1 Hydrogen As a Fuel – Different Properties Require

Home , Hydrogen safety

Hydrogen as a fuel – different properties require different optimization for safety

Olav Roald Hansen, LR Bergen

FLACS users group May 3, 2017

Working together for a safer world

Hydrogen as a fuel – Safety Aspects

Outline

• Hydrogen – fun facts, uses in the past and in future

• Lloyd’s Register and ORH experience with hydrogen

• Hydrogen incidents & accidents

• Special properties hydrogen

• Hydrogen consequence modelling vs experiments

• Example homework exercises LR training

• Summary– key aspects hydrogen safety

Material from internal training sessions at LR (4 x 60 min with modelling assignments) Part 1 - Intro & what makes hydrogen different wrt safety

Part 2 - Consequence modelling of hydrogen safety (FLACS), H2 specific guidelines, worksheet Part 3 - Advanced modelling topics (DDT,detonation, vessel rupture) Part 4 - Challenging modelling topics (how to apply knowledge to real studies…)

©Lloyd’s Register

1 Hydrogen fun facts

• Hydrogen – most common element in universe ~75% by mass (90% by mole)

• 10% of human body weight is hydrogen molecules (65-70% of atoms are H)

• Earth is heated by solar fusion reaction (~2 kg/s)

• Density solid (SG 0.086 <14 K), liquid (SG 0.09 < 20K) and gas (0.09 kg/m3)

• Forms both positive and negative ions, only element without neutron etc.

Solar fusion (Wikipedia)

©Lloyd’s Register

Hydrogen traditional use

• Used in air balloons and airships until 1937

• Key component in synthesis gas

• Steam reforming (CH4+H2O=> CO+3H2) • Ammonia production

• Hydrocarbon cracking refineries

• Methanol production

• HCl production

• Rocket propellant (liquid hydrogen and oxygen)

©Lloyd’s Register

2 Hydrogen in the future – why interesting

Hydrogen exhaust is pure water vapour

 great potential reducing local pollution from transportation

Primarily used in fuel cells

 Electricity is output – easy to combine with batteries

Distributed electrolyser production one option

 Off-peak renewable electricity (wind, solar, hydro) converted to hydrogen

 By-product oxygen required e.g. for fish farming …

©Lloyd’s Register

Hydrogen in the future – will it lift off the ground this time?

“New Hydrogen” age from 2004

 Utsira Hydrogen Society 2004-2006

 Hydrogen highway 2009-2012 – 7 filling stations Oslo to Stavanger

Similar trends elsewhere – USA, Japan, Canada, Europe

 Fork lifts, trains, buses/cars

Mainly R&D and subsidy dominated, challenges

• Efficient storage

• Scaling up

• Price

• Safety (?)

2008 => less dedication from many countries 2012 – Statoil changed focus from hydrogen to offshore wind … Today: Political dedication being GREEN is stronger, thus hope … ©Lloyd’s Register

3 Hydrogen – strong heritage in Norway

Hydrogen factories built to produce ammonia with energy from hydropower

1940-Rjukan largest hydrogen factory in the world

 Norsk Hydro electrolysers – now NEL-hydrogen

 Norsk Hydro ammonia/fertilizer – now Yara

 Hexagon also major player within composite storage

©Lloyd’s Register

Lloyd’s Register and hydrogen safety

LR has performed various H2 risk assessments: • Hydrogen-highway – several studies 2006-2012

• HyOP Høvik/Gardermoen, Uno-X Kjørbo/Hvam/Åsane/Søreide (2015-2017)

• Safety assessments hydrogen production unit (2016)

• ASKO Trondheim – H2 from solar to trucks & forklifts • Partner MoZEES R&D (Mobility Zero Emission Energy Systems)

Hydrogen expertise also important for

• Ammonia plants

• Refineries

• Nuclear industry

Colleagues with long experience from gas industry Close connection to LR Marine & shipping initiatives

