and unfold over long periods of time. The GT17R2F1 recommendation identifies and proposes contractual principles and mechanisms to adapt Design and Engineer in charge contracts to the specific characteristics of underground projects, while complying with the Public Procurement Code (including the MOP law) and taking into account the CCAG MOE. This article first reviews the objectives and key issues of the MOP law, the origins of the lump-sum remuneration for Design and Engineer in charge services, and the findings of the underground works profession regarding this lump sum, as observed within the AFTES Observatory of contractual practices. In its second part, the article presents the GT17R2F1 recommendation, highlighting the key messages of each of its chapters.