D. Makarov, V. Molkov

Hydrogen Safety Engineering and Research Centre, University of Ulster

Core research on hydrogen safety

and EPSRC Challenge project

“Integrated safety strategies for

onboard hydrogen storage system”

SUPERGEN HUB Research and Advisory Board meeting

Newcastle University, 30-31 July 2014

Core research

Work Package 3: Hydrogen and Fuel Cell Safety

Challenges: Breakthrough safety strategies and engineering solutions to

underpin safety of emerging HFC technologies

Approach: Fundamental research to close knowledge gaps:

hydrogen concentration decay in highly under-expanded jets

behaviour of under-ventilated hydrogen fires (self-extinction and

reignition)

combustion instabilities (e.g. Rayleigh-Taylor)

Applied research directions: Safety of hydrogen buses and cars

Rely on CFD expertise of HySAFER Centre

Control of release and blowdown from H2 storage

Reduction of separation distances for H2 releases, jet fires, pressure

effects of deflagrations/detonations

Research outline

PhD student Mr David Yates, thesis title “Innovative solutions to reduce separation

distances in hydrogen systems”

PhD student Mr Sergii Kashkarov, thesis title “Fire resistance of onboard high

pressure storage tanks for hydrogen-powered vehicles”

UU RA (to be hired), research to focus on hazards assessment of HFC technologies,

development of engineering safety models/tools

Other research activities in line with SUPERGEN Hub strategy:

EPSRC SUPERGEN Challenge project “Integrated safety strategies for onboard hydrogen storage” (EP/K021109/1),

EC FP7 project H2FC (“Research Infrastructures for Hydrogen & Fuel Cells facilities”),

FCH-JU projects HyIndoor (“Pre-normative research on safe indoor use of fuel cells and hydrogen systems”), SUSANA (“Support to safety analysis of hydrogen and fuel cell technologies”), HyResponse (“European Hydrogen Emergency Response training programme for First Responders”),

PhD study Mr Jean Meyer (PhD enrolment 2014, PhD title “Optimisation of onboard hydrogen storage fire resistance”),

seeding research on predicting hazards of delayed ignition of hydrogen jets

Separation distance (1/2)

“Innovative solutions to reduce separation distances in hydrogen systems”

Modelling of hazards from hydrogen jet fires

Scientifically informed measures to control of jets from H2 storage

Engineering solutions for reduction of separation distances (e.g.

plane nozzle)

Progress:

The model (Makarov and Molkov, 2013) was advanced to

account combustion in a near field and compressibility at far

field.

The results are in agreement with experiments (Mogi and

Horiguchi, 2009), and simulation (Makarov and Molkov, 2013)

Validity of applied approach in the later paper

Separation distance (2/2)

Mogi and Horiguchi,2009

Simulation

AR=12.8

AR=1.0

Simulated vs experimental H2 jet flame

Control of H2 release

Control of blowdown from H2 storage

Variable aperture TPRD targeting jet fire length ~1 m

Simulation using deforming grids (work-in-progress)

“Closed” TPRD:

700 bar, mass flow rate 4.0 g/s, flame length 1.1 m

“Open” TPRD:

37.5 bar, flow rate 1.3 g/s, flame length 0.62 m

Fire resistance

“Fire resistance of onboard high pressure storage tanks for hydrogen-powered vehicles”

Longer blowdown requires increase of storage fire resistance

Reduction of separation distances for H2 releases, jet fires

Pressure effects of tank raptures/deflagrations/detonations

Progress

Bare tank bonfire – formulated model, running simulations

Pre-tests with propane and methane-air burners in line with

Global Technical Regulation (GTR 2013)

Qualitative testing of simplified tank failure criterion (based on

temperature of resin melting)

Qualitative agreement with type 4 tanks Fire Resistance Rating

(FRR) approx. 15 min

Propane burner

Tamura et al., 2012

CFD, 336.5 kW

FRR=13 min

Methane-air pretest

KIT (Germany) facility

CFD, 336 kW

FRR=15 min

Bonfire dynamics

EPSRC project “Integrated safety strategies for

onboard hydrogen storage systems” (project No. EP/K021109/1 )

Progress at University of Ulster

Project overview

Participants

University of Ulster (Dr D Makarov, Prof V. Molkov, Dr P. Joseph)

University of Bath (Dr T. Mays)

University of Warwick (Prof J. Wen)

Motivation

Fire resistance rating (FRR) of tanks made of carbon-fibre reinforced polymer (CFRP) are

currently unacceptably low – 6-12 minutes.

