1. M. Tabarra, R.D. Matthews, B. Kenrick South Bank University, London
The Revival of Saccardo Ejectors: History, Fundamentals and Applications
M. Tabarra, R.D. Matthews, B. Kenrick
CenTAR
South Bank University, London

2. The Centre for Tunnel Aerodynamics Research (CenTAR)
First established in 1991 to study the aerodynamic characteristics of vehicle tunnels longitudinally ventilated by jet fans
Links with a number of industrial partners in the ventilation field developed over many years of research & consulting
A wide range of computational, experimental and full-scale test facilities

3. CTRL Project Genesis
The UK project to build a high speed rail link from the Channel Tunnel to London St. Pancras is called the Channel Tunnel Rail Link (CTRL)
The project began in 1997 to study the feasibility of using Saccardo ejectors in a high resistance rail tunnel under the river Thames
The alternative shaft ventilation system would cost 10 times more and take much longer in terms of completion time

4. History of Saccardo Ejectors & Literature Survey

5. Schematic of a Saccardo ejector operating in short tunnels

6. Schematic of a Saccardo ejector operating in long or highly resistive tunnels

7. Saccardo Ejector Characteristics
Brings in fresh ambient air and imparts thrust to the tunnel air at one point
Avoids the cost and dangers of electrical cablings
Can operate in short or long tunnels, entraining/rejecting air at the near portal
Can form a compact ventilation system near the portals of “small bore” road and rail tunnels, where head room is restricted

8. The Project Aims:
To provide a critical velocity of 2.5 m/s in the train annulus to prevent smoke back-layering (later reduced to 1.5 m/s for the max fire load of 7 MW)
To maintain an acceptable air velocity on the walkway not exceeding 11 m/s
To maintain a pressurised safe haven in the non-incident tunnel
To achieve the above with an ejector jet velocity not exceeding 30 m/s

9. Project Specification
The tunnel: m diameter twin-bored tunnel, 2.5 km long, plus 300 m long rectangular cut & cover sections at each end
Worst case scenario: 2 trains stationary in the tunnel; the leading train on fire & the other one stranded some distance behind, adding significantly to tunnel resistance

10. A 3-way Approach to the problem
A network analysis model of Saccardo ejectors operating against varying tunnel resistance
A 1/26th scale experimental rig of the stationary TGV train underneath the ejector at the interface of the tunnel and the cut & cover
A CFD model of the experimental rig as a predictive design tool in order to visualise the likely flow field anywhere

11. The Experimental Rig
1/26th scale of the southern cut & cover section interfacing with the bored tunnel + TGV train placed in several critical locations underneath the ejector
The effective resistance of the tunnel length with stationary trains was simulated by an iris damper to give the necessary pressure loss
Two ejector nozzle geometries (flush and cavity) were built and tested for comparison

12. Experimental Rig

13. Bird’s eye view of the tunnel interface region at CenTAR laboratories

14. Front view of the TGV with the wool tufts on the walkway providing visual indication of flow direction

15. 1/26th scale TGV train at the tunnel interface

16. The cut & cover region between the tunnel interface & the ejector

17. The Saccardo ejector: drawing air from ambient and delivering it as a 30 m/s at 10 ° inclination

20. The variable area iris damper allows for various tunnel lengths and resistance

21. Side elevation view of the flush nozzle ejector

22. Side elevation view of the cavity nozzle ejector

23. CFD Simulation CFX4 code from AEA, UK
Steady state simulation with pressure boundaries to emulate the experimental conditions
Both the pressure rise and recirculating flows in the jet diffusion area beneath the nozzle are modelled very well

24. Experimental results

25. Experimental results

26. Experimental vs CFD results of streamwise pressure variations
Saccardo
Nozzle

27. You say: “But I knew all that!”
Even though Saccardo ejectors have been around for a hundred years now, their positive features have not been adequately appreciated.
Especially in the flow rejection mode, they have often been avoided, perhaps for their poor hydraulic power efficiency.
But we plan to use them exactly in that mode, rejecting the flow at the near portal ...

28. Schematic Diagram of the CTRL Thames Tunnel

29. Creating a Safe Haven
In twin-bored tunnel designs it is possible to pressurise the non-incident tunnel by using two opposing Saccardo ejectors (one at each portal) blowing towards the centre and use this non-incident tunnel as a safe passage for escape
Typical Nozzle size: 4.2 m2 & Fan Flow rate = 130 m3/s
Environmental conditions in the jet diffusion zone have been studied and found to be acceptable
The tunnel flows and cross passage flows from safe haven to incident tunnel have now been investigated for a range of emergency scenarios (130 cases)

30. Conclusions
Saccardo ejectors have remained obscure and partly misunderstood for a variety of reasons
The Saccardo ejector provides an ideal, robust & economic design for emergency ventilation of tunnels
The presence of the train under the ejector had a favourable influence on the performance because of smaller effective cut & cover c.s. area
The novel design of two opposed Saccardos supports the evacuation philosophy for the Thames Tunnel by maintaining a pressurised safe zone …
as well as enabling smoke control in the incident tunnel in the desired direction