Minimum Ignition Energy and Ignition Probability For Methane, Hydrogen and Their Mixtures

Loughborough University
Institutional Repository
Minimum ignition energy

and ignition probability for
Methane, Hydrogen and
their mixtures
This item was submitted to Loughborough University's Institutional Repository

by the/an author.
Additional Information:
• A Doctoral Thesis. Submitted in partial fullment of the requirements for

the award of Doctor of Philosophy of Loughborough University.
Metadata Record: https://dspace.lboro.ac.uk/2134/19907
Publisher: c Hemant Mathurkar

Rights: This work is made available according to the conditions of the Creative
Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-
NC-ND 4.0) licence. Full details of this licence are available at: https://creativecommons.org/licenses/by-
nc-nd/4.0/
Please cite the published version.

<
! ill~lilil[mllllm I I I I I I I~
---------- - - ------
... to my daughter ...

Minimum Ignition Energy and Ignition probability for Methane,
Hydrogen and their mixtures
by
Hemant Mathurkar
A Doctoral Thesis Submitted in partial fulfillment

of the requirements for the award of
DOCTOR OF PHILOSOPHY
of Loughborough University
June, 2009
Department of Chemical Engineering
© Hemant Mathurkar, 2009

ABSTRACT
In the present hydrocarbon economy, energy is primarily derived from fossil fuels,
like Coal, Oil and Gas. The petroleum oil obtained from mother earth is further,
refined into gasoline, diesel, and natural gas. However, the burning of these
hydrocarbon fuels causes the emission of greenhouse gases and other pollutants.
Hydrogen the lightest of all gases and the most abundant element in the universe, is
being considered for use as an energy carrier (for storing and transporting energy)
for future generations. Emphasis on mitigating global climate change and reducing
pollution, strengthens the case of hydrogen over other fuels. The environmentally
benign nature of hydrogen coupled with the finite supply of fossil fuels supports the
hydrogen economy.
A possible transition to the full hydrogen economy is envisaged which will take place
through several phases. The current work is concerned with the transitional phase
and involves an investigation into the possibility of using the existing natural gas
infrastructure for transporting hydrogen as a natural gas-hydrogen mixture. Likely
impacts on the natural gas infrastructure as a consequence of the introduction of
hydrogen are being studied as part of a European Union funded research project
called Naturalhy. The work that is the subject of this thesis forms part of the safety
work package of the Naturalhy project. In turn the part of the safety work package
with which the work of this thesis is concerned is the changes that handling a mixture
of natural gas and hydrogen rather than natural gas will have on the risk that will be
posed to the general public. In particular, it is concerned with the changes that might
result to such parameters as the ease with which mixtures of hydrogen and natural
gas might be ignited compared with natural gas and hence the change to the
frequency with which such events as explosions within domestic properties might
increase.
The work commenced with a review of the literature on the subjects of failure
probability and ignition probability associated with natural gas infrastructure. The
analysis and the outcome of this literature review suggested that the most sensitive
area affected by the addition of hydrogen is accidental gas releases into confined
enclosures such as domestic property. The presence of hydrogen is likely to
increase the probability of fire and/or explosion due to the characteristic properties of
hydrogen (wide flammability range, lower minimum ignition energy etc.).
ii
-----------------------------------------------------------------
The ignition characteristics for the gases (methane, hydrogen and methane-hydrogen
mixtures) was studied using an experimental rig based on the principle of capacitive
spark discharge. Consequently, the data obtained through experiments was used to
calculate the Minimum Ignition Energy (MIE) of a particular gas and the Lowest
Ignition Energy at various flammable gas concentrations for a particular gas. The
results and observations were further analysed to provide information on the ignition
probability associated with various ignition energy values for all the gases. The
results for MIE are compared with the available data in the literature for methane and
hydrogen gas.
Generalised correlations for predicting the ignition energy for pure gases and for two
component (methane-hydrogen) gas mixtures were developed. Methane gas release
incidents are compared with hydrogen to estimate increases in the probability of fire
and/or explosion incidents using a few deterministic release rates for the two gases.
iii
------------------------------------ - - - - -
ACKNOWLEDGEMENT
I express my gratitude to Professor Geoff Hankinson, Chair of Safety Engineering,

Department of Chemical Engineering, Loughborough University, for his valuable
guidance, suggestions, constructive criticism and constant encouragement during the
work. The financial help and support made available by Professor Hankinson for
attending technical meetings and conferences on the subject matter has. really
helped a lot.
I want to thank Dr. Barbara Lowesmith, for her great ideas and advise; which helped
revise and update the chapters of this thesis.
I am also thankful 10 Dr. Michael Acton, Principal Consultant, and his team members
especially Dr. Mark Rogers and Dr. Phi I Cleaver from Advantica for their kind help,
useful technical discussions and valuable advice.
I would like to take this opportunity to thank Mr. Tony Eyre, Mr. Terry Neal, Mr. Jim
Muddimer, and other department staff who helped to fabricate the experimental rig.
I would like to make specific mention of my brothers, Narendra and Praveen, who
educated me on the electrical circuit of my test rig. I owe my electrical engineering
knowledge to them.
I am grateful to me parents because of whom, I am on this earth and trying to

understand the subject matter. Quite often during the course of my thesis my
personal and family life has suffered and my wife and daughter have borne the brunt
of it. I appreciate their understanding, and am grateful for their love and moral
support.
I would like to thank my organisation, NEERI and specially my Head of Department

for extending the study leave for the PhD work.
I am also grateful to the people of the town of Loughborough and Loughborough

