tunnels with FDS and 1D human behavior FCC Study Case Oriol RIOS, - - PowerPoint PPT Presentation

▶
tunnels with fds and 1d human behavior
SMART_READER_LITE
LIVE PREVIEW

tunnels with FDS and 1D human behavior FCC Study Case Oriol RIOS, - - PowerPoint PPT Presentation

Sep 05, 2022 210 likes •657 views

Estimating probability of fatalities in long tunnels with FDS and 1D human behavior FCC Study Case Oriol RIOS, Art ARNALICH, Saverio LA MENDOLA HSE-OHS-XP 02-October - 2018 FEMTC-2018. Gaithersburg, MD, USA 2 O. Rios, A. Arnalich, S. La

slide-1
SLIDE 1
slide-2
SLIDE 2

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Estimating probability of fatalities in long tunnels with FDS and 1D human behavior

FCC Study Case

Oriol RIOS, Art ARNALICH, Saverio LA MENDOLA HSE-OHS-XP

2

02-October - 2018

slide-3
SLIDE 3

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

3

What’s CERN?

~5000 members of personnel ~5000 associate members Since 1954 22 member states ~600 universities Two official languages

~100 km underground tunnels* with one-of-a-kind equipment

*and technical galleries

slide-4
SLIDE 4

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Underground tunnels at CERN

4

5 km 27km length

slide-5
SLIDE 5

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

5

Future Circular Collider. FCC project

In numbers ~ 98km ~ 300-500 m underground ~ 6-10km between shafts

slide-6
SLIDE 6

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

6

PBD process:

slide-7
SLIDE 7

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Safety Objectives

7 Life Environment Property Continuity of

  • peration

A B C D 1

Occupants shall be able to evacuate through protected areas, free from smoke/gas and other hazards at any time Limit the release of polluting (incl. activated) agents to the environment in case of incident The continuity of essential services and structural stability is assured in case of fire or gas release and other incidents Limiting the downtime in case of incident

2

Victims and other occupants, not able to self-evacuate, shall reach protected areas, and wait there to be rescued by the intervention teams Limit the volume of polluted (incl. activated) water released to the environment in case of incidents An incident shall not cause

  • ther potentially dangerous

accidental events

  • 3

Rescue teams shall be able to intervene safely and according to current CERN SOPs

  • Limiting the property loss in

case of incident

  • Fire induced radiological hazard (to environment, to

evacuees or interveners) is not in this study scope. Research in progress: FIRIA project.

slide-8
SLIDE 8

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

8

slide-9
SLIDE 9

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

9

Acceptance Criteria (for Life Safety Objectives)

  • Occupants

Occupants shall be able to evacuate through protected areas, free from smoke/gas and other hazards at any time

  • Victims

Victims and other occupants, not able to self-evacuate, shall reach protected areas, and wait there to be rescued by the intervention teams

  • Firefighters

Rescue teams shall be able to intervene safely and according to current CERN SOPs

02 Any victim outside the fire compartment of the seat

  • f the fire has reasonable opportunity of not reaching a

FED>0.1 before being rescued by intervention teams 03 Within the fire compartment of the seat of the fire, any of the two following criteria is met:

03a Fire compartment size and layout is such that there is a reasonable opportunity of not having a victim in the interior 03b Any victim has a reasonable opportunity of not reaching a FED>0.3 before being rescued by intervention teams

slide-10
SLIDE 10

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

10

PBD process:

slide-11
SLIDE 11

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Scenarios and Fire Designs

11

  • Those 3 scenarios are considered to be

more representatives ones

  • Only fire scenarios are considered.
  • Fire Scenarios resulting from explosions

are not accounted for.

CABLE TRAY DRUM KUKA

slide-12
SLIDE 12

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Fire#1 - Tray Fire

3.1 Fire designs: Enveloping cases

12

1000 2000 3000 4000 5000 6000 7000 8000 9000 10 20 30 40 50 60 70 80 90 100

HRR [kW] Time [min]

1000 2000 3000 4000 5000 6000 7000 8000 9000 10 20 30 40 50 60 70 80 90 100

HRR [kW] Time [min]

1000 2000 3000 4000 5000 6000 7000 8000 9000 10 20 30 40 50 60 70 80 90 100

HRR [kW] Time [min]

Fire#2 – Drum Fire Fire#3 - Kuka Fire

In all cases, the fire is assumed to be located at 1/3 of the length of the compartment

Fire#2 and Fire#3 have same growth up to 8min! If compartmentation works, with current ventilation system HRRmax < 5MW (ventilation limited)

slide-13
SLIDE 13

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

13

Geometry and safety measures

  • Fire compartment = 440m
  • Smoke extraction
  • Air supply

Baseline Safety Features

slide-14
SLIDE 14

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Ventilation: Proposed Strategy

14

Normal Ventilation Condition Normal Ventilation Condition

Uo=~ 0.5

slide-15
SLIDE 15

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Ventilation: Proposed Strategy

  • 3 compartment doors closes (fire compartment and neighbours)
  • Air inlets remain as normal condition.
  • All arc outlets close, except for:
  • 4 in the fire compartment
  • 2 in adjacent compartments
  • Exhaust duct extracts total flow of 14000m3/h (~1/3 of nominal fans power)
  • Overpressure created in neighbouring compartments
  • Normal longitudinal ventilation in all other compartments (or not, not influence)
  • Fast reaction. Only action dampers in the fire compartment. Duct under pressured.

