[PPT] - MECHANISTIC SAFETY ANALYSIS OF HYDROGEN BASED ENERGY SYSTEMS W. PowerPoint Presentation

SLIDE 1

1

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

MECHANISTIC SAFETY ANALYSIS OF HYDROGEN BASED ENERGY SYSTEMS

W. Breitung

Institute for Nuclear and Energy Technologies Karlsruhe Research Center Germany

Second European Summer School on Hydrogen Safety, University of Ulster, Belfast, 30 July- 8 August 2007

SLIDE 2

2

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

CONTENT OF PRESENTATION

Presentation consists of two topics which are treated in parallel

Analysis of hydrogen release in a private garage Analysis tools and related physics Combustion regimes Gas transport and mixing Turbulent deflagration Detonation Structural response Mixture generation Sequence of analysis steps Combustion Hazard potential Consequences

SLIDE 3

3

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

OUR EXAMPLE FOR HYDROGEN ANALYSIS

Oil peak behind us, hydrogen fueled cars in

widespread use

Returned from a trip late at night
There was some small collision but apperently

no domage to LH2-system

Park car in private garage
But at night the questions come ….

What would happen in case of a hydrogen leak? What mixtures could develop? Could they be flammable? How fast could the burn be? What would be the pressure loads? What could be the consequences?

SLIDE 4

4

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

GENERIC ARCHITECTURE OF AN LH2-TANK SYSTEM

Source: EU-Project EIHP-2, Final Report 2004

CRACK!

SLIDE 5

5

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

INVESTIGATED GARAGE SCENARIOS

2 22.3 34 100 0.34 1 under-neath trunk 22.3 34 10 3.40 Two times 10 x 20 cm² 70.2 Nr. Release Location Release Temp. (K) Total Mass (g) Duration (s) H2-Rate (g/s) Vent Openings Garage Volume (m³) CASE HYDROGEN SOURCE GEOMETRY 2 22.3 34 100 0.34 1 under-neath trunk 22.3 34 10 3.40 Two times 10 x 20 cm² 70.2 Nr. Release Location Release Temp. (K) Total Mass (g) Duration (s) H2-Rate (g/s) Vent Openings Garage Volume (m³) CASE HYDROGEN SOURCE GEOMETRY

A thermal energy deposition of 1 Watt into a cryogenic LH2-tank leads to a

boil-off of 170 g of gaseous hydrogen per day

Assume here 5 release pulses per day, 34 g H2 each, with two different release rates

SLIDE 6

6

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

WHAT ARE THE IMPORTANT RISK DETERMINING PARAMETERS?

Large spectrum of events possible, ranging from zero risk to destruction
f garage
What are the parameters influencing the outcome of such a leak scenario?
Obvious first step is to understand mixture generation, defines initial and

boundary conditions for further accident development …. …. …. …. …. …. …. …. …. …. …. …. …. …. …. ….

SLIDE 7

8

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Analysis of a hydrogen release scenario in a private garage Analysis tools and related physics Gas transport and mixing Sequence of analysis steps

SLIDE 8

9

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

AN INITIAL ESTIMATE ON HYDROGEN CONCENTRATION

We can make a first estimate on the hydrogen concentration in the garage by using a

single volume approach

Any risk?
Why is result independent of release rate?
Obviously the real situation is more complex
Next approach is a CFD model

% 0.5 m 70 H g /2 l 22.4 H g 34 garage

f

volume released H volume H fraction volume

3 2

2 2 2

≈ ⋅ = ≈

SLIDE 9

10

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

T>1000K IRWST Rekos

Physical models of 3d code GASFLOW (1)

Conservation equations of fluid flow (fully compressible, 3-dim. Navier- Stokes)
Thermophysical properties of components (JANAF)

(internal energy, specific heats, 25 components including two-phase water)

Molecular transport coefficients (CHEMKIN)

(thermal conductivity, dynamic viscosity, binary diffusion coefficients)

Convective and radiative heat transfer between gas and structure
Heat conduction within structures
Condensation and vaporization of water (film, droplets, sump)

SLIDE 10

11

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

T>1000K IRWST Rekos

Physical models of 3d code GASFLOW (2)

Lumped- parameter sump model
Boundary layer model for wall shear stress
Turbulence modeling (algebraic, k-ε)

(effects on molecular transport coefficients)

Accident mitigation measures

(Recombiner and igniter models, containment inertisation)

Ventilation systems

(1-dim. ducts, pipes, junctions, blowers, dampers, valves, filters, etc)

SLIDE 11

12

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

GASFLOW EQUATIONS

Fully compressible Navier-Stokes, expressed in integral form for finite volume discretisation

Momentum conservation

∂ ∂t

( )

dV dS pd dV dS

V V S

ρ ρ ρ τ u u A

S

g A

2 S S

= + −

∫ ∫ ∫ ∫ ∫

( )

D A

S d m V

dS S dV + ∫

∫

momentum flux pressure gradient gravity viscous stresses drag from internal surfaces & flow restrictions momentum sources

Mass conservation Total mass:

∂ ∂ ρ ρ

ρ

t dV dS S dV

V V

= uA

S

∫ ∫ ∫

+

Component α :

( )

∂ ρ ρ

α α α ρ α

∂t dV dS dS S dV

V V

uA

J A

S S

= +

∫ ∫ ∫ ∫

,

convection diffusion

chem. reaction

two-phase change convection sources (inflow, droplet depletion)

Internal energy conservation

( )

edV e dS p d p V V t dV

V V

ρ ρ ∂ ∂ u A

uA

S

S S

= − ⎛ ⎝ ⎜ ⎞ ⎠ ⎟

∫ ∫ ∫ ∫

( )

qA

S

dS S dV

e V

+ ∫

∫

e x e = ∑

α α α

pV work due to phase change energy flux energy sources (thermal, conductivity, ........) (combustion, phase change, heat transfer)