University, my colleagues and the office' staff within the Chemical Engineering
Department, for making my entire stay in the UK a truly memorable time.
I extend my heartfelt thanks and warmest gratitude to one and all.
iv
Table of contents
Certificate of Originality
Abstract ii
Acknowledgement iv
Table of Contents v
List of Figures ix
List of Tables xi
1 Introduction ........................................................................................................ 1
1.1 Background ................................................................................................ 1
1.2 Motivation .................................................................................................. 3
1.3 The Naturalhy projec!... .............................................................................. 5
1.4 Likelihood of failure .................................................................................. 11
1.4.1 The UKOPA pipeline fault database .................................................. 12
1.4.2 The EGIG report on gas pipeline incidents ....................................... 14
1.4.3 The US Department of Transport (DOT) data ................................... 15
1.4.4 Low pressure pipeline incidents ....................................................... 18
1.5 Impact of hydrogen .................................................................................. 19
1.5.1 Impact on pipeline failures ................................................................ 19
1.5.2 . Impact on ignition probability ............................................................ 21
1.6 The Thesis structure ................................................................................ 23
2 Literature Review ............................................................................................. 25
2.1 Ignition theory .......................................................................................... 25
2.2 Historical data on ignition of gas releases ................................................ 27
2.2.1 High pressure releases .................................................................... 27
2.2.2 Low pressure releases ..................................................................... 29
2.3 Sources of Ignition ................................................................................... 31
v
2.3.1 Common Ignition sources ................................................................. 31
2.3.2 Ignition sources for transmission pipeline release ............................ 34
2.3.3 Ignition sources for Low pressure release ........................................ 40
2.3.4 Analysis of Ignition sources .............................................................. 45
2.4 Ignition model for gas releases ................................................................ 47
2.4.1 Cox, Lees and Ang model ................................................................ 49
2.4.2 PIPESAFE - Ignition probability model for transmission pipeline ...... 50
2.4.3 WS Atkins model .............................................................................. 51
2.4.4 Ignition probability estimate for low pressure releases ...................... 53
2.4.5 Model limitation ................................................................................ 56
2.5 The Minimum Ignition Energy of gases .................................................... 57
2.5.1 MIE determination ............................................................................ 58
2.5.2 MIE and quenching distance ......................... 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.5.3 Historical MIE data ........................................................................... 63
2.5.4 Limitations of MIE apparatus ............................................................ 65
2.6 Ignition energy and ignition probability for gases ...................................... 65
3 Design and Operation of the Experimental facility ............................................ 67
3.1 Introduction .............................................................................................. 67
3.2 Conceptual design ................................................................................... 67
3.3 Gas supply, delivery and evacuation system ........................... ,................ 68
3.4 Cylindrical combustion chamber .............................................................. 71
3.5 The Capacitive spark discharge circuit.. ................................................... 73
3.6 Instrumentation ........................................................................................ 76
3.6.1 Pressure Transducers ...................................................................... 76
3.6.2 Temperature thermocouple .............................................................. 77
3.6.3 Oscilloscope ..................................................................................... 77
3.7 Data acquisition (LabVIEW interface) ....................................................... 78
3.8 Safety Systems ............................. : .......................................................... 82
vi
-----------------------------------
3.9 Experimental procedure ........................................................................... 83
3.9.1 Gas mixture preparation ................................................................... 85
3.9.2 Gas Ignition ...................................................................................... 86
3.10 Experimental programme ......................................................................... 88
3.10.1 Equivalence Ratio ............................................................................ 89
3.10.2 Flammability limits ............................................................................ 91
3.10.3 Targeted gas concentrations ............................................................ 92
4 Results, Data analysis and Discussions ........................................................... 93
4.1 Data analysis ........................................................................................... 93
4.1.1 Concentration estimation .................................................................. 93
4.1.2 Energy estimation ............................................................................ 95
4.1.3 Ignition probability computation ...................................................... 100
4.2 Error Analysis ........................................................................................ 102
.4.2.1 Error analysis for concentration ...................................................... 103
4.2.2 Error analysis for energy ................................................................ 107
4.3 Results for methane ............................................................................... 109
4.3.1 Comparison with previous data ...................................................... 111
4.3.2 Ignition probability .......................................................................... 112
4.4 Results for hydrogen .............................................................................. 114
4.4.1 Comparison with Previous data ...................................................... 116
4.5 Results for methane hydrogen mixtures ................................................. 120
4.5.1 Gas with 75 % methane and 25 % hydrogen .................................. 120
4.5.2 Ignition probability of 75% CH. + 25% H2 mixture ........................... 122
4.5.3 Gas with 50% methane and 50% hydrogen .................................... 124
4.5.5 Gas with 25% methane and 75% hydrogen .................................... 128
vii
4.6 Overall results ........................................................................................ 132
4.6.1 Ignition energy ................................................................................ 133
4.6.2 Prediction of Ignition probability ...................................................... 143
4.6.3 Ignition probability computation ...................................................... 145
4.6.4 Impact of hydrogen ........................................................................ 149
Chapter 5 .............................................................................................................. 152
5 Conclusions of the Experimental study ........................................................... 152
5.1 Main findings of the study ...................................................................... 153
5.1.1 Minimum Ignition Energy ................................................................ 153
5.1.2 Ignition Energy ............................................................................... 154
Chapter 6 .............................................................................................................. 158
6 Changes in gas incident probability ................................................................ 158
6.1 Comparing methane and hydrogen releases .......................................... 158
Chapter 7 .............................................................................................................. 162
7 Issues for future work ..................................................................................... 162
7.1 Limitations of the experimental setup ..................................................... 162
7.1.1 Modifications proposed to the experimental setup .......................... 163
7.2 Future work ............................................................................................ 165
References ............................................................................................................ 167
Appendix
Appendix A : Research work meetings, presentation and technical discussions 173
attended
Appendix B : Properties of Hydrogen and Methane gas 174
Appendix C : Calculations for the vessel design 175
Appendix D : Risk Assessment for the experimental rig 181
Appendix E : Operating procedure 192
Appendix F : Electrical circuit and Instrumentation 194
Appendix G : Specifications of the Pressure Transducers 196
Appendix H : Experimental runs with CH., H2 and CH.-H2 mixtures 197
viii
List of Figures
Figure 2.1 : Ignition of a flammable cloud (a) by spark (b) by auto-ignition .............. 25
Figure 2.2 : Ignition probability variation with time for gas leak inside house ........... 56
Figure 3.1 : Schematics of the experimental setup .................................................. 70
Figure 3.2 : Schematic representation of the cylindrical combustion chamber ......... 71
Figure 3.3 : The electrode assembly ....................................................................... 73
Figure 3.4 : Simplified electrical circuit for Capacitive spark discharge .................... 74
Figure 3.5 : Voltage trace for the spark discharge captured on the scope ............... 78
Figure 3.6 : LabVIEW (front panel) with data file and readings ................................ 81
Figure 3.7 : Flowchart for the experiment ................................................................ 84
Figure 4.1 : Pressure recording during gases introduction ....................................... 94
Figure 4.2 : Pressure recording during methane and hydrogen introduction ............ 95
Figure 4.3 : Overpressures during methane and hydrogen ignitions ....................... 96
Figure 4.4 : Overpressure for lean hydrogen ignitions ............................................. 97
Figure 4.5 : Errors analysis for experiments with 6.5% & 8% methane concentrations
............... :.............................................................................................................. 109
Figure 4.6 : Results of ignition experiments for methane ....................................... 110
Figure 4.7 : Comparision of methane ignition energy results with those from the
literature ................................................................................................................ 112
Figure 4.8 : Ignition probability at various energies along with exponential trend lines
for methane ........................................................................................................... 114
Figure 4.9 : Ignition energies for a range of hydrogen concentrations ................... 115
Figure 4.10 : Comparision of hydrogen igntion energy results with those from the
literature ................................................................................................................ 117
Figure 4.11: Ignition probability at various energies along with exponential trend
lines for hydrogen .................................................................................................. 120
Figure 4.12: Ignition energies for 75% CH. + 25% H2 mixture .............................. 121
for 75% CH. + 25% H2 mixture .............................................................................. 123
ix
Figure 4.14: Ignition energies for 50% CH, + 50% H2 mixture .............................. 125
for 50% CH, + 50% H2 mixture .............................................................................. 127
Figure 4.16 : Ignition energies for 25% CH, + 75% H2 mixture .............................. 129
for 25% CH, + 75% H2 mixture .............................................................................. 131
Figure 4.18 : Observed Ignition probability plotted against the calculated value .... 133
Figure 4.19 : Lowest ignition energy at various ER of flammable gases ................ 133
Figure 4.20 : Normalised ignition energy vs normalised ER for lean gas composition
.............................................................................................................................. 135
Figure 4.21 : Normalised Ignition energy vs normalised ER for rich gas composition
.............................................................................................................................. 1~
Figure 4.22 : Normalised Ignition energy vs normalised ER for all gases .............. 137
Figure 4.23 : MIE variation for various concentration of hydrogen in mixture ......... 137
Figure 4.24 : Equivalence Ratio variation with molecular weight of flammable gas 139
Figure 4.25 : Equivalence Ratio variation with molecular weight of flammable gas
including ethane and propane gas ......................................................................... 140
Figure 4.26 : MIE variation with Laminar burning velocity ...................................... 141
Figure 4.27 : Ignition energy for gases at various Equivalence Ratios ................... 142
Figure 4.28 : Observed and predicted values of Ignition energy ............................ 142
Figure 4.29 : Variation of parameter 'b' with Lowest Ignition energy ...................... 143
Figure 4.30 : Variation of parameter 'Em;n' against Lowest Ignition energy ............. 143
Figure 4.31 : Ignition probability variation for flammable gases at ER = 0.67 ......... 147
Figure 4.32: Ignition probability variation for flammable gases at ER = 0.93 ......... 148
Figure 4.33 : Ignition probability variation for flammable gases at ER = 1.2 ........... 148
Figure 4.34 : Lowest ignition energy at various flammable concentrations ............ 149
Figure 4.35 : Lowest ignition energy varation with increasing hydrogen percentage
.............................................................................................................................. 151
Figure 6.1 : Gas concentration built up for methane, hydrogen release in room .... 160
x
List of Tables
Table 1.1 : Qualitative change due to hydrogen addition ......................................... 11
Table 1.2: Failure Frequency and Causes for Pipeline Failures within UK .............. 13
Table 1.3 : Failure frequency and Incident cause for gas pipelines from EGIG
database ................................................................................................................. 15
Table 1.4 : Failure frequency and various causes for gas transmission pipeline
incidents within the USA. ......................................................................................... 17
Table 1.5 : Annual gas escapes requiring repair for year 2005/06 within UK ........... 19
Table 1.6 : Characteristics of hydrogen in comparison with other fuels ................... 22
Table 2.1 : Observed ignition incidents of released gases ....................................... 27
Table 2.2 : Pipeline Ignition incidents based on release size ................................... 28
Table 2.3 : Gas In Buildings events and ignition incidents for the UK ...................... 30
Table 2.4 : Common Ignition sources for reported incidents in process plants ......... 32
Table 2.5 : Ignition sources for major fire incidents .................................................. 33
Table 2.6 : Summary of ignition between rocks and metal. ...................................... 38
Table 2.7: Energies for various ignition sources ......................................................46
Table 2.8 : Ignition potential for various sources ...................................................... 47
Table 2.9 : Estimates for Ignition probability by various researchers ....................... 48
Table 2.10 : Ignition probability estimates for flammable releases ......................·..... 49
Table 2.11 : Source parameters for the WS Atkins model ....................................... 52
Table 2.12: Typical Ignition sources within homes .................................................. 54
Table 2.13 : MIE and quenching distance for flammable gases ............................... 61
Table 2.14: Theoretical estimates of MIE for gases ................................................ 63
Table 2.15: Minimum spark ignition energy data from various sources .................... 64
Table 3.1 : Energy delivered by Capacitor ............................................................... 75
Table 3.2 : Flammability limits of gas mixture .......................................................... 91
Table 3.3 : Target gas concentrations for the experimental programme .................. 92
Table 4.1 : Error in pressure measurement ........................................................... 103
xi
Table 4.2 : Nominal and actual capacitance of the capacitors ............................... 107
Table 4.3 : Error in voltage measurment... ............................................................. 107
Table 4.4 : Experimental conditions for methane tests .......................................... 109
Table 4.5 : Observed Ignition energies for various methane concentrations .......... 111
Table 4.6 : Ignition probability for various methane concentrations ........................ 113
Table 4.7 : Ignition energy at 1% ignition probability for methane .......................... 113
Table 4.8 : Experimental conditions for hydrogen .................................................. 115
Table 4.9 : Observed Ignition energies for various hydrogen concentration ........... 116
Table 4.10: Ignition probability for very lean (6%) hydrogen concentration ........... 118
Table 4.11 : Ignition probability for 20% to 45% hydrogen concentration ............... 118
Table 4.12 : Ignition probability for lean (10%) and rich (60%) H2 concentration .... 118
Table 4.13 : Ignition energy at 1% ignition probability for hydrogen ....................... 119
Table 4.14: Experimental conditions for 75% CH, + 25% H2 mixture .................... 120
Table 4.15 : Observed Ignition energies for 75% CH, + 25% H2 mixture ............... 121
Table 4.16: Ignition probability for 75% CH, + 25% H2 mixture ............................. 122
Table 4.17 : Energy at 1% ignition probability for 75% CH, + 25% H2 mixture ....... 123
Table 4.18 : Experimental conditions for 50% CH, + 50% H2 mixture .................... 124
Table 4.19: Observed Ignition energies for 50% CH, + 50% H2 mixture ............... 125
Table 4.20 : Ignition probability for 50% CH, + 50% H2 mixtu're .............................. 126
Table 4.21 : Energy at 1% ignition probability for 50% CH, + 50% H2 mixture ....... 126
Table 4.22: Experimental conditions for 25% CH, + 75% H2 mixture .................... 128
Table 4.23 : Observed Ignition energies for 25% CH, + 75% H2 mixture ............... 129
Table 4.24: Ignition probability for 25% CH, + 75% H2 mixture ............................. 130
Table 4.25: Energy at 1% Ignition probability for 25% CH, + 75% H2 mixture ....... 131
Table 4.26 : Ignition probability of flammable gases at various energies ............... 146
Table 4.27 : Ignition energy at various ignition probabilities ................................... 150
Table 6.1 : Ignition probability for methane releases within confined areas ........... 160
Table 6.2 : Ignition probability for hydrogen releases within confined areas .......... 161
xii
Nomenclature used
A the area of the region being in contact with the gas cloud (m 2 )
a probability of Ignition source being active
b constant
C Concentration (%)
t::,C Error in parameter C (%)
CP Capacitance (pF)
d Pipe diameter / shell diameter (m)
diameter of hole or puncture in pipeline (m)
Outside diameter of shell (m)
diameter of optical window (for the experimental vessel) (m)
E Energy or Capacitive spark energy (mJ)
EF Expansion Factor
ER Equivalence Ratio
f Allowable working stress (MPa)
Current magnitude (ampere)
lE Ignition Energy (mJ)
IG Number of successful Ignitions of the gas air mixture
IP Ignition probability
LFL Lower Flammability Limit
m mass release rate of gas (kg/s)
MIE Minimum Ignition Energy (mJ)
MW Molecular Weight (gm)
N Number of Ignition sources
NER Normalised Equivalence Ratio
NI number of Non - Ignitions for the gas air mixture
xiii
NP Number of moles of all Products formed
NR Number of moles of all reactants
P, P Pressure (kPa)
l'.P Error in parameter P (kPa)
ph Hoop stress (MPa)
po ignition potential (probability of ignition) of single source
qd Quenching distance (cm)
Q probability of Non Ignition
R Ratio of moles of methane to moles of hydrogen
l'.R Error in parameter. R
SPH Specific heat of gas (J/moI/K)
Su Laminar burning velocity (cm/s)
T Temperature (K)
Tf Final temperature (adiabatic flame temperature) (K)
Ti Initial temperature (room temperature) (K)
Tb Burnt gas temperature (K)
Tu Un-burnt gas temperature (room temperature) (K)
t time duration (s)
ta time period for which the source is active (s)
t; time period between each activation of the source (s)
ths Shell thickness (for the experimental vessel) (mm)
thw Window thickness (for the experimental vessel) (mm)
UFL Upper Flammability Limit
V Voltage (volt)
l'.V Error in parameter V (volt)
xiv
Greek symbols
f1 ignition sources density per unit area of source (11m2)
A frequency of activation of the source (1/s)
p molar density of gas (mol/m 3 )
n resistance (ohms)
Subscripts
av average value
gas flammable gas under consideration
gas-air flammable gas air mixture
MIE at the MIE condition of the flammable gas
max Maximum value
min Minimum value
lower Lower value
upper Upper value
LFL Lower Flammability Limit
UFL Upper Flammability Limit
vac vacuum condition
t time
xv
Chapter 1
Introduction
1.1 Background
Life on Earth is driven by energy. Plants take it from solar radiation and are often
referred to as the "initiator of life" on this planet. Energy captured slowly by
photosynthesis is stored up often in a more denser and useable form. As denser
reservoirs of energy were found over the course of Earth's history, we tended to use
more energy by exploiting these denser resources. The ability to use energy outside
the body, enables humans to use far more energy than any other living creature on
this planet. The control of fire and the exploitation of fossil fuels (denser resources)
have made it possible for us to release, vast amounts of energy, in a short time, that
took centuries to accumulate within our mother earth, often referred to as natural
resources.
Energy is critical to every aspect of our lives. All material things that are of any use to
us - our food, clothes, houses, automobiles, - require energy to make and energy to
use. All human activities like - working, managing, thinking, teaching, - require
human energy. All such useful activities are often called "work." Thus, all work
requires energy. In performing work, energy is always changed from more-
concentrated to less-concentrated forms. In fact, this natural tendency gives energy
its ability to perform work. Material things, such as food, wood, plastic and gasoline
actually are concentrated forms of energy. Matter can be changed into energy, as
when we eat food or burn gasoline. Usefulness can also be gained by changing the
form of energy, as in using heat to make electricity and electricity to produce light.
Whenever energy is used to perform work, it becomes more dispersed and
disorganized.
Everything we do is connected to energy in one form or another. The amount of .