To facilitate low FRR temperature activated pressure relief devices (TPRD) are designed to

release hydrogen as fast as possible, creating jet fire hazard

Hydrogen jet fires may reach up to 15 m length, safe separation distance – up to 50 m

Aim

Develop novel safety strategies and engineering solutions for onboard storage of hydrogen

Objectives

Conduct parametric studies of tank performance in fires to optimize its fire resistance

Test Type 4 tanks, demonstrate performance of proposed solutions to increase fire

resistance

Improve bonfire and TPRD test protocols, including input of fire loading;

Perform economic analysis and evaluate reduction in risk of HFC vehicles with longer fire

resistance.

Project structure

WP1. Safety strategies for onboard storage (leader UU, M1-36)

WP2. Degradation and failure mechanisms (leader UW, M1-48)

WP3. Fire resistance prediction tools (leader UU, M1-48)

WP4. Testing tank prototypes with increased fire resistance (UU,M12-36)

WP5. Novel storage and safety solutions (leader UB, M1-36)

WP6. Bonfire and PRD testing protocols, outreach programme (leader UU,

M12-48)

WP7. Socio-economic effects of hydrogen safety (leader UU, M30-48)

Fire resistant oboard storage

Fire resistance: 1-2 hour (instead of 5 min)

Flame length: less than 1 m (instead of 15 m)

Automated control of tank aging

Novel

FRR prediction tools (1/4)

Task 3.1. Fire resistance model (UU, M1-36)

Simplified failure criterion: unprotected (“bare”) tank performance

Criterion

Local temperature

at 0.44 of CFRP thickness

is equal to

resin melting temperature:

𝑇 ≥ 𝑇𝑚𝑒𝑙𝑡𝑖𝑛𝑔,

FRR prediction tools (2/4)

Task 3.2. Optimisation of fire protection solutions (UU, M18-48).

