mechanistic safety analysis of hydrogen based … · mechanistic safety analysis of hydrogen based...
TRANSCRIPT
1KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
MECHANISTIC SAFETY ANALYSIS OF HYDROGEN BASED ENERGY SYSTEMS
W. BreitungInstitute for Nuclear and Energy Technologies
Karlsruhe Research CenterGermany
Second European Summer School on Hydrogen Safety, University of Ulster, Belfast,30 July- 8 August 2007
2KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
CONTENT OF PRESENTATION
• Presentation consists of two topics which are treated in parallel
Analysis of hydrogen releasein 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
3KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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 apperentlyno 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 mixturescould develop?
Could they beflammable?
How fast couldthe burn be?
What would bethe pressure
loads?
What could be theconsequences?
4KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
GENERIC ARCHITECTURE OF AN LH2-TANK SYSTEM
Source: EU-Project EIHP-2, Final Report 2004
CRACK!
5KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
INVESTIGATED GARAGE SCENARIOS
222.3341000.34
1under-neath
trunk
22.334103.40Two times
10 x 20 cm²
70.2
Nr.Release Location
Release Temp.
(K)
Total Mass
(g)
Duration(s)
H2-Rate (g/s)
VentOpenings
Garage Volume
(m³)
CASEHYDROGEN SOURCEGEOMETRY
222.3341000.34
1under-neath
trunk
22.334103.40Two times
10 x 20 cm²
70.2
Nr.Release Location
Release Temp.
(K)
Total Mass
(g)
Duration(s)
H2-Rate (g/s)
VentOpenings
Garage Volume
(m³)
CASEHYDROGEN SOURCEGEOMETRY
• 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
6KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
WHAT ARE THE IMPORTANT RISK DETERMINING PARAMETERS?
• Large spectrum of events possible, ranging from zero risk to destructionof 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
…. …. …. ….
…. …. …. ….
…. …. …. ….
…. …. …. ….
8KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Analysis of a hydrogen releasescenario in a private garage
Analysis tools and related physics
Gas transport and mixing
Sequence of analysis steps
9KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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 /2l 22.4 H g 34
garage of volumereleased H volume
H fraction volume 32222
≈⋅
=≈
10KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
T>1000KIRWST
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)
11KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
T>1000KIRWST
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)
12KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
GASFLOW EQUATIONS
Fully compressible Navier-Stokes, expressed in integral form for finite volume discretisation
Momentum conservation
∂∂t
( )dV dS pd dV dSV V S
ρ ρ ρ τu u A - S g A2
S S
= + −∫∫ ∫∫ ∫
( )- D AS
d mV
dS S dV+ ∫∫
momentumflux
pressuregradient
gravity viscousstresses
drag frominternal surfaces
& flow restrictions
momentumsources
Mass conservation
Total mass:
∂∂
ρ ρ ρtdV dS S dV
V V
= uAS∫∫ ∫+
Component α :
( )∂ ρ ρα α α ρ α∂tdV dS dS S dV
V VuA - J A
S S
= +∫∫ ∫∫ ,
convection diffusion chem. reactiontwo-phase change
convection sources (inflow, droplet depletion)
Internal energy conservation
( )edV e dS p d pV
Vt
dVV V
ρ ρ ∂∂
u A - uA SS S
= − ⎛⎝⎜
⎞⎠⎟∫∫ ∫∫
( )- qAS
dS S dVeV
+ ∫∫ e x e= ∑ α αα
pV work due to phase change
energy flux energy sources(thermal,conductivity,........)