energy utilised may vary for each individual. Energy maintains our standard of living
and economy. The developed world often takes it for granted that energy will be
available whenever it is needed. Moreover industrial research and development is
1
aimed at bringing energy to the door step of each and every customer, hence the
need for "Energy transportation".
Energy can be obtained through various sources provided by nature. The most
common sources such as coal, oil, gas, wind, sun and nuclear fuels are often used
for satisfying energy demands. During the early stages, the emphasis was on
obtaining energy and less attention was paid towards its consequences. The impact
on surrounding ecology, environment and finally on humans was felt with the
increased utilisation of energy from various sources. In the present world most of this
energy comes from fossil fuels (coal, oil, and gas) which are depleting at an ever
increasing rate with the increase in population and our living standard. At current
rates of consumption, known reserves of oil & gas will only last for a few decades
and coal for a few centuries.
The technology for the commercial utilization of fossil fuels started in the early 18th
century and hence this technology for harnessing the energy contained within
resources such as coal, oil and gas is very well developed. The exploitation of these
resources has resulted in making life easier through utilising the energy released to .
provide services such as transport, cooking, communication etc. Since the 1970's the
adverse impact of combustion of these fossil fuels on the environment and the entire
ecosystem has been noted. Consequently, research is aimed at reducing the
environmental impact of fossil fuels by reducing the environmentally harmful
emissions. In addition, to ensure the security of supply of energy sources, research is
being undertaken to identify sustainable sources of energy. This will allow a transition
to be made from the finite fossil fuels to the infinite sustainable sources of energy.
The energy sources are often categorised as renewable and non-renewable.

