1. TS/CV/DC CFD Team CFD Simulation of a Fire in CERN Globe of Science and Innovation Sara C. Eicher CERN, CH

2. THE PROBLEM Sara C. Eicher The Globe is a spherical shape building made mainly of wood, with a crown vortex placed at its top. It consists of two floors and a basement room.

3. THE PROBLEM Sara C. Eicher The Globe is a spherical shape building made mainly of wood, with a crown vortex placed at its top. It consists of two floors and a basement room. The building is fitted with fire sprinkler, ventilation and smoke extraction systems.

4. THE PROBLEM Sara C. Eicher The Globe is a spherical shape building made mainly of wood, with a crown vortex placed at its top. It consists of two floors and a basement room. The building is fitted with fire sprinkler, ventilation and smoke extraction systems. Current emergency strategies to control smoke spread and evacuation are incomplete due to need to comply with established fire safety regulations.

5. THE PROBLEM Sara C. Eicher The Globe is a spherical shape building made mainly of wood, with a crown vortex placed at its top. It consists of two floors and a basement room. The building is fitted with fire sprinkler, ventilation and smoke extraction systems. Current emergency strategies to control smoke spread and evacuation are incomplete due to need to comply with established fire safety regulations. In the event of a fire in the first floor, the smoke extractor in place is believed to be inadequate for removal of the hot smoke accumulated in the ceiling of the first floor hall.

6. OBJECTIVES Sara C. Eicher To evaluate the performance of current emergency smoke extraction strategies of the Globe in the event of a fire.

7. OBJECTIVES Sara C. Eicher To evaluate the performance of current emergency smoke extraction strategies of the Globe in the event of a fire. To evaluate the effectiveness of installing hatches at the central vortex to provide natural smoke evacuation by comparison with the behaviour of the smoke system currently in place.

8. OBJECTIVES Sara C. Eicher To evaluate the performance of current emergency smoke extraction strategies of the Globe in the event of a fire. To evaluate the effectiveness of placing hatches at the central vortex to provide natural smoke evacuation by comparison with the behaviour of the smoke system currently in place. To optimise the improved smoke evacuation system by investigating the effect of location, size and overall quantity of hatches to be placed within the constraints of the vortex geometry.

9. THE CFD MODEL Sara C. Eicher Only the first floor hall is modelled.

10. THE CFD MODEL Sara C. Eicher Only the first floor hall is modelled. The fire is represented by a non-spreading fluid region and modelled as quadratically increasing source of heat placed in the middle of the first floor. The combustion process was not simulated. The fire maximum heat output was set equal to 250kW/m 3

11. THE CFD MODEL Sara C. Eicher Only the first floor hall is modelled. The fire is represented by a non-spreading fluid region and modelled as quadratically increasing source of heat placed in the middle of the first floor. The combustion process was not simulated. The fire maximum heat output was set equal to 250kW/m 3. Overall geometry takes into account the lift cage, inner ramp and wall partitions, 2 exit doors and the smoke extractor.

12. THE CFD MODEL Sara C. Eicher Only the first floor hall is modelled. The fire is represented by a non-spreading fluid region and modelled as quadratically increasing source of heat placed in the middle of the first floor. The combustion process was not simulated. The fire maximum heat output was set equal to 250kW/m 3. Overall geometry takes into account the lift cage, inner ramp and wall partitions, 2 exit doors and the smoke extractor. Fire sprinkler and ventilation systems are not considered. Smoke extractor assumed as a mass sink. Adiabatic wall conditions.

13. THE CFD MODEL Sara C. Eicher Only the first floor hall is modelled. The fire is represented by a non-spreading fluid region and modelled as quadratically increasing source of heat placed in the middle of the first floor. The combustion process was not simulated. The fire maximum heat output was set equal to 250kW/m 3. Overall geometry takes into account the lift cage, inner ramp and wall partitions, 2 exit doors and the smoke extractor. Fire sprinkler and ventilation systems are not considered. Smoke extractor assumed as a mass sink. Adiabatic wall conditions. Calculation strategy: start from an isothermal steady state solution and use results as initial guess to transient calculation accounting for the development of the fire.

14. GEOMETRY & MESH Sara C. Eicher Mesh containing around 400,000 tetrahedral shape cells Two cases studied: Case A – without hatches (current design configuration) Case B – with hatches at the crown vortex Transient model simulates 14 minutes of fire development Lift Cage Wall Partitions Crown Vortex Side Ramp Exit Door Wall Partitions Fire

15. RESULTS – TEMPERATURE FIELD Sara C. Eicher Case A Case B

16. RESULTS – VELOCITY FIELD Sara C. Eicher Case A Case B

17. CONCLUSIONS CFD model was able to predict the overall behaviour of the smoke extraction system currently in place, in case of a fire sprinkler failure. Sara C. Eicher

18. CONCLUSIONS CFD model was able to predict the overall behaviour of the smoke extraction system currently in place, in case of a fire sprinkler failure. Numerical results show expected temperatures stratification at the dome. Degree of stratification depends on the airflow pattern inside the hall, with temperatures at the ceiling up to 60˚C for case A. Sara C. Eicher

19. CONCLUSIONS CFD model was able to predict the overall behaviour of the smoke extraction system currently in place, in case of a fire sprinkler failure. Numerical results show expected temperatures stratification at the dome. Degree of stratification depends on the airflow pattern inside the hall, with temperatures at the ceiling up to 60˚C for case A. Installing of hatches at the vortex crown can improve the clearing of hot gases close to the ceiling, with temperatures about 20˚C lower than in the original configuration. Sara C. Eicher

20. CONCLUSIONS CFD model was able to predict the overall behaviour of the smoke extraction system currently in place, in case of a fire sprinkler failure. Numerical results show expected temperatures stratification at the dome. Degree of stratification depends on the airflow pattern inside the hall, with temperatures at the ceiling up to 60˚C for case A. Placing of hatches at the vortex crown can improve the clearing of hot gases close to the ceiling, with temperatures about 20˚C lower than in the original configuration. Both scenarios suggest that after 14 minutes of fire development, conditions at floor level remain acceptable for safe evacuation of people from the building. Sara C. Eicher sara.correia.eicher@cern.ch

21. CONCLUSIONS CFD model was able to predict the overall behaviour of the smoke extraction system currently in place, in case of a fire sprinkler failure. Numerical results show expected temperatures stratification at the dome. Degree of stratification depends on the airflow pattern inside the hall, with temperatures at the ceiling up to 60˚C for case A. Placing of hatches at the vortex crown can improve the clearing of hot gases close to the ceiling, with temperatures about 20˚C lower than in the original configuration. Both scenarios suggest that after 14 minutes of fire development, conditions at floor level remain acceptable for safe evacuation of people from the building. Sara C. Eicher sara.correia.eicher@cern.ch