(combustion, phase change,heat transfer)
∂∂t
HO2
convection pV work
J.R. Travis et al, Report FZKA-5994 (1998)
xx
R6
R5
Ring Room R9
He- Source
10 Vol % HeIsosurface
R6
R5
Ring Room R9
He- Source
10 Vol % HeIsosurface
Vertical He- jet into air (42 m/s) for 200 sVertical He- jet into air (42 m/s) for 200 s
0
5
10
15
20
0 200 400 600 800Time (s)
9KP000K76FZK/GASFLOW
0
5
10
15
20
0 200 400 600 800Time (s)
9KP000K76FZK/GASFLOW
Vertical distance from injection location (m)
0
40
80
0 2 4 6
FZK/GASFLOW
6kp270M18
Vertical distance from injection location (m)
0
40
80
0 2 4 6
FZK/GASFLOW
6kp270M18
0
5
10
15
0 200 400 600 800Time (s)
6KP270M18FZK/GASFLOW
0
5
10
15
0 200 400 600 800Time (s)
6KP270M18FZK/GASFLOW
Distance from jet axis (m)-2 -1 10 2
0
40
80
x
x
xx
x
xx x
EXPERIMENTFZK/GASFLOWx
Distance from jet axis (m)-2 -1 10 2
0
40
80
x
x
xx
x
xx x
EXPERIMENTFZK/GASFLOWx
13KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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
14KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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
GASFLOWExperiment
upper inj35 g/s
lower inj 135 g/s
lower inj 25 g/s
Inj. points
P. Royl, IKET
15KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Analysis of a hydrogen releasein a private garage
Analysis tools and related physics
Gas transport and mixing
Sequence of analysis steps
Mixture generation
16KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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
17KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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
18KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
RESULTING HYDROGEN CLOUD IN GARAGE
• Computed dimension of combustibleH2-air cloud in garage (4…75% H2)
• Characteristic size of combustible cloudexpressed as dCC = (Vcc)1/3
• Combustible cloud size strongly dependenton release rate, is result of balance betweensource strength and sinks, or releaserate and mixing mechanisms
Cas
e2
0.34
g H
2/s fo
r 100
sC
ase
13.
4 g
H2/s
for 1
0 s
d (cm)cc
0
10
20
30
40
50
60
70
80
90
0 50 100 150 200 250
stable
transient
0
20
40
60
80
100
120
140
160
180
0 10 20 30 40 50 60
d (cm)cc
Time (s)
Time (s)
19KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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?
20KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Analysis of a hydrogen releasein a private garage
Analysis tools and related physics
Combustion regimes
Transport and mixingMixture generation
Sequence of analysis steps
21KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
COMBUSTION REGIMES
• H2 – air mixtures can burn in different modes / combustion regimes
Inert, no stableflame propagationvfl = 0
Laminar deflagrationvfl ≈ 1 m/s, Ma << 1
Fast turbulent deflagrationvfl ≈ 300 m/s, Ma ≈ 1
Detonationvfl ≈ 1500 m/s, Ma >> 1
• Change of mode possible by transition process
Ignition Flame acceleration Deflagration-to-detonation transition (DDT)
2222
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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 1000flame speed (m/s)
overpressureratio (p-p)/p
max00
FZK-tube (closed) RUT ventedFLAME ventedplanar model,12%Hplanar model,28%H
2
2
Maximumacceptablestatic loadfor typical innercontainmentstructures(1 ton / m2)
(pm
ax-p
0)/p
0
p0
2323
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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:90
5
10
15
20
25
30
35
40
45
Frac
tion
of ig
nitio
n so
urce
[%]
Ignition sources
G:1 open fireG:2 mechanical sparkG:3 electrical sparkG:4 hot surfaceG:5 static discharge
G:6 catalytic surfaceG:7 self-ignitionG:8 othersG:9 unknownG:0 no ignition
Survey of 287 accidents with hydrogen
Kreiser et al, Report Univ. Stuttgart IKE 2-116 (1994)
2424
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
FLAME ACCELERATION
• Conservative conditions for flame acceleration in hydrogen mixtures were investigatedin closed obstructed tubes, e.g. FZK 12m-tube
25KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
RESULTS OF FLAME ACCELERATION EXPERIMENTS
D
flame velocity (m/s)
0 10 20 30 40 50 60 70x/D
0
200
400
600
800
1000
1200
v,m
/s
BR=0.6 H2-Air174mm
10%H211%H215%H2
520mm10%H211%H2
80mm10%H211%H213%H2
350mm9%H211%H217.5%H2
• Lean hydrogen mixtures in obstructedtubes with different tube diameters Dand 60% blockage ratio (BR)
• Two distinct regimes with slow and fast flame propagation are observed
H
0 5 10 15 20 25 30 35x/D
0
200
400
600
800
1000
1200
v, m
/s
350 mmBR=0.6
H2-air-CO2φ = 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
2626
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
FLAME ACCELERATION CRITERION
• Summary of experiments with different H2-O2- dilutend (N2, Ar, He) mixturesin obstructed tubes of differentscales
• Each point represents oneexperiment
• Results of data evaluation: expansion ratio σ is mixtureproperty which governsflame acceleration limit
• No flame acceleration forσ < 3.75 ± 0.1
(10.5% H2 in dry air)
fast flames
unstable flamesslow 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 LFZK
FZK-tube: 350 mmKI
Driver tube: 174 mmTorpedo tube : 520 mm
RUT : 2400 mmFZK / KI
δ ~ 0.1 mm~
30.1 1.0 10.0
4
5
6
7global quenchingquenching-reignitionchoked flamesquasi-detonation
T 300 K~~
Expa
nsio
n ra
tio σ
(= ρ
ub/ρ
b)
fast flames
unstable flamesslow 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 LFZK
FZK-tube: 350 mmKI
Driver tube: 174 mmTorpedo tube : 520 mm
RUT : 2400 mmFZK / KI
δ ~ 0.1 mm~
30.1 1.0 10.0
4
5
6
7global quenchingquenching-reignitionchoked flamesquasi-detonation
T 300 K~~
Expa
nsio
n ra
tio σ
(= ρ
ub/ρ
b)D
D
In lecture notes
27KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Fully obstructed tube (prototypic mode B)
accelerating flame DDT stable (quasi)detonation
spark ignition
pH / air2
burned gas
obstacle 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 / air2
high pressuresection
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 beenobserved
- shock focussing- detonation on-set in turbulent
flame brush
• Present here one example for DDTwith pressure wave emitted froman obstructed region and focussedin a conus
28KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
TURBULENT DEFLAGRATION EXPERIMENT WITHOUT DDT• Partially obstructed tube with conus, 15 % hydrogen in air
5 m
350 mm
29KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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 triggera detonation on other parts of the system
3030
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
CRITERION FOR DDT
• Correlation of all experimental data with given definitions of D and detonation cell size datashows that detonations areonly possible for D/λ > 7
• Current uncertainty in detonationcell size λ ≈ factor 2
2 3 5 2 3 5100 1000 10000Geometrical 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
s1s1
s1
s1
s1
s1
s1s1
s1
s1
s1
s1
t3
t3
t3
t3a1
a1a1a1
a1
a1
a1
a1a1a1a1
a1a1
a1
a1a1
a1a1a1
f1
f3
r2
r2
r5
r5
r5
r5r5
r5r5
r5
r5
r5
r5
r5
r5
r5r5
r5r5
r6
r6
v2
b1b1
b1
b1
b1
b2
b2
b2b2
b2
b3b3
b3
b3
b3
b3
b4b4b4b4b4
b4
b4
b5b5
b5
b5
b5
b6b6
b6b6b6
b6
m3m3 m4m4m4 m5
c3
d3
d3d3
d3
d3
d3d3d3d3d3d3
d3
d3 g1
g1
g1
g1
g1
g1g2
g2g6
g6
g6
g6
g7
g7
g7
g7
g8
g8
g8
g9
g9r1s2
s2
s2s2s2
s2s2
s2
t1
t1
r3r4r4r4r4
r4
r4
r4r4
r4
r7
r7
v1
d4
riri
ri
ri ririri
ri
d1
d1
t4
d2d2g3
s1a1a1
a1a1
a1
a1
a1a1a1
a1
a1
a1a1a1a1
f1
f3
r2r2
r5r5
r6v2
b1b1
b2
b2
b3b3
b4
b5b5b6b6
m3
m3m3
m4
m4m4 m5
m1 m2m6
m6m6m7m7m7 m8m8m8
g1g1
g1
g7
g7
g8
g8
r1 r3r4
r4
r4r4
r4
r4
r4
r4
r7r7v1
ri
riri
ri
ri
riD/
λ
Characteristic geometrical size of reacting mixture D (mm)
keinkein
Fehlergrenze
• Experiments on DDT in differently sized and shaped facilities have shown that a certainminimum scale is required for DDT
• In accident scenarios D/λ can varyby orders of magnitude, criterionhas therefore predictive capability
3131
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
DETONATION CELL SIZES
• Detonation cell sizes (in cm) of H2-air-steam mixtures at 375 Kand 1 bar initial pressure.