©Lloyd’s Register

4 Hydrogen experience Olav RH

24 years experience dispersion/explosion/CFD-modelling FLACS

 19y CMR/GexCon, 2y GL Noble Denton and 3y LR

 1998- focus to make FLACS a hydrogen safety software

 2001-2004 R&D test project – dispersion and explosion tests

 2001-2008 – Responsible R&D and global sale of FLACS CFD

 2004-2010 EU-project HySAFE (25 partners in Europe)

 2004-2012 IEA Expert Groups Hydrogen Safety

 Sold FLACS & trained e.g. Sandia NL (DoE), SRI, HRI, IRSN

 Various consulting work hydrogen safety

 15+ scientific articles within hydrogen safety

©Lloyd’s Register

Incidents and accidents hydrogen

• Hydrogen bomb – not relevant - Analogue - “NIMBY” & LNG-tanker vs 40 atombombs

• Chernobyl (1986) & Fukushima (2013) – some relevance

- Nuclear industry: challenges with waste & pressurized hydrogen

• Hindenburg (1937) – confirmed that hydrogen burns …

• Challenger (1986) – liquid H2 & O2 very reactive

©Lloyd’s Register

5 Incidents and accidents hydrogen

• Herøya-ammonia plant N1 (1985) –10-20 kg, building destroyed, broken windows 700m away • Stockholm Brahegatan (1983) –gas bottles on truck, broken windows 90m away • Gas generation batteries -potential for hydrogen+oxygen explosions? • Slow hydrogen generation in confined spaces - bio-facilities, metal waste in water, nuclear waste storage

Numerous smaller incidents: See h2tools.org/lessons

(risk for underreporting)

©Lloyd’s Register

Hydrogen properties - storage DoE

Efficient hydrogen storage challenging:

• Extreme pressures or temperatures

• Costly to pressurize to 700 bar or cool to -253˚C

Energy density

- Low by volume, high by mass

Similar challenges with alternative storage (metal hydrides/organic liquids)

• H2 small molecule, may diffuse through materials • Significant boil-off from liquid hydrogen

• Hydrogen cracking – special steel qualities required

©Lloyd’s Register

6 Hydrogen properties- leaks

In general a very high safety level when introducing hydrogen – CCTVs, detection ++

• Extreme pressures & temperatures challenging

• Hydrogen cracking

Will hydrogen cars and systems be allowed to age?

Costly and challenging to maintain safety level?

©Lloyd’s Register

Hydrogen properties- reactivity

DoE

Property Hydrogen Methane (NG) Flammability in air 4 to 75% 5 to 15% Burning velocity ~3 m/s 0.45 m/s Detonation energy 1 g TNT 1 kg TNT

2 Explosion pressure P ~ SL i.e. Hydrogen < 10% - limited reactivity Hydrogen >15% - reactivity worse than NG ©Lloyd’s Register

7 Hydrogen properties– releases (1)

Catastrophic rupture 350-700+ bar

• Lots of energy - may limit risk by location and choice of tanks

LR screening tool for tank explosjon =>

Sonic release – speed of sound 1270 m/s (2.8 x methane)

 Rapid dilution in air for free jet

Outdoor installations proper separation distance to walls/objects may limit gas accumulation > 15% H2 in air ©Lloyd’s Register

Hydrogen properties– releases (2) DoE

INERIS

H2 smallest molecule– 0.07 x density of air . Important safety principle – hydrogen shall disappear upwards

Exception: Liquid hydrogen (~20 K)

. Costly to cool (1/3 of combustion energy)

. Vapours denser than air, in particular pressurized releases

. Challenge: O2 and N2 “snow” due to condensaton/freezing of air HSL . O2-doped air may give strong increase of gas reactivity

Photos from: Ichard, M., Hansen, O.R., Middha, P. & Willoughby, D. (2012). CFD computations of liquid hydrogen releases. International Journal of Hydrogen Energy, 37: 17380-17389

HSL ©Lloyd’s Register

8 Hydrogen properties– ignition

DoE

Minimum ignition energy 0.019 mJ (methane 0.29 mJ)