∂ ∂t

HO

2

convection pV work

J.R. Travis et al, Report FZKA-5994 (1998)

x

R6 R5 Ring Room R9 He- Source 10 Vol % He Isosurface R6 R5 Ring Room R9 He- Source 10 Vol % He Isosurface Vertical He- jet into air (42 m/s) for 200 s Vertical He- jet into air (42 m/s) for 200 s

5 10 15 20 200 400 600 800 Time (s)

9KP000K76 FZK/GASFLOW

5 10 15 20 200 400 600 800 Time (s)

9KP000K76 FZK/GASFLOW

Vertical distance from injection location (m) 40 80 2 4 6 FZK/GASFLOW 6kp270M18 Vertical distance from injection location (m) 40 80 2 4 6 FZK/GASFLOW 6kp270M18 5 10 15 200 400 600 800 Time (s)

6KP270M18 FZK/GASFLOW

5 10 15 200 400 600 800 Time (s)

6KP270M18 FZK/GASFLOW

Distance from jet axis (m)

2
1

1 2 40 80 x x x x x x x x

EXPERIMENT FZK/GASFLOW

x Distance from jet axis (m)

2
1

1 2 40 80 x x x x x x x x

EXPERIMENT FZK/GASFLOW

x

SLIDE 12

13

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

GASFLOW VERIFICATION

3d code GASFLOW used and developed at FZK for hydrogen distribution simulation.

Large verification matrix:

Report FZKA-7085 (2005), www.fzk.de/hbm

SLIDE 13

14

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

EXAMPLE FOR RECENT GASFLOW VERIFICATION

German national benchmark, test TH7 in Thai facility with condensation
Blind pressure prediction of CFD codes

lower source Twall > 330K Twall < 300K

steam 1.0 1.2 1.4 1.6 1.8

[bar] [sec]

CFX STAR-CD Fluent Gothic GASFLOW Experiment

upper inj 35 g/s lower inj 1 35 g/s lower inj 2 5 g/s

Inj. points
P. Royl, IKET

SLIDE 14

15

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Analysis of a hydrogen release in a private garage Analysis tools and related physics Gas transport and mixing Sequence of analysis steps Mixture generation

SLIDE 15

16

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

GASFLOW SIMULATION OF GARAGE SCENARIO

Isosurface with 4 vol% H2, depicts flammable mixture in garage

Case 1: release rate 3.4 g H2 / s for 10 seconds

volume fraction H2

SLIDE 16

17

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

GASFLOW SIMULATION OF GARAGE SCENARIO

Isosurface with 4 vol% H2 , depicts flammable mixture in garage

Case 2: release rate 0.34 g H2 / s for 100 seconds

volume fraction H2

SLIDE 17

18

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

RESULTING HYDROGEN CLOUD IN GARAGE

Computed dimension of combustible

H2-air cloud in garage (4…75% H2)

Characteristic size of combustible cloud

expressed as dCC = (Vcc)1/3

Combustible cloud size strongly dependent
n release rate, is result of balance between

source strength and sinks, or release rate and mixing mechanisms

Case 2

0.34 g H2/s for 100s

Case 1

3.4 g H2/s for 10 s

d (cm)

cc

10 20 30 40 50 60 70 80 90 50 100 150 200 250

stable transient

20 40 60 80 100 120 140 160 180 10 20 30 40 50 60

d (cm)

cc

Time (s) Time (s)

SLIDE 18

19

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

WHAT IS RISK FROM COMBUSTIBLE CLOUD?

Case 1 Case 2

How would you judge the hazard in both cases?
Who would switch on lights in the garage?
What physical quantities determine hazard potential of a combustible H2-air cloud?

SLIDE 19

20

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Analysis of a hydrogen release in a private garage Analysis tools and related physics Combustion regimes Transport and mixing Mixture generation Sequence of analysis steps

SLIDE 20

21

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

COMBUSTION REGIMES

H2 – air mixtures can burn in different modes / combustion regimes

Inert, no stable flame propagation vfl = 0 Laminar deflagration vfl ≈ 1 m/s, Ma << 1 Fast turbulent deflagration vfl ≈ 300 m/s, Ma ≈ 1 Detonation vfl ≈ 1500 m/s, Ma >> 1

Change of mode possible by transition process

Ignition Flame acceleration Deflagration-to-detonation transition (DDT)

SLIDE 21

22

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

PEAK OVERPRESSURES FROM HYDROGEN-AIR FLAMES

The maximum flame speed generally governs the damage potential
Which combustion regime develops for given mixture and geometry?
How fast can it burn?

x p pmax

0.01 0.1 1 10 100 7 10 100 1000 flame speed (m/s)

v

e r p r e s u r e r a t i

(

p

p

) / p m a x

FZK-tube (closed) RUT vented FLAME vented planar model,12%H planar model,28%H

2 2

Maximum acceptable static load for typical inner containment structures (1 ton / m2)

(pmax-p0)/p0

p0

SLIDE 22

23

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

IGNITION

Combustion requires an ignition source

and a burnable mixture

Many potential ignition sources exist
More than 90% of incidents with GH2 lead

to ignition, cause often unknown

Ignition difficult to exclude in a hydrogen

safety analysis, conservatively the presence of an ignition source may be assumed

Controlling factor is then flammability of

mixture, well known for H2-air

G:1 G:7 G:5 G:4 G:3 G:0 G:2 G:8 G:6 G:9 5 10 15 20 25 30 35 40 45 Fraction of ignition source [%]

Ignition sources

G:1 open fire G:2 mechanical spark G:3 electrical spark G:4 hot surface G:5 static discharge G:6 catalytic surface G:7 self-ignition G:8 others G:9 unknown G:0 no ignition Survey of 287 accidents with hydrogen Kreiser et al, Report Univ. Stuttgart IKE 2-116 (1994)