Dry hydrogen concentrationis defined as H2 / (H2 + air)
H (trocken) (vol%)
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
3232
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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%0
4%
Luft
H 2
8%
x = 16%0
4%
air
σindex =
< 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 transitionFlame acceleration
3333
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Analysis of a hydrogen releasein a private garage
Analysis tools and related physics
Combustion regimes
Distribution and mixingMixture generation
Sequence of analysis steps
Hazard potential
34KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
COMPUTED HAZARD PARAMETERS FOR GARAGE SCENARIOSC
ASE
20.
34 g
H2/s
for1
00s
• 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
0
20
40
60
80
100
120
140
160
180
0 10 20 30 40 50 60
d (cm)cc
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0 10 20 30 40 50 60
V (m³)fa
0
0.5
1
1.5
2
2.5
3
0 10 20 30
DDT
no DDT
40 50 60
D7λ
CA
SE 1
3.4
g H
2/s fo
r10
s
no DDT
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0 50 100 150 200 250
D7λV (m³)fa
0
0.02
0.04
0.06
0.08
0.10
0.12
0 50 100 150 200 250
d (cm)cc
0
10
20
30
40
50
60
70
80
90
0 50 100 150 200 250
Time (s)
Time (s)Time (s)Time (s)
Time (s)Time (s)
35KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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 mitigationmeasures
• only small potential for slow deflagrations, naturalmixing processes sufficient
• release rate (and mass) seems tolerable forpresent garage design
- Case 2:(0.34 g H2/s)
36KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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 inthe garage?
• Outcome uncertain, experiments performed in test chamber simulating thegarage
37KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Analysis of a hydrogen releasein a private garage
Analysis tools and related physics
Combustion regimes
Distribution and mixingMixture generation
Sequence of analysis steps
Combustion (Experiments)
Hazard potential
3838
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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 g0.5 g
Hydrogen injection device cubes covered with plastic,filled with stoichiometric H2 - air mixture
39KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
0R
Pressure transducers
mH2pmax
i v ∫>Δ
+ Δ=0
)(p
dttpI
UNCONFINED TEST OF COMBUSTION UNIT
• Peak overpressure and impulsemeasured as function of distanceto characterize blast effects fromcombustion unit
4040
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
FLAME SPEEDS IN COMBUSTION UNITS
0.5 1 2 4 8 160
200
400
600
800
1000
1200
1400
1600
1800
2000
0 0.1 0.2 0.3 0.4 0.5
R, m
Flam
e sp
eed,
m/s
8 g
8 g16 g
16 g4 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
4141
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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+
0
1
2
3
4
5
0 5 10Distance R from center of cube, m
Peak
over
pres
sure
Δp+
, bar 2g
2g1g
1g
0.5g
0
2
4
6
8
10
12
0 5 10
16g
16g
8g8g
4g
Distance R from center of cube, m
Peak
over
pres
sure
Δp+
, bar
4242
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
IMPULSE VS DISTANCE
Δp+
t
I+
• Measured positive impulse I+ values from unconfinedcombustion units
0
20
40
60
80
100
120
140
160
180
0 5 10
Distance R, m
Impu
lse,
Pa*
s
16g
16g
8g8g
4g
0102030405060708090
100
0 5 10
Distance R, m
Impu
lse,
Pa*
s
2g
2g1g
1g
0.5g
4343