Renewable energy sources are those which are continually being replaced such as
energy from the sun (solar) and wind. If an energy resource is being used faster than it
can be replaced (for example, coal takes millions of years to form) then it will
eventually run out. Hence all the effort that is being expended to conserve these non-
renewable energy sources.
2
These days, more emphasis is being focussed on the renewable sources which are
believed to supply sustainable source of energy. Solar, Wind, Tidal, Wave, Biomass,
Hydroelectric, Geothermal, etc. are often referred to as renewable energy sources but
it is not guaranteed ttiat they will produce energy all of the time. For example,
renewable energy sources, like the sun, can not produce energy all the time. The sun
does not always shine, so a device to store this energy and transport it from the
location of generation to the location of utilization is required. Some experts believe
that hydrogen can store this energy until it is needed. Then it can be transported to
where it is needed, and used without producing harmful emissions. Hence, hydrogen
has the capacity to form part of the energy infrastructure for future generations.
Indeed some experts think that hydrogen will form the basis of the energy
infrastructure that will power future societies, replacing today's natural gas, oil and
coal. They see a new "hydrogen economy" to replace our current "fossil fuel-based
economy".
1.2 Motivation
Hydrogen, the lightest of all gases and the most abundant element in the universe, is
being considered for use as an energy carrier (a means of storing and transporting
energy obtained from sustainable sources) for the twenty first century because the
recovery of energy stored in the form of hydrogen can be obtained in environmentally
benign ways. The idea of using hydrogen as an energy carrier has been around at
least since 1870, when Jules Verne incorporated the concept in his science-fiction
classic, 20,000 Leagues under the Sea. There is a widespread assumption that the
increased use of hydrogen as an energy carrier will bring significant benefits. Some
emphasise the prospect of cleaner air in cities as vehicles stop emitting pollutants,
others point towards mitigating global climate change.
The hydrogen economy is in its infancy. Research is required to increase