15

When Fire is detected in a compartment

> 7000 m3/h ~7000 m3/h + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + > 3500 m3/h > 3500 m3/h ~3500 3/h 2080m3/h ~3500 m3/h 2080m3/h

Smoke Exhaust: 14k m3/h

slide-16
SLIDE 16

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

16

OUTPUTS

Slide Files

  • Temperature
  • Velocity

Vertical Trees

  • Temperature
  • O2
  • CO
  • Visibility
  • FED

Center of the walking lane 2m from floor 44 trees, 9 vertical positions Δz = 0.4m Δx = 10m

FDS set up 1 Compartment (440m) modelled

~ 𝟑 − 𝟐𝟕 ∙ 𝟐𝟏𝟕 cells HPC to get reasonable computational times

slide-17
SLIDE 17

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Visibility Model & Human Behavior

Human Speed

Reduced velocity due to visibility

Yields of toxic and asphyxiants

𝑤𝑑𝑝2 = 1, 𝑌𝑑𝑝2 < 2% 𝑓𝑑𝑝2/5, 𝑌𝑑𝑝2 < 2% Unimpeded Human speed, 𝑤𝑝 = 1 m/s Hyperventilation

As: ISO/TS 13571(International Standards, 2012, Purser, 2008)

  • S. Arias, E, Ronchi, Lund University

𝑤′ = 0.34𝑊 + 0.31 𝑊

𝑑𝑝𝑠 = min(𝑤′, 𝑤𝑝)

Soot Yield = 0.1

Walking speed as a function of visibility by Fridolf et al (2016).

slide-18
SLIDE 18

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

18 t t

d

t te

e

v

Detection, alarm on, ventilation off End of evacuation

tin

t

Start of Fire fighter intervention Ongoing evacuation

x v

Fire detection deactivated (IS37) Active fire detection Hot works

ti= 0

Ignition

slide-19
SLIDE 19

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Berlin Baseline. Det. 510s. No supply. Emergency extraction 300s to ramp up. FIRE#2 and FIRE#3

19

Detection Door Closes Exhaust Starts

Smoke propagates to next compartment downstream!!!

001_FCC_20cm_23msh_200ptv_510det_300t2ev_1sh_devc_noLeak

slide-20
SLIDE 20

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Visibility and Temperature: Fire#3

20

Life Safety objective met !

slide-21
SLIDE 21

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Victims: FED calculations

21

FED calculations

𝐺𝐹𝐸𝑢𝑝𝑢𝑏𝑚 = 𝐺𝐹𝐸𝑑𝑝 + 𝐺𝐹𝐸𝐷𝑂 + 𝐺𝐹𝐸𝑂𝑃𝑦 × 𝐼𝑊

𝐷𝑃2

+𝐺𝐹𝐸𝑃2

CO intoxication O2 ODH CO2 hyperventilation

Further knowledge of combustion material not available

Visibility at Z=0.4m 1 2

03 Within the fire compartment of the seat of the fire, any of the two following criteria is met:

03a Fire compartment size and layout is such that there is a reasonable opportunity of not having a victim in the interior 03b Any victim has a reasonable opportunity of not reaching a FED>0.3 before being rescued by intervention teams

slide-22
SLIDE 22

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

𝑮𝑭𝑬 𝒖 = 𝒃𝒇𝒄(𝒖−𝒖𝒑)

22 Different to

𝒖𝒑 → 𝑮𝑭𝑬 𝒖𝟏 > 𝟐𝒇 − 𝟐𝟕

Extremely good fitting!

(data coming from CFD model...)

Extrapolation required

slide-23
SLIDE 23

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

23 Fire#1 Fire#2 Fire#3

slide-24
SLIDE 24

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Victims rescue by fire fighter

24

. . .

n victims...

slide-25
SLIDE 25

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Fire Fighting Victims rescue

25

SPS Fire Safety Study Group / Firefighting Tactical Approach , July 2015

Operation Speed

Blind progression 0.8m/s Carrying victim 0.26m/s

Rescue mask considered!

slide-26
SLIDE 26

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Fire Fighting Victims rescue

26 T Show Up [min] Case 15 FF ready at the bottom of each shaft (LTS, ST) 30 FF on each shaft (surface) 60 FF every two shaft (surface)

slide-27
SLIDE 27

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

On FED distribution

27

𝐌𝐩𝐡𝐎𝐩𝐬𝐧𝐛𝐦(𝝂 = 𝟏, 𝝉 = 𝟐)

FED P dying* 0.1 1% 0.3 11% 1 50% 2 76% 3 86% 4 92%

Meaning

*or being incapacitated

slide-28
SLIDE 28

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

FED to every victim and dying probability

TshowUp = 30min Victims Location = 50m from door

FED FED → Dying prob. Expected #fatalities| #vict.