SLIDE 23

24

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

FLAME ACCELERATION

Conservative conditions for flame acceleration in hydrogen mixtures were investigated

in closed obstructed tubes, e.g. FZK 12m-tube

SLIDE 24

25

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

RESULTS OF FLAME ACCELERATION EXPERIMENTS

D

f l a m e v e l

c

i t y ( m / s )

10 20 30 40 50 60 70

x/D

200 400 600 800 1000 1200

v, m/s BR=0.6 H2-Air

174m m

10%H2 11%H2 15%H2

520m m

10%H2 11%H2

80m m

10%H2 11%H2 13%H2

350m m

9%H2 11%H2 17.5%H2

Lean hydrogen mixtures in obstructed

tubes with different tube diameters D and 60% blockage ratio (BR)

Two distinct regimes with slow and fast

flame propagation are observed H

5 10 15 20 25 30 35

x/D

200 400 600 800 1000 1200

v, m/s 350 mm BR=0.6 H2-air-CO2

φ = 0.5

40%CO2 35%CO2 30%CO2 27.5%CO2 25%CO2 20%CO2 15%CO2 10%CO2 5%CO2 2.5%CO2 0%CO2

FA x

SLIDE 25

26

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

FLAME ACCELERATION CRITERION

Summary of experiments with different

H2-O2- dilutend (N2, Ar, He) mixtures in obstructed tubes of different scales

Each point represents one

experiment

Results of data evaluation:

expansion ratio σ is mixture property which governs flame acceleration limit

No flame acceleration for

σ < 3.75 ± 0.1 (10.5% H2 in dry air)

fast flames unstable flames slow flames

σ = 3.75 FZK FZK FZK FZK FZK / KI KI KI KI KI

Tube diameter L / Laminar flame thickness δ x 10-3

different length scales L

FZK FZK-tube: 350 mm KI Driver tube: 174 mm Torpedo tube : 520 mm RUT : 2400 mm FZK / KI δ ~ 0.1 mm ~

3 0.1 1.0 10.0 4 5 6 7 global quenching quenching-reignition choked flames quasi-detonation T 300 K ~ ~

Expansion ratio σ (= ρub/ρb)

fast flames unstable flames slow flames

σ = 3.75 FZK FZK FZK FZK FZK / KI KI KI KI KI

Tube diameter L / Laminar flame thickness δ x 10-3

different length scales L

FZK FZK-tube: 350 mm KI Driver tube: 174 mm Torpedo tube : 520 mm RUT : 2400 mm FZK / KI δ ~ 0.1 mm ~

3 0.1 1.0 10.0 4 5 6 7 global quenching quenching-reignition choked flames quasi-detonation T 300 K ~ ~

Expansion ratio σ (= ρub/ρb) D

D In lecture notes

SLIDE 26

27

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Fully obstructed tube (prototypic mode B)

accelerating flame DDT stable (quasi) detonation spark ignition p H / air

2

burned gas

bstacle section

flame precursor shock conus 1 m 5-6 m

Partially obstructed tube with conus (prototypic mode A)

0,5 m

spark ignition

Shock tube with conus (idealized mode A)

p He H / air

2

high pressure section burst membrane low pressure section conus 3 m/1m 9 m 0.35 m 3m 1m

DEFLAGRATION-TO DETONATION TRANSITION

Two different modes of DDT have been
bserved
shock focussing
detonation on-set in turbulent

flame brush

Present here one example for DDT

with pressure wave emitted from an obstructed region and focussed in a conus

SLIDE 27

28

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

TURBULENT DEFLAGRATION EXPERIMENT WITHOUT DDT

Partially obstructed tube with conus, 15 % hydrogen in air

5 m 350 mm

SLIDE 28

29

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

TURBULENT DEFLAGRATION EXPERIMENT WITH DDT

Partially obstructed tube with conus, 16.5 % hydrogen in air
Result: focussing of pressure waves emitted from a fast turbulent flame can trigger

a detonation on other parts of the system

SLIDE 29

30

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

CRITERION FOR DDT

Correlation of all experimental

data with given definitions of D and detonation cell size data shows that detonations are

nly possible for D/λ > 7
Current uncertainty in detonation

cell size λ ≈ factor 2

2 3 5 2 3 5

100 1000 10000

Geometrical size L, mm

2 3 5 2 3 5 2 3 5 2

1 10 100

D /λ

No DDT, BR<0.5 No DDT, BR>0.5, rooms DDT, BR<0.5 DDT, BR>0.5, rooms L = 7λ λ accuracy limits d1 d1 d1 d1 d1 d1 d1 d1 c1 c1 t4 t4 t4 c2 d2 d2 d2 d2 d2 d2 d2 d2 g3 s1 s1 s1 s1 s1 s1 s1 s1 s1 s1 s1 s1 t3 t3 t3 t3 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 f1 f3 r2 r2 r5 r5 r5 r5 r5 r5 r5 r5 r5 r5 r5 r5 r5 r5 r5 r5 r5 r6 r6 v2 b1 b1 b1 b1 b1 b2 b2 b2 b2 b2 b3 b3 b3 b3 b3 b3 b4 b4 b4 b4 b4 b4 b4 b5 b5 b5 b5 b5 b6 b6 b6 b6 b6 b6 m3 m3 m4 m4 m4 m5 c3 d3 d3 d3 d3 d3 d3 d3 d3 d3 d3 d3 d3 d3 g1 g1 g1 g1 g1 g1 g2 g2 g6 g6 g6 g6 g7 g7 g7 g7 g8 g8 g8 g9 g9 r1 s2 s2 s2 s2 s2 s2 s2 s2 t1 t1 r3 r4 r4 r4 r4 r4 r4 r4 r4 r4 r7 r7 v1 d4 ri ri ri ri ri ri ri ri d1 d1 t4 d2 d2 g3 s1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 f1 f3 r2 r2 r5 r5 r6 v2 b1 b1 b2 b2 b3 b3 b4 b5 b5 b6 b6 m3 m3 m3 m4 m4 m4 m5 m1 m2 m6 m6 m6 m7 m7 m7 m8 m8 m8 g1 g1 g1 g7 g7 g8 g8 r1 r3 r4 r4 r4 r4 r4 r4 r4 r4 r7 r7 v1 ri ri ri ri ri ri