• Static electricity sparks ~1 mJ

Ignition often hard to explain

. Atmospheric releases (low momentum) - snow or thunder

. Charging of equipment - static electricity and discharges

. Sparks from dust/particles in a jet

. Shockwaves heating pockets of gas/air above AIT (Inverse Joule-Thomson effect should not cause ignition)

Jet fire:

• Extremely noisy

• Often invisible

• Hot flame, limited radiation

1 kg/s SRI-release (DoE-Sandia) Practically always ignition ©Lloyd’s Register DoE

Hydrogen consequence modelling

Hydrogen very different

 Specialist knowledge and quality calculations required KIT tests FLACS –well validated with guidelines

 Good results in blind tests, gas dispersion and explosion

INERIS Blind test Frequently used CFD-tools Fire models

 “Colourful Fluid Dynamics”

FLUENT CFX

©Lloyd’s Register

9 Fraunhofer ICT Hydrogen large scale experiments

Fraunhofer ICT

©Lloyd’s Register Sandia NL

SRI (Sandia hydrogen jet-experiments) Hydrogen DoE/Sandia & NFPA-2

. Sandia(DoE) – leading US hydrogen safety competence

. Trained them using FLACS – showed how

Later: Various papers showing FLACS vs experiment

. Ignited jet – pressure in front of and behind wall

. Release and explosion in tunnel

. Release and explosion in garage

Sandia/DoE managers: “We understand hydrogen safety – our models work”

©Lloyd’s Register

10 Example slide from LR internal training: Hydrogen model validation & important guidelines - DISPERSION

Small orifice jets (< 1mm) overestimated by CFD

INERIS jet adjusted: High momentum Cd=0.788 Z=1.12 => rate=94.5% of ideal gas • INERIS jets (200 bar/0.5 mm) Adjusted pseudo-jet • FZK-jets (200 bar/0.25-1.00 mm) v=0.9c (967 m/s at 196 K) T=196 K (80K less than JET-utility) • HSL-jets (12-140 bar/3-12mm) Overprediction: • Sandia-impinging (ignited) Previously +100% Now ~ +35% (still conservative) • HSL/Shell impinging (ignited) Pseudo-source to be revisited?

SRI (Sandia hydrogen jet-experiments)

©Lloyd’s Register

Example slide from LR internal training: Hydrogen model validation & important guidelines – Explosions – premixed large scale tests

Partly confined (DDT cases in green)

• Sandia FLAME facility

• KI-KOPER facility

• Fh-ICT lane experiments

• FM Global - chamber

Unconfined

• Fh ICT 20m diameter balloon

• Shell/HSL pipe arrays & refueling station

• Sandia/SRI tests; balloon, pipe array etc.

Observations:

 Most well reproduced by FLACS

 Exception: FM Global vented chamber – flame instabilities important for empty, vented rooms – model and report with care

 Pre-ignition turbulence important (ref. previous slides) DDT to be kept in focus – handled in Session 3 ©Lloyd’s Register

11 Hydrogen safety screening tool (ORH)

Internal LR tool to estimate hazard distances (like NFPA-2)

• Jet flame radiation levels

• Hydrogen concentration distances

• Compressibility effect included

Further enhanced to include:

• Blast (pressure, impulse, duration) for tank rupture

• Explosive cloud sizes hydrogen jets

• Estimating effect of confining planes (ground)

3m Demo: 4m

Example: LR Consequence Modelling Assignment 1 - Explosion

Explode 20m diameter hemispherical stoichiometric hydrogen cloud, central ignition at ground Report transient pressures at 5m, 18m, 35m and 80m distance (1m above ground) + 3D video flame+PMAX

12 Example: LR Modelling Assignment 2 - Dispersion

1 g/s vertical hydrogen release for 240 seconds from a h=26cm and D=12cm release chamber at central room axis from orifice with D=20mm, model dispersion for 1 hour. Garage dimension 7.2m x 3.78m x 2.88m – two small openings D=5cm at h=7.5cm +/- 7.5 off room axis. Release chamber 3.8m into room on central axis (1.9m from either side wall)

Report transient concentrations at sensor 4, sensor 12 and sensor 16 Make video of development of flammable plume Brain-twister: Do the two small openings matter? What could happen if these were closed? What setting is required to model this in FLACS?

Homework-input from colleagues:

Example: LR Assignment 4a: HP impinging jet dispersion + ignition

4a Ignited free jet Horizontal release 1.2m above ground P=205 bara – orifice D=6.4mm (tube dia 9.5mm) ignition 2m from release point at 600 & 800 ms Estimate release rate, report length to 4%, 8%, 15% and Q9 and FLAM-volume at ignition/steady-state Report transient overpressure at sensor 1

HSL

13 Example: Assignment 5: DDT prediction

Explosion in chamber + lane (Fraunhofer ICT)

3m x 1.5m x 1.5 box, ignition at centre of inner end opposite there is a vent opening (0.5m x 0.5m centrally)

Outside the box there is a 12m x 3m x 3m gas cloud between two solid walls 12m x 3m Halfway down walls there is one tube on each side Hydrogen gas concentration 21%

Report Potential for DDT (DPDX-plots) + likely spot for DDT PMAX and FLAME PLOTS Transient pressures at P1 (middle of chamber), P3 (0.6m outside vent), P7 (3.6m outside) and P13 (8m outside) DDT onset well modelled by colleagues – one used LR DDT and detonation automatic functionality cs*CLOUD “spot disease” rediscovered by colleagues: [CV_RESOLUTION=16 to be used] ©Lloyd’s Register

Example: Assignment 7: Fire simulation

Hydrogen jet fire towards wall (Sandia/SRI) 138 bara, 3.175 mm orifice 1.37m from wall (pointing at centre of wall) 2.4m x 2.4m wall

Report radiation at R1, R2, R3 and R4

Repeat for free jet without wall => Report R2 and R3

Modelling by colleagues (KFX):

14 Summary: Important safety principles (Compressed H2 350-700 bar)

Avoid gas accumulation at concentrations > 10-15%

• Limit amount vs room volume OR design so that gas will quickly disappear by buoyancy

• Limit congestion – pressurized releases should dilute in air AND flame acceleration be limited

• DDT (detonation) must be avoided – challenge for design

• Ignition source control fine, but no safety barrier If DDT – flames faster than sound • Blast waves/flames not to threaten people or structural integrity  Vent opening not seen

E.g. ship design – risk based design required

• Storage tanks at well ventilated locations, protected against impact

Machine room (and all hydrogen segments in enclosures)

• Consider hydrogen inventory & maximum leak rate VERSUS compartment size & ventilation

• Limit congestion in all areas where hydrogen may accumulate

QUESTIONS or COMMENTS?

Olav Roald Hansen Senior Principal Consultant Bergen T +47 91 17 17 87 E [email protected]

Lloyd’s Register www.lr.org

Working together for a safer world

15 Refence list – scientific articles within hydrogen safety

• Ichard, M., Hansen, O.R., Middha, P. and Willoughby, D. (2012). CFD Computations of liquid hydrogen releases. Journal of Hydrogen Energy, 11/2012

• Ham, K., Marangon, A., Middha, P., Versloot, N., Rosmuller, N., Carcassi, M., Hansen, O.R., Schiavetti, M., Papanikolaou, E., Venetsanos, A., Engebø, A:, Saw, J. L., Saffers, J-B., Flores, A. & Serbanescu, D. (2011). Benchmark exercise on risk assessment methods applied to a virtual hydrogen refuelling station. International Journal of Hydrogen Energy, 36: 2666-2677.

• Middha, P., Engel, D. & Hansen, O.R. (2011). Can the addition of hydrogen to natural gas reduce explosion risk? Int. Journal of Hydrogen Energy, 36: 2628-2636.

• Venetsanos, A.G., Adams, P., Azkarate, I., Bengaouer, A., Brett, L., Carcassi, M.N., Engebø, A., Gallego, E., Gavrikov, A.I., Hansen, O.R., Hawksworth, S., Jordan, T., Kessler, A., Kumar, S., Molkov, V., Nilsen, S., Reinecke, E., Stöcklin, M., Schmidtchen, U., Teodorczyk, A., Tigreat, D. & Versloot, N.H.A. (2011). On the use of hydrogen in confined spaces: Results from the internal project InsHyde, International Journal of Hydrogen Energy, 36: 2693-2699.

• García, J., Baraldi, D.,Gallego, E., Beccantini, A., Crespo, A., Hansen, O.R., Høiset, S., Kotchourko, A., Makarov, D., Migoya, E., Molkov, V., Voort, M.M. & Yanez, J. (2010). An intercomparison exercise on the capabilities of CFD models to reproduce a large-scale hydrogen deflagration in open atmosphere. International Journal of Hydrogen Energy, 35: 4435-4444.

• Middha, P., Hansen, O.R., Grune, J. & Kotchourko, A. (2010). CFD calculations of gas leak dispersion and subsequent gas explosions: Validation against ignited impinging hydrogen jet experiments. Journal of Hazardous Materials, 179: 84-94.

• Venetsanos, A.G., Papanikolaou, E., Middha, P., Hansen, O.R., Garcia, J., Heitsch, M., Baraldi, D. & Adams, P. (2010). HySafe standard benchmark problem SBEP-V11: Predictions of hydrogen release and dispersion from a CGH2 bus in an underpass. International Journal of Hydrogen Energy, 35: 3857-3867.

• Makarov, D., Verbecke, F., Molkov, V., Roe, O., Skotenne, M., Kotchourko, A., Lelyakin, A., Yanez, J., Hansen, O.R, Middha, P., Ledin, S., Baraldi, D., Heitsch, M., Efimenko, A. & Gavrikov, A. (2009). An inter-comparison exercise on CFD model capabilities to predict a hydrogen explosion in a simulated vehicle refuelling environment. Int. Journal of Hydrogen Energy, 34 (6): 2800-2814.

• Makarov, D., Baraldi, D., Kotchourko, A., Lelyakin, A., Yanez, J., Middha, P., Hansen, O.R., Gavrikov, A., Efimenko, A., Verbecke, F. & Molkov, V. (2009). An inter-comparison exercise on CFD model capabilities to simulate hydrogen deflagrations in a tunnel. Int.Journal of Hydrogen Energy, 34 (18): 7862-7872.

• Middha, P. & Hansen, O.R. (2009). Using computational fluid dynamics as a tool for hydrogen safety studies. Journal of Loss Prev. in Process Industries, 22 (3): 295-302.

• Middha, P. & Hansen, O.R. (2009). CFD simulation study to investigate the risk from hydrogen vehicles in tunnels. Int. Journal of Hydrogen Energy, 34 (14): 5875-5886.

• Middha, P., Hansen, O.R. & Storvik, I.E. (2009). Validation of CFD-model for hydrogen dispersion. Journal of Loss Prev. in the Process Industries, 22: 1034-1038.

• Venetsanos, A.G., Papanikolaou, E., Delichatsios, M., Garcia, J., Hansen, O.R., Heitsch, M., Huser, A., Jahn, W., Jordan, T., Lacome, J-M., Ledin, H.S., Makarov, D., Middha, P., Studer, E., Tchouvelev, A.V., Teodorczyk, A., Verbecke, F. & van der Voort, M.M. (2009). An inter-comparison exercise on the capabilities of CFD models to predict the short and long term distribution and mixing of hydrogen in a garage. International Journal of Hydrogen Energy, 34 (14): 5912-5923.

• Middha, P. & Hansen, O.R. (2008). Predicting deflagration to detonation transition in hydrogen explosions. Process Safety Progress, 27 (3): 192-204.

• Hansen, O.R. & Middha, P. (2007). CFD-based risk assessment for hydrogen applications. Process Safety Progress, 27(1): 29-34.