Analysis of the ventilation systems in the Dartford tunnels using a multi-scale modelling approach

F. Colella, G. Rein, R. Carvel, P. Reszka, J. L. Torero

Research output: Contribution to journalArticlepeer-review

43 Scopus citations


The capabilities of the ventilation systems in the two road tunnels at Dartford (UK) are analysed using a multi-scale modelling approach. Both tunnels have complex semi-transverse ventilation systems with jet fans to control longitudinal flow. The construction and ventilation systems in the tunnels are described and the current emergency ventilation strategies are presented. The analysis includes a coupling of a 1D network model with 3D components, representing the operational jet fans, built using computational fluid dynamics. The jet fans were experimentally characterized on-site and the findings were compared to the model predictions. The predicted ventilation flows for each of the emergency ventilation strategies are presented and discussed. In cold-flow conditions, ventilation velocities significantly above 3. m/s can be generated throughout the tunnels. However, it is observed that 1/3 of the flow generated in the East tunnel is diverted from the tunnel up the extract shafts. The model was used to simulate various reduced fan combinations and thus the level of redundancy in each of the systems has been estimated. It is found that an acceptable level of ventilation may be produced in the West tunnel, even if several pairs of jet fans are disabled. In the East tunnel there is less redundancy, but an acceptable level of ventilation control can be maintained with one or two jet fans disabled.

Original languageEnglish
Pages (from-to)423-432
Number of pages10
JournalTunnelling and Underground Space Technology
Issue number4
StatePublished - Jul 2010
Externally publishedYes


  • Dartford tunnels
  • Jet fans
  • Multi-scale modelling
  • Network model
  • Redundancy
  • Ventilation system


Dive into the research topics of 'Analysis of the ventilation systems in the Dartford tunnels using a multi-scale modelling approach'. Together they form a unique fingerprint.

Cite this