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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
0 2 4 6
Scaled distance R (p0 / E)1/3
Scal
edpe
akov
erpr
essu
reΔ
P+/p
0
16g
16g
8g8g
4g
Gas detonationTNT (= energy)
0.01
0.1
1
10
0 5 10
2g2g
1g
1g0.5g
Gas detonation
TNT (= energy)
Scal
edpe
akov
erpr
essu
reΔ
P+/p
0
Scaled distance R (p0 / E)1/3
4444
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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 chamberpossible
• All ventilation systems explosion protected
• Test cell used for simulation of garage /confined volume
Ventilation ducts
Ventilation turbines
Test chamber(garage)
Lower compartment
45KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
0100
200300
400500
0
50
100
150
200
250
300
0
200400
6008000
100200
300400
500
0
50
100
150
200
250
300
0
200400
600800
DruckaufnehmerBeschleunigungs-sensoren
3A
3B
1B
13A
7B
2A
1A
4A5A
6A 7A
8A15A
14A
2B
4B
5B
6B8B
16A
Höhe [cm
]
Breite [cm] Länge [
cm]
TürZünd-ort
INSTRUMENTATION OF GARAGE
• The instrumentation included pressure and acceleration sensors at different locations,covering flat surfaces, (2d) edges and (3d) corners
Pressure sensorsAccelerationsensors
location of combustion unit
Height (cm)
Width (cm)Length (cm)
46KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
LOCAL HYDROGEN EXPLOSIONS IN A GARAGE
Experimentwith 8g H2
H2 - mass:
- 1g- 2g- 4g- 8g- 16g
47KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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 H2Experiment2Experiment3
Vol
t
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 H2Experiment2Experiment3
Vol
t
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 Experiment2Experiment3
Vol
t
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 Experiment2Experiment3
Vol
t
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
Experiment2Experiment3
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
Experiment2Experiment3
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 Experiment2Experiment3
Übe
rdru
ck [b
ar]
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
Experiment1Experiment2Experiment3
Übe
rdru
ck [b
ar]
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
0,06Druckaufnehmer 5A1g H2
Experiment1Experiment2Experiment3
Übe
rdru
ck [b
ar]
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 Experiment2Experiment3
Übe
rdru
ck [b
ar]
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 Experiment2Experiment3
Übe
rdru
ck [b
ar]
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
Experiment1Experiment2Experiment3
Übe
rdru
ck [b
ar]
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
0,06Druckaufnehmer 5A1g H2
Experiment1Experiment2Experiment3
Übe
rdru
ck [b
ar]
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
0,06Druckaufnehmer 5A1g H2
Experiment1Experiment2Experiment3
Übe
rdru
ck [b
ar]
Zeit [s]
REPRODUCIBILITY OF MEASURED DATA
• The experiment with 1 g H2was performed three times
• Acceleration and pressuresensors show very good reproducibility of measuredsignals
• Complex, but reproduciblepressure waves are createdin confined local explosionsof H2-air mixtures
2
48KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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 1g2g4g8g16g
Übe
rdru
ck[b
ar]
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 1g2g4g8g16g
Übe
rdru
ck[b
ar]
Zeit [s]
0,193 0,194 0,195 0,196 0,197
-0,1
0,0
0,1
0,2
0,3
0,41g2g4g8g16g
Übe
rdru
ck [
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.