production fr.om sustainable sources. Research is needed on CO 2 sequestration, if H2
is produced from a source containing carbon. Research is needed on the utilization
of hydrogen such as fuel cells and combustion equipment.
3
While moves towards an increased use of hydrogen are starting to gather speed, this
growth is restricted by a number of constraints at the political, commercial, technical
and social levels. Safety concerns are still widespread in relation to the use of
hydrogen in our daily routine life. The public perception of the dangers of hydrogen
transportation and distribution need to be addressed if widespread hydrogen use is
envisaged in the future. At a practical level, there are real issues in terms of how
hydrogen is stored and transported. Hydrogen is a very light gas making it far more
difficult to work with than other liquid I gaseous fuels.
In the meantime, in order to be ready to swing into action, when the above research
bears fruit, work needs to be done on ways of transporting hydrogen from the point of
production to the point of utilization. The most economical, safest and environmentally
friendly method of transporting large quantities of liquids and gases on a regular basis
is by pipeline. Hence the construction of the natural gas, and water pipeline networks
and pipelines for gasoline, etc. A small number of purpose built hydrogen pipelines
exist. However, the construction of a nation wide hydrogen network will only
commence, if at all, once large quantities can be produced and a market exists for its
use.
From an economic point of view, the costs of switching over to a full hydrogen based
economy are high. For example, the capital investment required to convert existing
gasoline stations to provide hydrogen to vehicle drivers will run into millions of pounds.
The present phase can be termed as a transition phase for the hydrogen economy. A
means of carrying hydrogen from its production to its utilization sites is required during
the period of transition; from now until the full hydrogen economy is introduced. A
system is required that can readily increase in volume and throughput as production
and demand increases and can also manage the mismatch between production and
demand. Such systems already exits for natural gas and an examination of the use
that the natural gas infrastructure can play in the transport of hydrogen, in particular
during the transition to the full hydrogen economy, must be undertaken.
The distance between the production and utilization points for hydrogen can be a few
metres, to hundreds of kilometres. This requires a connection between the production
and the end-use facilities. This link can be made by means of road, rail, ship, pipeline
or air transport. As mentioned above, pipelines have been shown to be the most
4
economical, safest and environmental friendly method of transporting large quantities'
of material such as liquid or gas on a regular long term basis. This most viable option
has resulted in the development of the extensive natural gas infrastructures within
Europe, USA, Asia, etc. and is still expanding to other areas of the world.
Europe's Natural Gas (NG) pipelines are some of the most advanced in the world and
the infrastructures are evenly spread. These infrastructures are also connected to
other continents such as Asia and Africa. It is prudent to investigate whether or not
these existing NG infrastructures can be utilised to assist a transfer to the hydrogen
economy.
A possible vision of the use of the NG infrastructures to assist a transition to the

hydrogen economy is through various phases where increasing amounts of hydrogen
are blended with the NG to produce a mixture. The financial burden during the
transition period will be eased considerably by making use of the existing NG
infrastructure with its inherent economic value. It will also help in achieving a
seamless move to the hydrogen economy.
Several research and development activities have been initiated by the European
commission (EC) to study the subject of hydrogen transportation, including by
pipelines. These include the projects NATURALHY, HYSAFE, HyWays,
Roads2HyCom etc. The NATURALHY project is focussed especially on studying the
practicality of using the existing natural gas pipeline networks to assist in the transition
to the hydrogen economy by transporting mixtures of natural gas and hydrogen.
1.3 The Naturalhy project

The NATURALHY project is a major "Integrated Project" which has been funded by
the European Commission within the sixth framework programme. The purpose is to
prepare the European natural gas industry for the introduction of hydrogen by
assessing the capability of natural gas infrastructures to accept mixtures of hydrogen
and natural gas.
5
The objective of the NATURALHY project, is to investigate the feasibility of using the
natural gas infrastructures (pipeline networks) for transporting hydrogen-natural gas
mixtures from the point of hydrogen production to the point of use where the
hydrogen is extracted from the mixture. It is envisaged that this would aid the
transition to a hydrogen economy, i.e. under an expected situation where the
demand for hydrogen is increasing from a low base and the hydrogen supply is
carried out in parallel with the supply of natural gas.
The hydrogen injected into the natural gas network could be extracted by techniques
like membrane separation and Pressure Swing Adsorption (PSA) to utilise hydrogen
in a pure form. However, whatever separation technique is adopted, not all the
hydrogen can be extracted, and the gas supplied to users connected to the network
is likely to contain some amount of hydrogen. As hydrogen combustion will not result
in CO 2 formation there is some environmental benefit and this is often referred to as
the "greening of gas". Hence the greening of gas is a secondary objective of the
Naturalhy project and will result in Environmental benefits towards a reduction in CO 2
emissions. In 2004, the total primary energy consumption in the EU15 countries
amounted to 6.54 x 104 kJ of which 24% was provided by natural gas. If 1%
(equivalent to 3% by volume) of the energy content of the natural gas were replaced
by hydrogen (produced through CO 2 free production technique), the total CO 2
emission of the EU15 would be reduced by about 7.4 Million-tons/year (in the case of
oil 11 Million-tons/year and in the case of coal 13 Million-tons/year). This is about 0.2
% of the total annual CO 2 emission in the EU 15 countries. Accordingly, the potential
of hydrogen-natural gas mixtures for the reduction of CO 2 emissions could be
significant. (Florisson, et. aI., 2006). To summarise, the NATURALHY project
objectives are two fold
• Means of transporting hydrogen from point of production to point of use by

hydrogen fuelled applications
• 'Greening' of gas by replacing some NG by H2 , resulting in lower emissions
NATURALHY has a total budget of 17.3 million euros and started in May 2004, for a
duration of five years. The project partners comprise of gas companies, research
institutes, academic institutes, consultants, government bodies and planning and
implementing organisations, with Gasunie Research (N.V. Nederlandse Gasunie)
assigned the role of project coordinator.
6
The NATURALHY project is subdivided into eight Work Packages (WP) each based on
an area of research and development. The WP project titie, WP lead organisation and
a short description of the work being conducted are presented below:
Work Package Project Title Project Leader
Socio-economic and
WP1 Loughborough Univ., UK
Life Cycle Analysis
WP2 Safety Loughborough Univ., UK
WP3 Durability Gaz de France, France
Gas - und Umweltlechnik GmbH (DBI),
WP4 Integrity
Germany
WP5 End Use University of Oxford, UK
Instituto de Soldadura e Oualidade
WP6 Decision Support Tool
(ISO), Portugal
WP7 Dissemination Exergia, Greece
WP8 Project management Gasunie Research, The Netherlands
WP1 Life Cycle and Socio-economic Assessment comprises a comparison of the

existing natural resource requirements, environmental impacts, employment
consequences and economic costs over the complete life cycle of current natural gas
and related energy systems; proposed transitional natural gas/hydrogen systems;
and future complete hydrogen systems from source to point of use.
WP2 Safety is discussed at the end of this section in detail.
WP3 Durability determines the effect of hydrogen on the durability of materials and
components used in the natural gas infrastructure. For example, the permeability of
pipeline materials to hydrogen is being assessed and any effect on "ageing".
Hydrogen impact on pipeline materials is being evaluated through experimentation,
the results processed to produce mathematical durability lifetime models. A durability
assessment tool developed for assessing the ageing of pipeline materials and
components, will help provide a practical way of evaluating the lifetime of pipeline
systems under an increasing percentage of hydrogen.
7
WP4 Integrity assesses the effectiveness of NDE (Non Destructive Evaluation) tools
and techniques to monitor the condition of pipelines (transmission and distribution)
exposed to natural gas-hydrogen mixtures. The maintenance and repair procedures
for such pipeline systems are evaluated. Critical defects leading to pipeline failures,
under gaseous atmosphere are studied to arrive at the defect assessment criteria
relevant for natural gas-hydrogen mixtures. The output of this package is to develop
an integrity management tool for a pipeline system conveying a natural gas -
hydrogen mixture.
WP5 End Use examines the implications of providing distributed pipeline natural
gas-hydrogen mixtures to end-users and the effect on eXisting appliances. The
impact of added hydrogen on the performance characteristics of domestic and
industrial burners are studied. This will lead to greater understanding of the
requirements for appliances to operate safely and efficiently with increasing levels of
hydrogen. This work package is also studying the potential of using high efficiency
membranes for separating hydrogen from the gas mixture for end-use applications as
an alternative to PSA. Methods for controlling the gas quality within the network as
hydrogen is added and removed in the network are also being studied.
WP6 Decision support tool brings together the results of WP1 to WP5 into a
software based decision support tool, which can be used by pipeline operators to
assess the possibility of adding hydrogen to their network or part of their network.
WP2: Safety
The aim of Work Package 2 is to evaluate the change in risk presented to the public
by the natural gas infrastructure, if it were used to carry a mixture of natural gas and
hydrogen rather than simply natural gas. The existing gas pipeline infrastructure was
designed, constructed and operated based on the premise that natural gas is the
material to be conveyed. Since, hydrogen has different properties to natural gas,
which may adversely affect the risk presented to the public, this must be re-evaluated
for the mixture.
8
Risk is defined as a combination of likelihood of an undesired event (such as pipeline
failure, gas release, gas ignition etc.) and the consequence of that event (such as the
severity of heat load from fire or overpressure wave from explosion). Mathematically
risk is expressed as :
RISK = L likelihood x consequences

All events
For gas pipelines, the likelihood can be due to 'pipeline failure' leading to gas
release. The likelihood of ignition of released gas often termed as 'ignition
probability', can be also considered (along with pipeline failure) to arrive at the
likelihood of a fire. Consequences depend on the properties of released gas
(flammability limits, heat of combustion, specific heat, flame temperature, burning
velocity etc.) and also the surrounding conditions (weather, confined/unconfined
release etc.). Different scenarios/events can be considered to calculate risk through
summation of all these events to arrive at the risk level for the gas pipeline system.
Adding hydrogen to the gas infrastructure may affect both the likelihood and
consequence (severity) of undesired ·events and hence potentially change the risk
level. It is important to quantify this effect in order to establish if the risk remains
acceptable and to identify the maximum hydrogen concentration that can be added to
the natural gas without this risk becoming unacceptably high.
To this end, the Safety Work Package is examining the fire and/or explosion hazard
situations pertinent to the gas infrastructure, based on a natural gas-hydrogen
mixture being involved. To assess the change in consequences small and large scale
experiments are undertaken to provide data to aid model development and
validation. Specifically these are:
• Laboratory experiments to assess the burning velocities of methane/hydrogen

mixtures
• Large scale experiments of
o gas build up and explosions in a domestic environment
9
o gas build up and explosions in an industrial enclosure/building
o explosions in congested regions (Vapour Cloud Explosions)
o high pressure Jet fires
o fire hazards following failure of a transmission pipeline
Consequence models are being developed and validated using this experimental
data, These models are then used to assess the impact of different levels of
hydrogen addition on the severity of the hazards. Fire and explosion consequences,
and impact of hydrogen on these consequences, are not part of this thesis, as the
focus here is on the likelihood of the event.
The impact of hydrogen on the likelihood of an adverse event is studied through

experiments in this work package and also by work within the work packages on
Durability and Integrity. Data from WP3 and WP4, will help re-evaluate the failure
frequencies of pipelines and associated equipment. Within WP2, Laboratory scale
experiments are being conducted to help to re-assess ignition probability.
Specifically, ignition probability and the minimum ignition energy of natural gas-
hydrogen mixtures are being studied and this is the main focus of this thesis.
The above work on likelihood and consequences will then be combined to produce a
risk assessment tool, which can be applied to gas transmission networks conveying
methane-hydrogen mixtures. The risk to the public can be calculated for differing
levels of hydrogen introduction and compared with the risk for a system conveying
natural gas. The risk assessment tool will evaluate both individual and societal risk to
the surrounding population. A risk assessment methodology for low pressure
releases within a property, addressing the risk of an explosion will also be
considered. Again the situation with natural gas-hydrogen mixtures will be compared
to that with natural gas alone.
Natural gas properties are often considered as similar to pure methane, since it is the
major constituent of the natural gas. A comparison of hydrogen and methane
properties suggest changes (in likelihood and consequences parameters) as
10
presented in Table 1.1. Only qualitative changes are presented here, since
quantification is the objective of study within the NATURALHY project.
Table 1.1 : Qualitative change due to hydrogen addition
Parameter Potential qualitative change compared to

natural gas
Leaks (pipeline failures) Increase
Ruptures (pipeline failures) Increase
Ignition energy for gas Decrease
Ignition probability Increase
Energy content per m3 Decrease
Flammable range in open atmosphere Increase
Buming velocity Increase
Heat radiation from flames Decrease
Explosion severity Increase
Total Impact on risk Unknown
As a first step towards quantifying the impact of hydrogen addition, on the likelihood
of adverse events, the failure probability of high pressure pipelines and low pressure
systems conveying natural gas are considered in section 1.4. The effect of hydrogen
on this failure probability is considered in section 1.5.
1.4 Likelihood of failure

Despite the fact that pipelines are considered reliable for transmitting energy, there
have been incidents of gas release resulting in· loss of life and property damage. This
has prompted the need to understand the various modes of failure of pipelines. and
increasing effort is directed towards reducing the frequency of these accidents. The
first step is to estimate the failure probability (frequency analysis) for existing pipelines
operating in different parts of the world.
11
Pipeline failure frequency estimation based on historical incident data has formed the
basis for many risk assessment studies. (Acton 2002 and Willcocks et. al. 2000). The
source of data considered in the present analysis are worldwide pipeline failure events.
Pipeline failure databases from the UK, Europe and the US are considered. Low
pressure failures pertaining to the UK, are also discussed.
1.4.1 The UKOPA pipeline fault database
Natural gas for consumption in the UK comes to shore from the North sea and the Irish
sea. The gas, after processing, is transferred through high pressure transmission
pipelines, normally up to one metre in diameter, with a pressure of around 70 bar.
Compressor stations located at intervals of about 70 km along the transmission
pipelines are used to boost the gas pressure. Finally, the local distribution of gas is
carried out at low pressure through small distribution pipes from 20 mm to 180 mm in
diameter. The gas is then passed to customers through a gas metering system
(www.transco.uk.com). The United Kingdom Onshore Pipeline Operators Association
(UKOPA) provides the view of the UK pipeline Operators on strategic issues related to
safety of pipelines in the UK.
The UKOPA database presents pipeline and product loss incident data from onshore
Major Accident Hazard Pipelines (MAHPs) operated within the UK by National Grid,
Shell UK, BP, Huntsman and Powergen UK, covering operating experience of 654,732
km.yr. (for 21,727 km of pipeline by the end of 2004). The database is designed to
reflect the ways in which the UKOPA operators design, build, operate, inspect and
maintain their pipeline systems.
12
Table 1.2 : Failure Frequency and Causes for Pipeline Failures within UK
Product Loss Cause Percentage
Girth Weld Defect 19
External Interference 22
Internal Corrosion 1
External Corrosion 18
Unknown 5
Other 23
• Internal cracking due to wet towns gas 17
• Pipe-Fitting Welds 2
• Leaking Clamps 1
• Lightning 0.6
• Soil stress 0.6
• Threaded Joint 0.6
• Electric Cable Arc Strike 0.6
Pipe Defect 8
Ground Movement 3
Seam Weld Defect 1
Total (based on 172 incidents) 100
Failure freguencll
Average failure frequency (1962 to 2004 ) per 1000 km. yr. 0.263
Failure frequency for the last 5 years (2000 - 2004) per 1000 km.yr. 0.028
The overall failure frequency over the period 1962 to 2004 is 0.263 incidents per 1000
km.yr .. The failure frequency over the last 5 years (2000 - 2004) is 0.028 incidents
per 1000 km.yr., which is a better representation of the pipeline failure frequency for
13
new generation pipelines. The causes of the failures of the pipelines for the various
incidents as observed are given in Table 1.2.
1.4.2 The EGIG report on gas pipeline incidents
As per 2003 figures, 485 billion cubic meters of gas was supplied to Europe with
62% indigenous production and 38% imports. As per gas supply and consumption
statistics Russia, Algeria, Norway and Netherlands are likely to remain as gas
suppliers for the European countries (Stern, 2002 & Beckervordersandforth, 2004).
The unintentional release of gas from transmission pipeline systems is compiled to

produce a comprehensive database for the European pipeline infrastructure. The
initiative to gather data was started in the year 1982 by six system operators and by
the year 2007 a total of fifteen companies were participating, comprising all of the
major gas transmission system operators in Western Europe. This co-operation of
gas companies is named EGIG (European Gas pipeline Incident data Group).
The EGIG report on gas pipeline incidents is based on the combined operating
experience (3.15 million km. yr. from 1970 to 2007) of fifteen European gas
transmission and system operators. The failure frequency and the distribution of the
causes of the pipeline incidents for EU pipelines are reported in 7th EGIG report (2008)
and is given in Table 1.3.
14
Table 1.3 : Failure frequency and Incident cause for gas pipelines from EGIG database
Incident cause Percentage
Extemal interference 49.6
Construction defecVmaterial failure 16.5
Corrosion 15.4
Ground movement 7.3
Hot-tap made by error 4.6
Other 6.7
Failure freguencl1
Overall failure frequency (1970 to 2004) per 1000 km.yr. 0.37
Failure frequency for the 5 years (2000 - 2004) per 1000 km.yr. 0.17
A total of 1172 incidents during this period gives the failure frequency of 0.37
incidents per 1000 km.yr. The failure frequency for the five years (2000 - 2004) is 0.17
incidents per 1000 km.yr. for a total exposure of 5.7 x 105 km. yr. during this period.
The five year average failure frequency is less than half of the average frequency
between 1970 - 2007, suggesting that improvement in the pipeline technology and
better maintenance practices have had a favourable impact resulting in the reduction
of failure frequency. However, these EGIG values are far greater than the UKOPA
values.
1.4.3 The US Department of Transport (DOT) data
The U.S. natural gas transmission pipeline network consists of a total of 477,012 km.
(298,133 miles) of pipelines consisting of 31 ,782·km. (19,864 miles) of gathering lines,
plus 445,230 km. of transmission lines. These transmission pipelines feed over 2.96
million km. (1.85 million miles) of distribution lines (mains and service lines) operated
by around 2,500 natural gas pipeline operators. The United States currently consumes
15
about 1.8 billion cubic meters (63 billion cubic feet) of natural gas daily (0.65 trillion
cubic meters annually) (OPS, USA).
The office of pipeline safety (DOT, US) has recorded 2042 incidents from the natural
gas transmission operators for the period 1 January 1986 to 31 December 2007. The
data suggests an average of 102.1 incidents per calendar year. If we consider that the
total transmission pipelines. for the US is around 477,000 kms. (US DOT, 2007), then
the failure frequency for these lines are approximately 0.214 incidents per 1000 km.yr.
This value is comparable with the EGIG data. The causes of onshore incidents for
natural gas transmission and gathering systems during the period 1985-2001 is
presented in Table 1.4.
16
Table 1.4 : Failure frequency and various causes for gas transmission pipeline
incidents within the USA
Cause Percentage
External interference (Third-party damage) 28
External corrosion 17
Internal corrosion 11
Natural forces 10
Miscellaneous 2
Incorrect operation 3
Unknown 7
Other failure 4
Construction/installation 6
Manufacturer 6
Previously damaged pipe 4
Malfunction 1
Stress corrosion cracking 2
Vandalism -
Failure freguencll
Overall failure frequency (1986 to 2007) per 1000 km.yr. 0.214
Failure frequency for the last 5 years (469 incidents from 2000 - 2004 0.209
per 1000 km.yr.
(Source: Transportation research board, US)
The pipeline incidents data bases considered in the above analysis show that these
events have a very low frequency. However, they are important for risk assessment,
17
since they may have very large impact (consequences). The next section considers
the low pressure side of the NG infrastructure.
1.4.4 Low pressure pipeline incidents
In risk analyst terminology, these incidents are typically high frequency, but relatively
low consequences event. The incidents considered here are of gas releases from the
distribution network, the meter area, carcass pipe-work (internal) and the home
appliances.
The actual number of low pressure gas escapes are difficult to quantify since this
number is based on 21.7 million gas customers in UK (year 2006) and public reporting
of these gas escape incidents. Moreover a single gas escape is often reported by
numerous customers and presently there is no provision to eliminate multiple calls for
the same gas escape. The National Gird handled nearly 2.3 million emergency calls (in
the year 2006 - 07) on the UK national gas emergency number on behalf of all the
network operators (www.nationalgrid.co.uk).
To arrive at any gas release frequency the pipeline operators often only consider the
mains and the service pipeline failure incidents, since maintenance of these pipelines
are under the control. of the gas distribution companies. The Gas Safety
(Management) Regulations 1996 makes it mandatory for operators to report and
investigate gas escapes that occur on their networks. The HSE's Safety Performance
Indicator report (2005/06) suggest release frequencies for a period 1 April 2005 to
31 March 2006 as presented in Table 1.5.
18
Table 1.5 : Annual gas escapes requiring repair for year 2005/06 within UK
Incident Annual number reported

,
Gas escapes requiring repair by pipeline operator 157,969
Gas In Building (GIB) cases 954
Major accidents (major injury or structural damage) 6
Fatalities 0
(source HSE, 2006)
Only serious incidents of gas escape get analysed as minor smelly leaks are often
stopped as soon as the release is identified, either by the person reporting gas
release or by the service engineer. Such releases are not given great importance and
may not get recorded.
1.5 Impact of hydrogen
In the present work, hydrogen is considered to be injected into the existing natural gas
network and the mixture (natural gas and hydrogen) is considered to be transported
through the channels of transmission, distribution and finally to the end use appliance.
If such a distribution is considered feasible, then it is imperative to consider the impact
of hydrogen as presented in Table 1-1 earlier. Accordingly impact on pipeline failures
is considered in the next section.
1.5.1 Impact on pipeline failures
Any gas coming into contact with a solid can be adsorbed, i.e. it can adhere to the
surface without penetrating the material, or it can also be absorbed, i.e. it can diffuse
into the material and get stored in its bulk. Because of its small size, the hydrogen
atom can rather easily diffuse and accumulate into most structural steels, where it can
cause severe degradation, potentially yielding catastrophic failures.
19
Atomic hydrogen is believed to cause Hydrogen Induced Cracking (HIC) in steel,
leading to blisters in the pipe wall surface, but the impact of hydrogen on the overall
failure frequency needs to be quantified in order to provide meaningful results. Impact
on plastic pipes (polyethylene, PVC, etc.) introduced into the gas distribution system
are unknown. Even if, its believed that these pipelines are not affected by hydrogen,
the smaller molecular size of hydrogen may lead to more leakages from the pipelines
as compared to the steel pipes. The change in risk level because of the introduction of
hydrogen needs to be quantified to help decision makers to evaluate the option of
hydrogen introduction in the present infrastructure or to compare natural gas and
hydrogen pipeline from a safety perspective.
Hydrogen is likely to increase the corrosion rate especially by hydrogen induced