Towards development of intumescent paint coated tank bonfire performance

Tank structure Tank structure Tank structure

Unreacted coating Unreacted coating

Tank structure

Unreacted coating

reacting coating reacting coating

char

char

char

Heat flux

1. Physical process

Melting: The polymer matrix melts and degrades to form a viscous fluid

Intumescence: Components within the coating decomposes producing

gas, some fraction is trapped within the molten matrix

Char formation The molten fluid hardens

Char degradation: The char layer pyrolyses leaving inert porous matrix

Flow chart

Initialize

Calculate gas trapped

among bubbles

Calculate expansion

ratio

Compute mass fraction

in each cells

Determine bubble burst

and expansion stop

Calculate volume

fraction of bubbles and

condensed phase

Determine material

properties

Calculate temperature

field

𝐾𝑗= 𝐴𝑗𝑒𝑥𝑝 −𝐸𝑖𝑅𝑇

, 𝑗 = 1,2,3

𝜕𝑚1

𝜕𝑡= −𝑚0,1𝐾1𝑌1,

𝜕𝑚2

𝜕𝑡= −𝑚0,2𝐾2𝑌2,

𝜕𝑚3

𝜕𝑡= −𝑚0,3𝐾3𝑌3,

𝜕𝑚𝑐

𝜕𝑡= 𝜈𝑐𝑚0,3𝐾3𝑌3,

𝑚 𝑔𝑜𝑢𝑡-𝑚 𝑔

𝑖𝑛 =𝜕𝑚𝑔

𝜕𝑡 - 𝜕 𝜀𝑥𝜚𝑔

𝜕𝑡

𝑥𝑠 = 𝑥0 1 − 𝜀0𝑚𝑠

𝑚0, 𝜀 =

𝑥−𝑥𝑠

𝑥

𝜚=𝑊𝑔𝑃0

𝑅𝑇

Mass Conservation

Mass continuity

Solid phase volume

State equation

𝜕

𝜕𝑥𝜆𝜕𝑇

𝜕𝑥= 𝑚𝑠𝐶𝑠+𝑚𝑔𝐶𝑔

𝜕𝑇

𝜕𝑡 + 𝐶𝑔𝑇

𝜕 𝜀𝑥𝜚𝑔

𝜕𝑡 + 𝐶𝑠𝑇

𝜕𝑚𝑠

𝜕𝑡 + 𝐶𝑔

𝜕 𝑚 𝑔𝑇

𝜕𝑥

𝜕 𝑚𝐶𝑇𝜕𝑡

+ 𝐻𝑗𝜕𝑚𝑗

𝜕𝑡

∆𝐻

𝜆∗ = 𝜆𝑠(𝜆𝑔

𝜆𝑠𝜀23 + 1 − 𝜀

23 )/(

𝜆𝑔

𝜆𝑠𝜀23 − 𝜀 + 1 − 𝜀

23 + 𝜀)

𝜆𝑔 = 𝜆𝑐𝑜𝑛𝑑+𝜆𝑟𝑎𝑑 𝜆𝑐𝑜𝑛𝑑 = 4.815 × 10−4𝑇0.717 𝜆𝑟𝑎𝑑 =2

3× 4𝑑𝑒𝜎𝑇3

Energy conservation

2. Intumescent paint model (following Di Blasi)

Node updating

Δ𝑥𝑖 = Δ𝑥0 + ∆𝑡𝑇𝑖𝛽𝑅

𝑃0𝑊𝑉2

𝜕 𝑚2

𝜕𝑡, 𝑇𝑚 < 𝑇𝑖 < 𝑇𝑐

…(3/4)

4. Preliminary simulation of expansion process (Ansys FLUENT 14.5)

𝐾𝑗= 𝐴𝑗𝑒𝑥𝑝 −𝐸𝑖𝑅𝑇

, 𝑗 = 1,2,3

𝜕𝑚1

𝜕𝑡= −𝑚0,1𝐾1𝑌1,

𝜕𝑚2

𝜕𝑡= −𝑚0,2𝐾2𝑌2,

𝜕𝑚3

𝜕𝑡= −𝑚0,3𝐾3𝑌3,

𝜕𝑚𝑐

𝜕𝑡= 𝜈𝑐𝑚0,3𝐾3𝑌3,

Mass Conservation

𝜕

𝜕𝑥𝜆𝜕𝑇

𝜕𝑥= 𝜚𝐶𝑝

𝜕𝑇

𝜕𝑡 + 𝐻𝑗

𝜕𝑚𝑗

𝜕𝑡

∆𝐻

Energy conservation

Node updating

Δ𝑥𝑖 = Δ𝑥0 + ∆𝑡𝑇𝑖𝛽𝑅

𝑃0𝑊𝑉2

𝜕 𝑚2

𝜕𝑡, 𝑇𝑚 < 𝑇𝑖 < 𝑇𝑐

Void

Paint

Air

Computational Domain

Expanding surface

…(4/4)

UU progress summary

Research in both SUPERGEN Hub (core) and in SUPERGEN

Challenge projects are being developed as planned

PhD projects are targeted at safety solutions readily available for

practical implementation:

Reduction of separation distances in hydrogen systems

Fire resistance of high pressure storage tanks for hydrogen-

powered vehicles

Focus of SUPERGEN Challenge project is engineering of integrated

safety solutions for Type 4 tanks to be used in automotive industry

Model of bonfire tank performance was formulated; modelling of

thermal protection (intumescent paint coating) is under

development

Contribution to regulations is expected as a particular result of the

project

24

Progress at University of Warwick

Prof J. Wen, Dr Z. Saldi

WP2. Degradation and failure mechanisms

Task 2.2. Finite element analysis of tank failure

Elmer

• Open source FEM solver for multiphysical

problems.