Ove
rpre
ssur
e[b
ar]
Ove
rpre
ssur
e[b
ar]
Time [s] Time [s]
49KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Analysis of a hydrogen releasein a private garage
Analysis tools and related physics
Combustion regimes
Turbulent deflagration
Distribution and mixingMixture generation
Sequence of analysis steps
Combustion
Hazard potential
5050
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Re =t u'L t
ν
Turb. integral length scale Lt / Laminar flame thickness δ
Evolution inobstructedlarge scalecombustion
Turb
ulen
cein
tens
ityu‘
/ La
min
arfla
me
velo
city
SL Da < 1
Da = 1
Da > 1, Ka >1
Ka = 1
Ka < 1
Ret < 1
Da = Turb. Transport time (makro)
Laminar reaction timeLt / u'
δL/ SL=
Ka = Laminar reaction time Turbulent transport time (Kolmogorov scale)
δL / SLlΚ / u 'K
=
Flameshapes
PDFs
laminarflame
foldedflame
wrinkledflame
homogeneousreaction
Unburned gas
Burned gas
Reaction zoneKa < 1 Da > 1 Da = 1 Da << 1
thickened flame
TURBULENT DEFLAGRATION REGIMES
5151
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
COM3D EQUATIONS
A. Kotchourko, IKET
• COM3D under development at FZK for simulation of turbulent deflagration
5252
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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
5353
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
A. Kotchourko, IKET
COM3D VERIFICATION (2)
Distance (m)
Overpressure(bar)
0,2 0,3 0,4 0,5 0,6
Time (s)
0
5
10
15
20
25
30
35
Experiment RUT13H=11%BR=0,3p=1barT=283K
2
0
0
COM-Code c=7 f
Experiment
0
2
4
6
8
10
12
Distance (m)
Overpressure(bar)
0
2
46
8
10
12
14
16
100 150 200 250 300
Time (ms)
0
5
10
15
20
25
30
35
Experiment RUT21H=12,5%BR=0.6p=1barT=283K
2
0
0
COM-Code c=7 f
Experiment
Distance (m)
Overpressure(bar)
Distance (m)
Overpressure(bar)
0
2
4
6
8
10
150 200 250 300 350
Time (ms)
0
5
10
15
20
25
30
35
Experiment RUT23H=11,2%BR=0.6p=1barT=283K
2
0
0
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
Dis
tanc
e (m
)
Dis
tanc
e (m
)
Dis
tanc
e (m
)
5454
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Analysis of a hydrogen releasein a private garage
Analysis tools and related physics
Combustion regimes
Turbulent deflagration
Distribution and mixingMixture generation
Sequence of analysis steps
Combustion simulation
Hazard potential
55KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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
57KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
COM3D COMBUSTION SIMULATION
• 3d pressure field, calculated isosurface for 1.1 bar
• Test with 8g H2
58KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
COMPARISON OF OVERPRESSURES
• Good agreement, remaining differences are due to geometry simplification and rigid wall modelin simulation
over
pres
sure
[bar
]
over
pres
sure
[bar
]ov
erpr
essu
re[b
ar]
over
pres
sure
[bar
]
Time [s]
Time [s] Time [s]
Time [s]
59KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Analysis of a hydrogen releasein a private garage
Analysis tools and related physics
Combustion regimes
Turbulent deflagration
(Detonation)
Structural response
Distribution and mixingMixture generation
Sequence of analysis steps
Combustion simulation
Hazard potential
60KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
STRUCTURAL RESPONSE
• What are effects of blast loads on the structure?