D / λ

Characteristic geometrical size of reacting mixture D (mm)

kein kein Fehlergrenze

Experiments on DDT in differently sized and shaped facilities have shown that a certain

minimum scale is required for DDT

In accident scenarios D/λ can vary

by orders of magnitude, criterion has therefore predictive capability

SLIDE 30

31

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

DETONATION CELL SIZES

Detonation cell sizes (in cm) of

H2-air-steam mixtures at 375 K and 1 bar initial pressure. Dry hydrogen concentration is defined as H2 / (H2 + air)

H ( t r

c

k e n ) ( v

l

% ) 2

H0 (vol%)

2

State of the Art Report by a Group of Experts „Flame Acceleration and Deflagration – to – Detonation Transition in Nuclear Safety“, Nuclear Safety NEA/CSNI/R(2000)7, August 2000

SLIDE 31

32

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

SUMMARY OF CRITERIA

Criteria for possible occurrence of fast combustion regimes were evaluated from many

experiments with various H2-mixturs on different scales

Transition phenomena cannot be modeled numerically on large building scale
Criteria allow selection of fastest possible combustion mode from computed

H2-air cloud composition and scale

H 2

8% x=16% 4%

Luft H 2

8% x = 16% 4%

air σ index = < 1 ausgeschlossen = > 1 möglich

σ (Durchschnittsmischung) σ (T)

critical

Flammenbeschleunigung

σindex

=

< 1 excluded = > 1 possible σ(average mixture) σ

(T)

critical

7λ

Übergang zur Detonation

= < 1 exluded = > 1 possible (Cloud volume)

1/3 σindex >1

7λ (average mixture) D

Detonation transition Flame acceleration

SLIDE 32

33

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Analysis of a hydrogen release in a private garage Analysis tools and related physics Combustion regimes Distribution and mixing Mixture generation Sequence of analysis steps Hazard potential

SLIDE 33

34

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

COMPUTED HAZARD PARAMETERS FOR GARAGE SCENARIOS

CASE 2

0.34 g H2/s for 100s

Volume of cloud with potential for

spontaneous flame acceleration (10.5 to 75 % H2)

DDT index of cloud

(10.5 to 75 % H2)

Dimension of combustible cloud,

4 to 75 % H2, dcc = (Vcc)1/3

20 40 60 80 100 120 140 160 180 10 20 30 40 50 60

d (cm)

cc

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 10 20 30 40 50 60

V (m³)

fa

0.5 1 1.5 2 2.5 3 10 20 30

DDT no DDT

40 50 60

D 7λ

CASE 1

3.4 g H2/s for 10 s

no DDT

0.2 0.4 0.6 0.8 1.0 1.2 1.4 50 100 150 200 250

D 7λ

V (m³)

fa

0.02 0.04 0.06 0.08 0.10 0.12 50 100 150 200 250

d (cm)

cc

10 20 30 40 50 60 70 80 90 50 100 150 200 250

Time (s) Time (s) Time (s) Time (s) Time (s) Time (s)

SLIDE 34

35

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

HAZARD POTENTIAL FOR GARAGE SCENARIOS

Risk parameters show strong dependence on H2 release rate
Case 1:

(3.4 g H2/s)

Only Case 1 followed in further safety analysis
Continuous potential for slow deflagration

(≈ 20 g of 34 g)

potential for supersonic combustion regimes (and ignition)

during the release period

high release rate not tolerable without mitigation

measures

only small potential for slow deflagrations, natural

mixing processes sufficient

release rate (and mass) seems tolerable for

present garage design

Case 2:

(0.34 g H2/s)

SLIDE 35

36

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

COMBUSTION EXPERIMENTS FOR CASE 1

Up to 20 g of hydrogen would be in burnable concentrations
A significant part of this could potentially burn with high flame speeds
What would be pressure loads and consequences from a local explosion in

the garage?

Outcome uncertain, experiments performed in test chamber simulating the

garage

SLIDE 36

37

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Analysis of a hydrogen release in a private garage Analysis tools and related physics Combustion regimes Distribution and mixing Mixture generation Sequence of analysis steps Combustion (Experiments) Hazard potential

SLIDE 37

38

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

COMBUSTION UNITS

To obtain conservative pressure loads, combustion units were developed providing the

fastest possible flame speed for a given H2 mass

Cubes were made for 0.5, 1, 2, 4, 8 and 16 g of H2, which can be inserted into each other
Wire grids 6.5 x 0.65 mm, 12 layers between cubes

4 g 0.5 g Hydrogen injection device cubes covered with plastic, filled with stoichiometric H2 - air mixture

SLIDE 38

39

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

R

Pressure transducers mH2 pmax i v

∫

> Δ +

Δ = ) (

p

dt t p I UNCONFINED TEST OF COMBUSTION UNIT

Peak overpressure and impulse

measured as function of distance to characterize blast effects from combustion unit

SLIDE 39

40

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

FLAME SPEEDS IN COMBUSTION UNITS

0.5 1 2 4 8 16 200 400 600 800 1000 1200 1400 1600 1800 2000 0.1 0.2 0.3 0.4 0.5 R, m Flame speed, m/s