𝐌𝐩𝐡𝐎𝐩𝐬𝐧𝐛𝐦(𝝂 = 𝟏, 𝝉 = 𝟐)

slide-29
SLIDE 29

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

FED

Fire#1 Fire#2 Fire#3

FED FED → Dying prob. Expected #fatalities| #vict.

slide-30
SLIDE 30

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

FED

30 Fire#1 Fire#2 Fire#3

FED FED → Dying prob. Expected #fatalities| #vict.

slide-31
SLIDE 31

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Conclusions on Victims assessment

  • Vict. Loc.

FIRE #1 FIRE #2 FIRE #3

Show Up t

30m 280m 30m 280m 30m 280m 15 min ALL 2 1 30 min 6 1 60 min

31

*A victim with a dying probability > 0.8 → fatality

Victims (impeded occupants) that can be potentially saved*.

10 victims (impeded occupants) considered

slide-32
SLIDE 32

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Conclusions

  • PBD approach was a sound approach to help designing a safe

infrastructure and proof the safety systems performance

  • FDS was necessary to solve the ventilation scenarios and corresponding

flows

  • Some uncertainties and discrepancies behind FED calculations. If used

consistently, it could be an indicator for probabilistic quantification of consequences

  • Quantitative risk analysis might be necessary to stablish the cost-benefit

solutions and check the results against acceptable risk profile limits

32

slide-33
SLIDE 33

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Further work

  • Use FED as a robust indicator for Quantitative Risk assessments.
  • Exploration of FED addition throughout victims and probabilistic

meaningful

  • Better characterization of yields of our combustibles

33

slide-34
SLIDE 34
slide-35
SLIDE 35

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

FED

Fire#1 Fire#2 Fire#3

slide-36
SLIDE 36

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

FED

Fire#1 Fire#2 Fire#3

slide-37
SLIDE 37

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

FED

Fire#1 Fire#2 Fire#3

slide-38
SLIDE 38

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

FED

Fire#1 Fire#2 Fire#3

slide-39
SLIDE 39

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

FED

Fire#1 Fire#2 Fire#3

slide-40
SLIDE 40

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

Scenarios

41 A probabilistic approach might be considered for the TDR

slide-41
SLIDE 41

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

FED

Fire#1 Fire#2 Fire#3

slide-42
SLIDE 42

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

43

PBD process: EVALUATE TRIAL DESIGN Victim’s FED assessment

slide-43
SLIDE 43

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

45

Abstract: Life Safety is frequently the primary goal in a Performance Based Design (PBD) process. Tenability limits based on visibility, temperature and heat flux thresholds are often used to establish the performance criteria. However, exposure to these limits does not imply fatalities but rather impeded evacuees requiring Fire Service intervention in order to be rescued. In the scope of a preliminary design of a large new particle accelerator, featuring a 98 km tunnel ring laying 400 m underground, CERN FSE team conducted a full PBD, where the impact of the fire brigade intervention on the survivability of potential victims was studied in

  • depth. Given a simple layout scenario from an evacuation standpoint, 1D egress models were

coupled with FDS simulations to quantify the Fractional Effective Dose (FED) received by both evacuating occupants and victims unable to self-evacuate. The Fire Service search&rescue

  • peration was also simulated by using experimental data on zero-visibility intervention access
  • speeds. Multiple tentative configurations of the in-house fire brigade (in terms of locations

and staffing) were tested to evaluate the potential benefit of each one. The results show the advantage and capability of the PBD approach to deal with non-standard complex infrastructures out of the scope of the regulatory framework.

slide-44
SLIDE 44

FEMTC-2018. Gaithersburg, MD, USA

  • O. Rios, A. Arnalich, S. La Mendola

46

When Fire is detected in a compartment

> 7000 m3/h ~7000 m3/h + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + > 3500 m3/h > 3500 m3/h ~3500 3/h 2080m3/h ~3500 m3/h 2080m3/h

Smoke Exhaust: 14k m3/h

If Fire is detected in a adjacent compartment too

> 7000 m3/h ~7000 m3/h + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + > 3500 m3/h > 3500 m3/h ~3500 3/h 2080m3/h ~7000 m3/h ~3500 m3/h > 7000 m3/h

Smoke Exhaust: 21k m3/h

If fire spreads: Replicate logics

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +