CFD Applications of PHOENICS on Building Environment and Fire Safety Design 1 Qian Wang, PhD, CFD Specialist Kenneth Ma, Senior Associate Micael Lundqvist, Senior Fire Engineer Ove Arup Pty Ltd Level 10, 201 Kent Street, Sydney NSW 2000, Australia

2Software PHEONICS PC version 3.3 & 3.4 Computers COMPAQ WorkStation Pentium 3/800MHz & Pentium 4/1.6GHz & Pentium 4/1.6GHz INTRRODUCTION Abstract ARUP has been using PHOENICS for many years to deal with various CFD modellings in building thermal comfort design, indoor environment, fire safety control, etc. and has gained good results recognised by the clients. This paper summarises some selected CFD studies of normal & emergency ventilation controls.

3 Description of Project Project Type Urban underground railway station with connecting tunnel to the ground level, heavily occupied with diesel trains CFD Objectives 1.Platform: thermal comfort & air quality. 2.Tunnel:smoke ventilation control during emergency fire in the connecting tunnel.

Diagram Description of Project 4

5 CFD Tasks & Outcomes To provide all detailed information to support the final design of mechanical ventilation system.To provide all detailed information to support the final design of mechanical ventilation system. Temperature, air velocity, concentration of pollutant gases (CO, CO 2, NO, etc) in winter & summer seasons.Temperature, air velocity, concentration of pollutant gases (CO, CO 2, NO, etc) in winter & summer seasons. Station Normal Ventilation

CFD Domain 6 Station Normal Ventilation

7 CFD Input Conditions Station Normal Ventilation

Station Normal Ventilation - winter 8 Velocity Vector

Station Normal Ventilation - winter 9

Station Normal Ventilation - summer Velocity Vectors 10

Station Normal Ventilation - summer Velocity Vectors 11

Station Normal Ventilation 12Summary Winter Allowable concentration of CO (25ppm) is acceptable.Allowable concentration of CO (25ppm) is acceptable. Gas fume is accumulating near ceiling.Gas fume is accumulating near ceiling.Summer Well mixed fluid domain.Well mixed fluid domain. Containment materials may be driven towards the platforms.Containment materials may be driven towards the platforms. Hot layer within T > 30°C is broader and thicker, may result in discomfort to the passengers.Hot layer within T > 30°C is broader and thicker, may result in discomfort to the passengers. The 0.082% CO level (25ppm) is quite close to the platforms – greater ventilation capacity is required.The 0.082% CO level (25ppm) is quite close to the platforms – greater ventilation capacity is required.

13 CFD Tasks & Outcomes To evaluate the smoke control policy during emergency fires in the sloped tunnel (450m x 8.9m x 6m).To evaluate the smoke control policy during emergency fires in the sloped tunnel (450m x 8.9m x 6m). Transient air velocities, temperatures and smoke concentrations.Transient air velocities, temperatures and smoke concentrations. Behaviours of backlayering of smoke towards station platform.Behaviours of backlayering of smoke towards station platform. Tunnel Fire Smoke Control

14

Ventilation Shaft Fire Source Entrance of Ground Level Connection to Station Jet Fan Train Train Fire in Entry Tunnel 8.9m 6m Tunnel Fire Smoke Control 15

16 Tunnel Fire Smoke Control T < 200°C if hot layer is above 1.5m from floor level.T < 200°C if hot layer is above 1.5m from floor level. T <60°C if hot layer is below 1.5m and/or the visibility not be less than 6m (ie the optical density should not exceed 0.14m -1 ).T <60°C if hot layer is below 1.5m and/or the visibility not be less than 6m (ie the optical density should not exceed 0.14m -1 ). Fire Safety Design Criteria Fire Scenario Description 1 - Carriage Fire 'worst credible' fire breaking out inside a train that is stopping in the tunnel between the two vent shafts. Assumed to be an exponentially fast growing fire with the maximum fire size 15 MW, as a fully developed carriage fire. 2 - Suppressed Diesel Fire 'worst credible scenario for a fire outside a stopped train between the two vent shafts and is leaking diesel. The fire is assumed to be a fast (0.047kW/s²) growing fire, which is suppressed upon activation of the foam suppression system at track level. 3 - Unsuppres sed Diesel Fire This fire scenario is a sensitivity analysis of the diesel fire outside the carriage (Scenario 2) in the event of failure of the foam suppression system. The fire therefore continuos to grow to its maximum size 40 MW, involving both the diesel and a carriage.

t [min] Q [MW] a b d e Scenario 1-- abc Scenario 2-- abe Scenario 3-- abcd Fire starts All jet fans stop F1 operates: 180m 3 /s c 17 Tunnel Fire Smoke Control

18 Tunnel Fire Smoke Control - Scenario 1 TemperatureConcentration

19 Tunnel Fire Smoke Control - Scenario 3 Temperature Concentration Backlayering Distance

20 Double click the image to play ! Tunnel Fire Smoke Control

21 Tunnel Fire Smoke Control Summary It is proved that the design smoke control/ventilation system during different fires will be able to provide a reasonable fire safety condition according to the calculated internal temperature, CO concentration levels. Conclusions of PHOENICS PHOENICS applications on building internal air quality control and emergency fire smoke control strategy have been carried out. Very detailed thermal and fluid behaviours of internal air have been analysed, which either identified the efficiency of the ventilation systems or provided the optimisation to the design features. All these results prove that PHOENICS can deal with very broad fluid dynamic modelling, and is the most cost effective tool in professional engineering consultant services.