6161
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
SINGLE-DEGREE-OSCILLATOR MODEL FOR STRUCTURAL RESPONSE
• Simplest model for structural response is SDO model
• Describes ground mode (first harmonic) of 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)
mt
D
6262
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
BLAST LOADED ELASTIC OSCILLATOR (1)
0 0) (t x 00) x(twith
ep kx xm
ep kx -
F xm
load
load
Tt
Tt
i i
====
Δ=+
Δ+=
=
−+
−+
∑
&
&&
&&
( )( ) ⎥
⎦
⎤⎢⎣
⎡+ω−
ω+
=Δ
−+
loadt/T
loadload
load etcosωT
t sinωT
ωT/kp
x(t)2
2
1
/kp x0x
max+Δ=
=&&
( )osc
1/2
T2πperiod oscillatork/mω ===
• 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)
)Tf( /kp
xload
max ω=Δ +
k
m t
x
loadT/tep)t(p −+Δ=
t
p(t)
Δp+
. scal
edde
form
atio
nx m
axscaled loading time T
mkT =ω
T = 1 ms 5 ms
scal
edde
form
atio
nx m
axscaled loading time T
mkT =ω
/ (P
* / k
)
6363
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
BLAST LOADED ELASTIC OSCILLATOR (2)
• Asymptotes for maximum deflection /deformationcan be computed from energy balances
• Quasistatic loading realm (T load >>Tosc)- strain energy = work on structure
½ kx2max = Δp+ · xmax
2k/p
xmax =Δ +
• Impulsive loading realm ( T load << Tosc)- initial kinetic energy = strain energy
2max
2
2max
20
kx21
2mI
kx21mv
21
=
=
loadload/max T T)
mk(
k/px
ω==Δ +
21or I)
km1(x 1/2max ⋅=
maximum deformation is proportional to blast wave impulse I
Impulsiveasymptote
Quasistaticasymptote
Scaled loading time ωT
loadT/t Tp epI load ⋅Δ=Δ= +∞ −+∫0
dynamic maximum deflection is two times staticdeflection (DLF = 2) S
cale
ddi
spla
cem
ent
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
6464
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
max load
loadmax
21loadmax
x~ I T pT
1 x(km) p
T1
kxp
=Δ
=Δ
ω=
Δ
+
+
+
OSCILLATOR RESPONSE: ANOTHER VIEW
• Often oscillator response is presented with inverted ordinate and unscaled load parametersΔp+ and Tload
• Quasistatic asymptote
Maximum deflagration xmax is only proportional to applied peak overpressure Δp+, independent of loadduration• Impulsive asymptote
Maximum deflection xmax is proportionalto applied impulse
positive peakoverpressure
logΔp+
log of positive overpressure duration Tload
I2I1
xmax1
xmax2
xmax2
xmax1
Δp+2
Δp+1
2kx p
21
kxp
max
max
=Δ
=Δ
+
+
6565
KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
Analysis of a hydrogen releasein a private garage
Analysis tools and related physics
Combustion regimes
Turbulent deflagration
(Detonation)
Structural response
Distribution and mixingMixture generation
Sequence of analysis steps
Combustion simulation
Hazard potential
Consequences
68KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
STRUCTURAL DAMAGE FROM CONFINED LOCAL H2 EXPLOSIONS IN GARAGE
10000Duration of positive overpressure T [ms]
Pos
itive
pea
kov
erpr
essu
reΔ p
+[P
a]
103
104
105
106
Δp+I+
t
T+=2I+/Δp+
T+
Unconfined gaseous detonations
1,4m
1,4m
2,4m
2,4m
2,8m
5.6m
4,0m
6,3m4,0m
FZK Exp. 2g H2
FZK Exp. 16g H2
0.1 1 10 100 1000
Partial demolition, 50%- 75% of walls destroyed
Major structural damage, some wrenchedload bearing members fails
Minor structural damage, wrenchedjoints and partitions
2g H2
1 m
glass breakage
Area 1m2
4.7 mm thick
Area 3m2
4.7 mm thick
16g H2
2 m
5 m
450 Pas
300 Pas110 Pas
10000[ms]
Δ+
[Pa]
103
104
105
106
Δp+I+
t
T+=2I+/Δp+
T+
Unconfined gaseous detonations
1,4m
1,4m
2,4m
2,4m
2,8m
5.6m
4,0m
6,3m4,0m
FZK Exp. 2g H2
FZK Exp. 16g H2
0.1 1 10 100 1000
Partial demolition, 50%- 75% of walls destroyed
Major structural damage, some wrenchedload bearing members fails
Minor structural damage, wrenchedjoints and partitions
2g H2
1 m
glass breakage
Area 1m2
4.7 mm thick
Area 3m2
4.7 mm thick
16g H2
2 m
5 m
10000[ms]
Δ+
103
104
105
106
Δp+I+
t
T+=2I+/Δp+
T+
Unconfined gaseous detonations
1,4m
1,4m
2,4m
2,4m
2,8m
5.6m
4,0m
6,3m4,0m
FZK Exp. 2g H2
FZK Exp. 16g H2
0.1 1 10 100 1000
Partial demolition, 50%- 75% of walls destroyed
Major structural damage, some wrenchedload bearing members fails
Minor structural damage, wrenchedjoints and partitions
2g H2
1 m
glass breakage
Area 1m2
4.7 mm thick
Area 3m2
4.7 mm thick
16g H2
2 m
10000+ [ms]
Δ+
103
104
105
106
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
6,3m4,0m
FZK Exp. 2g H2
FZK Exp. 16g H2
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 wrenchedload bearing members fails
Minor structural damage, wrenchedjoints and partitions
2g H2
1 m
glass breakage
Area 1m2
4.7 mm thick
Area 3m2
4.7 mm thick
16g H2
2 m
5 m5 m
450 Pas
300 Pas110 PasResults: - 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
71KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
0.1 1 10 100 1000 10000
NASA
16g H2
Pos
itive
pea
k ov
erpr
essu
re Δ
p+[P
a]
103
104
105
106
lung rupture thresholdΔp+
I+
t
T+=2I+/Δp+
T+
Unconfined gaseous detonations
1,4m
1,4m
2,4m
2,4m
2,8m
5.6m
4,0m
6,3m4,0m
FZK Exp. 2g H2
FZK Exp. 16g H2
increasinggrade ofinjury
increasinggrade ofinjury
50 % eardrum rupture
1 % eardrum rupture
Baker
Duration of positive overpressure T+ [ms]
1 m
2 m
2g H25 m
0.1 1 10 100 1000 10000
NASA
16g H2
Pos
itive
pea
k ov
erpr
essu
re Δ
p+[P
a]
103
104
105
106
lung rupture thresholdΔp+
I+
t
T+=2I+/Δp+
T+
Unconfined gaseous detonations
1,4m
1,4m
2,4m
2,4m
2,8m
5.6m
4,0m
6,3m4,0m
FZK Exp. 2g H2
FZK Exp. 16g H2
FZK Exp. 2g H2
FZK Exp. 16g H2
increasinggrade ofinjury
increasinggrade ofinjury
50 % eardrum rupture
1 % eardrum rupture
Baker
Duration of positive overpressure T+ [ms]
1 m1 m
2 m2 m
2g H25 m5 m
HUMAN EFFECTS FROM CONFINED LOCAL H2 – EXPLOSIONS IN GARAGE
Results: - high probability of ear drum rupture- no long damage
72KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
SUMMARY OF MECHANISTIC SAFETY ANALYSIS OF
HYDROGEN BASED ENERGY SYSTEMS
73KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund Universität Karlsruhe (TH)
PROCEDURE FOR HYDROGEN SAFETY ANALYSIS
• A complete, self-consistent and mechanistic analysis procedure has been developed whichaddresses all important physical phenomena of hydrogen behaviour in accidental release scenarios
CONSEQUENCEANALYSIS
Mechanical andthermal loads
Structural
Human effects
CRITERIA FORHAZARD
POTENTIAL
Flammability
y
y
Flame Acceleration
y
COMBUSTIBLEMIXTURE
GENERATION
GASFLOW
FLAME3D
Fast turbulentdeflagration
COM3D
DetonationDET3D
Slow deflagration
CONSEQUENCEANALYSIS
responseSDO
ABAQUS
Detonationtransition
COMBUSTIBLEMIXTURE
GENERATION
Problem geometry
Mitigation
Scenario
Sources
DistributionGP - Program
COMBUSTIONSIMULATION
74KIT – die Kooperation vonForschungszentrum Karlsruhe GmbHund 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)
Accidentprogression
If one level of defencehas beenoptimized, workon next barrierfor accidentprogression