• Fluid dynamics, structural mechanics,

electromagnetics, heat transfer, acoustics.

• Thermomechanical problem (relevant in this

project)

• http://www.csc.fi/english/pages/elmer

25

Thermomechanical solver

• Heat transfer

• Linear elasticity

26

rcp¶T

¶t+ u ×Ñ( )T

æ

èç

ö

ø÷-Ñ× kÑT( ) = t :e + rSh

convective term (neglected) frictional viscous heating (neglected)

r¶2d

¶t2-Ñ×t = f

Initial test case: pressure vessel

• Ref:

www.openeering.com/sites/default/files/

Thermo_Elasticity_Scilab.pdf

Constant internal pressure of 1MPa

28

Results: temperature

Elmer

Ref

29

Test case: Acetylene cylinders in fire

• Ferrero et al., J. Loss

Prevention in the Process

Industries 25 (2012) 364-372

• Heat conduction problem,

convection neglected

• With heat release due to

decomposition reaction

• Fully enveloped by fire (heat

flux b.c due to convection &

radiation)

Acetylene

Steel AISI 4340

31

Sh = -Ae

-Ea

RT

æ

èç

ö

ø÷

Qdec

Initial results: inner shell temperature

(no decomposition reaction)

Cylinder radius = 0.115 m, height = 1.06 m 32

Test case for decomposition

reaction

33

Top & bottom: insulated wall

Cylinder heated at 325 K

Convective

outlet

Inlet vel =

5e-4 m/s

cA =

1000 mol/m3

Decomposition reaction: A kA¾ ®¾ F kA = Ae-Ea

RT

æ

èç

ö

ø÷

COMSOL model gallery, ID: 2164

Test case for decomposition

reaction

36

Left/right: temperature/species concentration

Top/bottom: without/with heat release

Stronger thermal decomposition in the hot area

close to the cylinder

Test case for decomposition

reaction

37

Summary

• Satisfying results from test cases on

thermo-mechanical & thermal

decomposition problem using Elmer

• Work in progress to couple thermo-

mechanical & thermal decomposition

aspects

• Elmer promising tool

• Basic know-how achieved 38

WP5: Novel storage and safety solutions

Co-I, Tim Mays, Chemical Engineering, Bath (15/10/13 – 14/10/17) 36 month FT PDRA, Dr Nuno Bimbo (17/2/14 – 16/2/17)

(funded from project) 36 month FT PhD student , Leighton Holyfield (1/10/14 – 30/9/17) (funded from University and EPSRC CDT in Sustainable Technologies)

Task 5.1: Review of innovative storage solutions (M1-6)

Focus on hybrid sorbent / high-pressure gas storage systems

Task 5.2: Comparative analysis of storage safety options (M3-36)

What are the key design and safety challenges and options for hybrid

sorbent / high-pressure gas storage systems?

Workpackage Details @ Bath

Concept

Safety aspects of integrating nanoporous sorbents into Type IV tanks

metal-organic frameworks

polymers of intrinsic microporosity

activated arbons

…

currently 70 MPa, 300 K

Compare physisorption with compressed gas

PC Storage pressure

Adsorptive storage

Compressed gas

Adsorption

favoured

Compression

favoured

Am

ou

nt o

f sto

red

hyd

rog

en

Modelling

0 5 10 15 20 25

0

10

20

30

40

5089 K - TE7 carbon beads

Hydro

gen u

pta

ke /

g L

-1

Absolute pressure, P / MPa

empty of adsorbent

quarter full of adsorbent

half full of adsorbent

full of adsorbent

Sorption Data

Research Questions

• Nanoporous sorbents in tanks increase volumetric

capacity at low pressures and temperatures

• Do these benefits extend to temperatures near

ambient?

• Are some sorbents “better” than others?

• Charge hybrid tank at cryogenic temperatures

and low pressures then allow warming to

ambient, with a consequent pressure rise. Is

this safe and techno-economically feasible?

Thank you!