8 g 8 g 16 g 16 g 4 g cube border

The flame acceleration inside the combustion units was measured with photodiodes
For 8 and 16 g H2 detonation speeds are obtained at the outer edge of the cube

center of the cube

SLIDE 40

41

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

MAXIMUM OVERPRESSURE VS DISTANCE

Measured peak overpressures Δp+ in unconfined tests with combustion units of 0.5 to 16 g H2
Data are well reproducible

Δp+ t I+

1 2 3 4 5 5 10 Distance R from center of cube, m Peak overpressure Δp+, bar

2g 2g 1g 1g 0.5g

2 4 6 8 10 12 5 10

16g 16g 8g 8g 4g

Distance R from center of cube, m Peak overpressure Δp+, bar

SLIDE 41

42

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

IMPULSE VS DISTANCE

Δp+ t I+

Measured positive impulse I+ values from unconfined

combustion units

20 40 60 80 100 120 140 160 180 5 10 Distance R, m Impulse, Pa*s

16g 16g 8g 8g 4g

10 20 30 40 50 60 70 80 90 100 5 10 Distance R, m Impulse, Pa*s

2g 2g 1g 1g 0.5g

SLIDE 42

43

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

SCALED PEAK OVERPRESSURES VS DISTANCE

Use of Sachs scaling collapses measured peak overpressures to universal correlation for

≥ 1 g H2, E = total energy of explosive charge

Combustion units provide conservative and well defined overpressures

0.01 0.1 1 10 2 4 6 Scaled distance R (p0 / E)1/3 Scaled peak overpressure ΔP+/p0

16g 16g 8g 8g 4g Gas detonation TNT (= energy)

0.01 0.1 1 10 5 10

2g 2g 1g 1g 0.5g Gas detonation TNT (= energy)

Scaled peak overpressure ΔP+/p0 Scaled distance R (p0 / E)1/3

SLIDE 43

44

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

TEST CELL FOR GARAGE SIMULATION

Dimensions 5.5 x 8.5 x 3.4 m,

about 160 m3

Air flow ≤ 24.000 m3/h,

up to 1 air exchange in 24s

Controlled air flows in chamber

possible

All ventilation systems

explosion protected

Test cell used for simulation of

garage /confined volume

Ventilation ducts Ventilation turbines Test chamber (garage) Lower compartment

SLIDE 44

45

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

100 200 300 400 500 50 100 150 200 250 300 200 400 600 800 100 200 300 400 500 50 100 150 200 250 300 200 400 600 800 Druckaufnehmer Beschleunigungs- sensoren

3A 3B 1B 13A 7B 2A 1A 4A 5A 6A 7A 8A 15A 14A 2B

4B

5B 6B 8B 16A

H ö h e [ c m ] Breite [cm] Länge [cm] Tür Zünd-

rt

INSTRUMENTATION OF GARAGE

The instrumentation included pressure and acceleration sensors at different locations,

covering flat surfaces, (2d) edges and (3d) corners

Pressure sensors Acceleration sensors location of combustion unit

Height (cm) Width (cm) Length (cm)

SLIDE 45

46

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

LOCAL HYDROGEN EXPLOSIONS IN A GARAGE

Experiment with 8g H2 H2 - mass:

1g
2g
4g
8g
16g

SLIDE 46

47

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

0,188 0,190 0,192 0,194 0,196 0,198 0,200 0,202 0,204

0,020
0,015
0,010
0,005

0,000 0,005 0,010 0,015 0,020

Beschleunigungsaufnehmer 1B 1g H2 Experiment2 Experiment3 Volt Zeit [s]

0,188 0,190 0,192 0,194 0,196 0,198 0,200 0,202 0,204

0,020
0,015
0,010
0,005

0,000 0,005 0,010 0,015 0,020

Beschleunigungsaufnehmer 1B 1g H2 Experiment2 Experiment3 Volt Zeit [s]

0,19 0,20 0,21 0,22 0,23

0,03
0,02
0,01

0,00 0,01 0,02 0,03 0,04

Beschleunigungsaufnehmer 3B 1g H2 Experiment2 Experiment3 Volt Zeit [s]

0,19 0,20 0,21 0,22 0,23

0,03
0,02
0,01

0,00 0,01 0,02 0,03 0,04

Beschleunigungsaufnehmer 3B 1g H2 Experiment2 Experiment3 Volt Zeit [s]

0,20 0,21 0,22 0,23 0,24 0,25 0,26 0,27

0,02
0,01

0,00 0,01 0,02 0,03

Beschleunigungsaufnehmer 9A 1g H

2

Experiment2 Experiment3 Volt Zeit [s]

0,20 0,21 0,22 0,23 0,24 0,25 0,26 0,27

0,02
0,01

0,00 0,01 0,02 0,03

Beschleunigungsaufnehmer 9A 1g H

2

Experiment2 Experiment3 Volt Zeit [s]

0,192 0,194 0,196 0,198 0,200 0,202 0,204

0,04
0,03
0,02
0,01

0,00 0,01 0,02 0,03 0,04 0,05 0,06 0,07 0,08 0,09

Druckaufnehmer 16A 1g H2 Experiment2 Experiment3 Überdruck [bar] Zeit [s]

0,185 0,190 0,195 0,200 0,205 0,210

0,10
0,05

0,00 0,05 0,10 0,15 0,20 0,25

Druckaufnehmer 2B 1g H2 Experiment1 Experiment2 Experiment3 Überdruck [bar] Zeit [s]

0,190 0,195 0,200 0,205 0,210 0,215

0,05
0,04
0,03
0,02
0,01

0,00 0,01 0,02 0,03 0,04 0,05 0,06

Druckaufnehmer 5A 1g H2 Experiment1 Experiment2 Experiment3 Überdruck [bar] Zeit [s]

0,192 0,194 0,196 0,198 0,200 0,202 0,204

0,04
0,03
0,02
0,01

0,00 0,01 0,02 0,03 0,04 0,05 0,06 0,07 0,08 0,09

Druckaufnehmer 16A 1g H2 Experiment2 Experiment3 Überdruck [bar] Zeit [s]

0,192 0,194 0,196 0,198 0,200 0,202 0,204

0,04
0,03
0,02
0,01

0,00 0,01 0,02 0,03 0,04 0,05 0,06 0,07 0,08 0,09

Druckaufnehmer 16A 1g H2 Experiment2 Experiment3 Überdruck [bar] Zeit [s]

0,185 0,190 0,195 0,200 0,205 0,210

0,10
0,05

0,00 0,05 0,10 0,15 0,20 0,25

Druckaufnehmer 2B 1g H2 Experiment1 Experiment2 Experiment3 Überdruck [bar] Zeit [s]

0,190 0,195 0,200 0,205 0,210 0,215

0,05
0,04
0,03
0,02
0,01

0,00 0,01 0,02 0,03 0,04 0,05 0,06

Druckaufnehmer 5A 1g H2 Experiment1 Experiment2 Experiment3 Überdruck [bar] Zeit [s]

0,190 0,195 0,200 0,205 0,210 0,215

0,05
0,04
0,03
0,02
0,01

0,00 0,01 0,02 0,03 0,04 0,05 0,06

Druckaufnehmer 5A 1g H2 Experiment1 Experiment2 Experiment3 Überdruck [bar] Zeit [s]

REPRODUCIBILITY OF MEASURED DATA

The experiment with 1 g H2

was performed three times

Acceleration and pressure

sensors show very good reproducibility of measured signals

Complex, but reproducible

pressure waves are created in confined local explosions

f H2-air mixtures

2

SLIDE 47

48

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

0,182 0,183 0,184 0,185 0,186 0,187 0,188

0,4

0,0 0,4 0,8 1,2 1,6

1g 2g 4g 8g 16g Überdruck [bar] Zeit [s]

0,182 0,183 0,184 0,185 0,186 0,187 0,188

0,4

0,0 0,4 0,8 1,2 1,6

1g 2g 4g 8g 16g Überdruck [bar] Zeit [s] 0,193 0,194 0,195 0,196 0,197

0,1

0,0 0,1 0,2 0,3 0,4 1g 2g 4g 8g 16g Überdruck [ bar] Zeit [ s]

COMPARISON OF OVERPRESSURES

Pressure sensor 2 B,

floor near combustion unit

Pressure sensor 8 A,

back wall, half wall height

Pressure signals very consistent in timing, amplitudes increase systemarically with H2 mass,

reproducible pattern of reflected pressure waves in confined volume. Overpressure [bar] Overpressure [bar] Time [s] Time [s]

SLIDE 48

49

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Analysis of a hydrogen release in a private garage Analysis tools and related physics Combustion regimes Turbulent deflagration Distribution and mixing Mixture generation Sequence of analysis steps Combustion Hazard potential

SLIDE 49

50

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Re =

t

u'L t ν

Turb. integral length scale Lt / Laminar flame thickness δ

Evolution in

bstructed

large scale combustion

Turbulence intensity u‘/ Laminar flame velocity SL Da < 1 Da = 1 Da > 1, Ka >1 Ka = 1 Ka < 1 Ret < 1 Da = Turb. Transport time (makro) Laminar reaction time Lt / u' δL/ SL = Ka = Laminar reaction time Turbulent transport time (Kolmogorov scale) δL / SL lΚ / u 'K = Flame shapes PDFs laminar flame folded flame wrinkled flame homogeneous reaction Unburned gas Burned gas Reaction zone Ka < 1 Da > 1 Da = 1 Da << 1 thickened flame

TURBULENT DEFLAGRATION REGIMES

SLIDE 50

51

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

COM3D EQUATIONS

A. Kotchourko, IKET
COM3D under development at FZK for

simulation of turbulent deflagration

SLIDE 51

52

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

COM3D VERIFICATION (1)

Large scale experiments performed in RUT facility near Moscow (FZK, CEA, partly

NRC), H2-air, H2-air-steam

Total length 62 m
Total volume 480 m3
First channel with obstacles
Second part without obstacles

SLIDE 52

53

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

A. Kotchourko, IKET

COM3D VERIFICATION (2)

D i s t a n c e ( m ) O v e r p r e s u r e ( b a r )

0,2 0,3 0,4 0,5 0,6

Time (s)

5 10 15 20 25 30 35 Experiment RUT13 H=11% BR=0,3 p=1bar T=283K

2

COM-Code c=7

f

Experiment

2 4 6 8 10 12

D i s t a n c e ( m ) O v e r p r e s u r e ( b a r )

2 4 6 8 10 12 14 16

100 150 200 250 300

Time (ms)

5 10 15 20 25 30 35 Experiment RUT21 H=12,5% BR=0.6 p=1bar T=283K

2

COM-Code c=7

f

Experiment

D i s t a n c e ( m ) O v e r p r e s u r e ( b a r ) D i s t a n c e ( m ) O v e r p r e s u r e ( b a r )

2 4 6 8 10

150 200 250 300 350

Time (ms)

5 10 15 20 25 30 35 Experiment RUT23 H=11,2% BR=0.6 p=1bar T=283K

2

Experiment COM-Code, c=6

f

Numerical simulation of large scale RUT experiments with hydrogen-air and

hydrogen-air steam mixtures. Standard k-ε and Eddy-Break-up model.

Venting in experiments, no venting in simulation

Distance (m) Distance (m) Distance (m)

SLIDE 53

54

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Analysis of a hydrogen release in a private garage Analysis tools and related physics Combustion regimes Turbulent deflagration Distribution and mixing Mixture generation Sequence of analysis steps Combustion simulation Hazard potential

SLIDE 54

55

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

SIMULATION OF UNCONFINED TESTS

The unconfined tests with different combustion units were simulated with COM3D
The COM3D combustion model was fitted to the measured flame speed in the combustion units
The calculated peak overpressures agree with the experimental values and follow Sachs scaling

Δp+/p0 R (p0 / E)1/3

SLIDE 55

57

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

COM3D COMBUSTION SIMULATION

3d pressure field, calculated isosurface for 1.1 bar
Test with 8g H2

SLIDE 56

58

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

COMPARISON OF OVERPRESSURES

Good agreement, remaining differences are due to geometry simplification and rigid wall model

in simulation

verpressure [bar]
verpressure [bar]
verpressure [bar]
verpressure [bar]

Time [s] Time [s] Time [s] Time [s]

SLIDE 57

59

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Analysis of a hydrogen release in a private garage Analysis tools and related physics Combustion regimes Turbulent deflagration (Detonation) Structural response Distribution and mixing Mixture generation Sequence of analysis steps Combustion simulation Hazard potential

SLIDE 58

60

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

STRUCTURAL RESPONSE

What are effects of blast loads on the structure?

SLIDE 59

61

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

SINGLE-DEGREE-OSCILLATOR MODEL FOR STRUCTURAL RESPONSE

Simplest model for structural response is

SDO model

Describes ground mode (first harmonic)
f structural element which is

represented by lumped values for mass, stiffness and damping of motion

Tool to understand basic effects of

transient pressure loads on global displacement of element

In FEM analysis also higher modes

included, but superposition of different effects, results not so transparent

xmax

k p(t) m t

D

SLIDE 60

62

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

BLAST LOADED ELASTIC OSCILLATOR (1)

0) (t x 0) x(t with e p kx x m e p kx

F

x m

load load

T t T t i i

= = = = Δ = + Δ + = =

− + − +

∑

& & & & &

( ) ( )

⎥ ⎦ ⎤ ⎢ ⎣ ⎡ + ω − ω + = Δ

− +

load

t/T load load load

e t cos ωT t sin ωT ωT /k p x(t)

2 2

1 /k p x x

max +

Δ = = & &

( )

sc

1/2

T 2π period

scillator

k/m ω = = =

Equation of Motion

Static maximum deflection

Solution

where

Damage is determined by maximum displacement xmax,

can be found from solution by setting x(t) = 0

Scaled displacement = f(scaled loading time)

) T f( /k p x

load max

ω = Δ

+

k m t x

load

T / t

e p ) t ( p

− +

Δ =

t p(t) Δp+

.

scaled deformation xmax scaled loading time

T m k T = ω

T = 1 ms 5 ms

scaled deformation xmax scaled loading time

T m k T = ω

/ (P* / k)

SLIDE 61

63

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

BLAST LOADED ELASTIC OSCILLATOR (2)

Asymptotes for maximum deflection /deformation

can be computed from energy balances

Quasistatic loading realm (T load >>Tosc)
strain energy = work on structure

½ kx2

max = Δp+ · xmax

2 k / p xmax = Δ

+

Impulsive loading realm ( T load << Tosc)
initial kinetic energy = strain energy

2 max 2 2 max 2

kx 2 1 2m I kx 2 1 mv 2 1 = =

load load / max

T T ) m k ( k / p x ω = = Δ

+ 2 1

r

I ) km 1 ( x

1/2 max

⋅ =

maximum deformation is proportional to blast wave impulse I

Impulsive asymptote Quasistatic asymptote

Scaled loading time ωT

load T / t

T p e p I

load

⋅ Δ = Δ =

+ ∞ − +

∫

dynamic maximum deflection is two times static deflection (DLF = 2) Scaled displacement

2

W.E. Baker, P.A. Cox, P.S. Westine, J.J. Kulesz, R.A. Strehlow; Explosion Hazards And Evaluation; Fundamental Studies in Engineerings, 5; Elsevier

SLIDE 62

64

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH) max load load max 2 1 load max

x ~ I T p T 1 x (km) p T 1 kx p = Δ = Δ ω = Δ

+ + +

OSCILLATOR RESPONSE: ANOTHER VIEW

Often oscillator response is presented with inverted ordinate and unscaled load parameters

Δp+ and Tload

Quasistatic asymptote

Maximum deflagration xmax is only proportional to applied peak overpressure Δp+, independent of load duration

Impulsive asymptote

Maximum deflection xmax is proportional to applied impulse positive peak

verpressure

logΔp+ log of positive overpressure duration Tload I2 I1 xmax1 xmax2 xmax2 xmax1 Δp+

2

Δp+

1

2 kx p 2 1 kx p

max max

= Δ = Δ

+ +

SLIDE 63

65

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

Analysis of a hydrogen release in a private garage Analysis tools and related physics Combustion regimes Turbulent deflagration (Detonation) Structural response Distribution and mixing Mixture generation Sequence of analysis steps Combustion simulation Hazard potential Consequences

SLIDE 64

68

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

STRUCTURAL DAMAGE FROM CONFINED LOCAL H2 EXPLOSIONS IN GARAGE

10000 Duration of positive overpressure T [ms] Positive peak overpressure Δp

+ [Pa]

10

3

10

4

10

5

10

6

Δp+ I+ t T+=2I

+/Δp+

T+

Unconfined gaseous detonations

1,4m 1,4m 2,4m 2,4m 2,8m 5.6m 4,0m 6,3m 4,0m

FZK Exp. 2g H

2

FZK Exp. 16g H

2

0.1 1 10 100 1000

Partial demolition, 50%

75% of walls destroyed

Major structural damage, some wrenched load bearing members fails Minor structural damage, wrenched joints and partitions 2g H

2

1 m glass breakage

Area 1m

2

4.7 mm thick Area 3m

2

4.7 mm thick

16g H

2

2 m 5 m 450 Pas 300 Pas 110 Pas

10000 [ms] Δ + [Pa] 10

3

10

4

10

5

10

6

Δp+ I+ t T+=2I

+/Δp+

T+

Unconfined gaseous detonations

1,4m 1,4m 2,4m 2,4m 2,8m 5.6m 4,0m 6,3m 4,0m

FZK Exp. 2g H

2

FZK Exp. 16g H

2

0.1 1 10 100 1000

Partial demolition, 50%

75% of walls destroyed

Major structural damage, some wrenched load bearing members fails Minor structural damage, wrenched joints and partitions 2g H

2

1 m glass breakage

Area 1m

2

4.7 mm thick Area 3m

2

4.7 mm thick

16g H

2

2 m 5 m

10000 [ms] Δ + 10

3

10

4

10

5

10

6

Δp+ I+ t T+=2I

+/Δp+

T+

Unconfined gaseous detonations

1,4m 1,4m 2,4m 2,4m 2,8m 5.6m 4,0m 6,3m 4,0m

FZK Exp. 2g H

2

FZK Exp. 16g H

2

0.1 1 10 100 1000

Partial demolition, 50%

75% of walls destroyed

Major structural damage, some wrenched load bearing members fails Minor structural damage, wrenched joints and partitions 2g H

2

1 m glass breakage

Area 1m

2

4.7 mm thick Area 3m

2

4.7 mm thick

16g H

2

2 m

10000

+[ms]

Δ + 10

3

10

4

10

5

10

6

Δp+ I+ t T+=2I

+/Δp+

T+ Δp+ I+ t T+=2I

+/Δp+

T+

Unconfined gaseous detonations

1,4m 1,4m 2,4m 2,4m 2,8m 5.6m 4,0m 6,3m 4,0m

FZK Exp. 2g H

2

FZK Exp. 16g H

2

FZK exp. 2g H2 FZK exp. 16g H 2

0.1 1 10 100 1000

Partial demolition, 50%

75% of walls destroyed

Major structural damage, some wrenched load bearing members fails Minor structural damage, wrenched joints and partitions 2g H2 1 m glass breakage

Area 1m

2

4.7 mm thick Area 3m

2

4.7 mm thick

16g H2 2 m 5 m 5 m 450 Pas 300 Pas 110 Pas

Results: - windows and light garage components (door) wold break

damage to masonary walls only from nearly explosion of 16 g H2
wooden framework construction would be destroyed

SLIDE 65

71

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

0.1 1 10 100 1000 10000

NASA 16g H2

Positive peak overpressure Δp+ [Pa] 103 104 105 106 lung rupture threshold Δp+ I+ t T+=2I+/Δp+ T+

Unconfined gaseous detonations

1,4m 1,4m 2,4m 2,4m 2,8m 5.6m 4,0m 6,3m 4,0m

FZK Exp. 2g H2 FZK Exp. 16g H2

increasing grade of injury increasing grade of injury 50 % eardrum rupture 1 % eardrum rupture

Baker

Duration of positive overpressure T+ [ms]

1 m 2 m 2g H2 5 m

0.1 1 10 100 1000 10000

NASA 16g H2

Positive peak overpressure Δp+ [Pa] 103 104 105 106 lung rupture threshold Δp+ I+ t T+=2I+/Δp+ T+

Unconfined gaseous detonations

1,4m 1,4m 2,4m 2,4m 2,8m 5.6m 4,0m 6,3m 4,0m

FZK Exp. 2g H2 FZK Exp. 16g H2 FZK Exp. 2g H2 FZK Exp. 16g H2

increasing grade of injury increasing grade of injury 50 % eardrum rupture 1 % eardrum rupture

Baker

Duration of positive overpressure T+ [ms]

1 m 1 m 2 m 2 m 2g H2 5 m 5 m

HUMAN EFFECTS FROM CONFINED LOCAL H2 – EXPLOSIONS IN GARAGE

Results: - high probability of ear drum rupture

no long damage

SLIDE 66

72

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

SUMMARY OF MECHANISTIC SAFETY ANALYSIS OF HYDROGEN BASED ENERGY SYSTEMS

SLIDE 67

73

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

PROCEDURE FOR HYDROGEN SAFETY ANALYSIS

A complete, self-consistent and mechanistic analysis procedure has been developed which

addresses all important physical phenomena of hydrogen behaviour in accidental release scenarios CONSEQUENCE ANALYSIS

Mechanical and thermal loads Structural Human effects

CRITERIA FOR HAZARD POTENTIAL

Flammability y y Flame Acceleration y

COMBUSTIBLE MIXTURE GENERATION

GASFLOW FLAME3D Fast turbulent deflagration COM3D Detonation DET3D Slow deflagration

CONSEQUENCE ANALYSIS

response SDO ABAQUS Detonation transition

COMBUSTIBLE MIXTURE GENERATION

Problem geometry Mitigation Scenario Sources Distribution GP - Program

COMBUSTION SIMULATION

SLIDE 68

74

KIT – die Kooperation von Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH)

MITIGATION MEASURES

The proposed analysis procedure allows identification of possible mitigation measures

for risk reduction

Exlude severe scenarios by design changes
Limit hydrogen sources
Support hydrogen dispersion and mixing processes
Exclude ignition sources
Suppress flame acceleration

(low confinement and turbulence generation)

Avoid detonation transition processes

(lean mixtures, small scale)

Confine consequences

(strong enclosure) Accident progression

If one level of defence has been

ptimized, work
n next barrier

for accident progression