cracking and the fatigue crack growth rate because of the inherent properties of
hydrogen gas. Hydrogen is known for its deteriorating effect on carbon steel, often
referred as hydrogen embrittlement and includes surface cracking, slow crack growth,
loss of ductility, and decrease in fracture resistance. This deterioration can lead to
premature failure, possibly with little warning (Sofronis et. al. 2005). The quantification
of this hydrogen effect, is being undertaken within the Work Packages on Durability
and Integrity (WP3 and WP4) of the NATURALHY project.
Incidents involving the failures of the pipelines as per the UKOPA, EGIG and the
DOT, US suggests following causes:
• External interference (or third party damage) is the single largest cause of
failure contributing around 50 % for the pipelines
• Corrosion is second largest cause of pipeline failures contributing around

15 % to 19 %, which includes internal, external corrosion and stress corrosion
cracking.
• Natural forces such as ground movement I lightning I earthquake etc. and

human error contributes around 17 % for the failures
• Construction defect and material failure contributes around 16 % (EGIG data)
20
Examination of the above data suggests that the introduction of hydrogen into the
pipeline is not likely to modify the causes for human error and natural forces (ground
movement, lightening, earthquake etc.). These suggest that for around 17% of the
causes of pipeline failure (for pipelines carrying natural gas and hydrogen mixture) are
likely to remain the same. It is believed that hydrogen may have an impact on the other
factors contributing to the failure of pipelines.
1.5.2 Impact on ignition probability
Hydrogen has some unique properties compared to other fuels. Some of these
unique characteristics can make it safer than natural gas, while others make it more
hazardous (www.humboldt.edu). Ignition energy, flammability range and leakage rate
can be considered as key parameters influencing the ignition probability in accidental
release conditions. Furthermore, the low viscosity and small molecular size of
hydrogen may give it a greater propensity to leak than natural gas. For a given
pressure and hole size, hydrogen will leak approximately 2.8 times faster than natural
gas on a volumetric basis (if pressure is constant). However, the energy density of
hydrogen is much lower than that of methane; therefore, the energy leakage rate for
hydrogen would be less than that for methane for a given pressure and hole size. As
shown in Table 1.6, the flammability range of hydrogen in air covers a much wider
range of concentrations than for methane, propane, or gasoline vapour. Furthermore, .
it can be much more readily ignited under most circumstances due to its extremely
low ignition energy. Hydrogen is also detonable over a very wide range of
concentrations when confined.
21
Table 1.6 : Characteristics of hydrogen in comparison with other fuels
Properties Hydrogen Methane Propane Gasoline
LFL - UFL a (vol %) 4 - 75 5.3 -15 2.1 - 9.5 1 - 7.8
Stoichiometric concentration
29.5 9.5 4.1 1.8
with air (vol %)
Minimum ignition energy

0.019 0.29 0.26 0.24
(mJ)
LDL - UDL b (vol %) 11/18 - 59 6.3 - 13.5 3.1 - 7 1.1 - 3.3

(source: Schatz research centre)
a LFL _ Lower Flammability Limit b LDL - Lower Detonability Limit

UFL - Upper Flammability Limit UDL - Upper Detonability Limit
The principal hazard foreseen for hydrogen systems is the uncontrolled combustion of
accidentally released hydrogen. Various common physical processes (open flames,
hot surfaces, friction, electrical spark and static discharge) can serve as sources of
ignition for most gas releases. The much wider range of flammability and low
Minimum Ignition Energy (MIE) may result in hydrogen having a higher ignition
probability than other gases/vapours. Increasing dosage of hydrogen to the NG
infrastructure is likely to lower the MIE for the gas mixture. This reduction in MIE for the
gas mixture, could then increase the number of ignition sources with the potential to
provide ignition to the released flammable gas - air mixture. The propensity of these
ignition sources, in regions in which infrastructure handling hydrogen-natural gas
mixtures exist, needs to be considered.
The above characteristics would tend to indicate that there is a greater risk of ignition
for hydrogen-natural gas mixtures. The ignition risk can be quantified for the gas
system through the concept of ignition probability. The substantially lower MIE of
hydrogen (15 times lower than methane) suggests that the ignition probability for a
hydrogen system will be greater than that for a natural gas system.
22
However, these comparisons cannot be so straight forward as they appear. In many'
accidental situations the lower flammable limit (LFL) is more important. The LFL for
hydrogen (4%) is almost similar to that of methane (5%), and about twice that of
propane. In addition, the minimum ignition energy for hydrogen at the LFL is also
similar to that for methane.
The light nature of hydrogen along with its higher diffusivity, differentiate hydrogen
from natural gas. Hydrogen is more diffusive and more buoyant than natural gas,
(methane, ethane, and propane) and therefore tends to disperse more rapidly. For low-
momentum, hydrogen leaks, buoyancy may affect the gas dispersion and mixing more
significantly. For high-momentum leaks, which are more likely in high-pressure
systems, buoyancy effects are less significant, and the direction of the release will
determine the gas motion. Localized air currents due to wind or ventilation will also
affect gas movement. At low concentrations the effect of buoyancy becomes less
significant because the density of the flammable gas -air mixture will be similar to that
of air.
All the above considerations relate to hydrogen. However, there is no information

available concerning methane-hydrogen mixture. The methane-hydrogen mixture will
certainly behave differently than the individual gases, methane or hydrogen. Moreover
the minimum ignition energy for the gas mixture is unknown and no correlation is
available for determination of the ignition energy for various methane-hydrogen
mixtures. Ignition probability, which is believed to be associated with the ignition
energy, needs to be evaluated for the gas mixtures and also for the individual gases
(methane and hydrogen). Hence in the work reported here, the impact of hydrogen on
the ignition probability of methane-hydrogen gas mixtures is studied through
experiments carried out at different ignition energy levels.
1.6 The Thesis structure
In this thesis the MIE, the lowest ignition energy at various concentrations, and the
associated ignition probabilities are considered for the flammable gases under
consideration. The impact of hydrogen on the ignition energy and associated
probability are discussed based on experimental results from methane-hydrogen
23
mixtures. In particular the minimum ignition energy for various gases is identified.
The thesis contents are outlined below.
Chapter 2 summaries the Literature review as carried out for the MIE, ignition energy,
flammability limits and the ignition probability for the pure gases methane and
hydrogen. The past data pertaining to the fire and/or explosion for the gas releases
and the ignition probability models proposed for accidental gas releases are also
discussed.
Chapter 3 summaries the experimental set-up designed and fabricated at

Loughborough University for the measurement of the ignition energy for various gases.
Detailed operating procedures and the relevant safeguards during the experimentation
are also outlined.
Chapter 4 presents the results, observations and the analysis of the experimental data.
The Minimum Ignition Energy (MIE) and ignition probability for all the gases studied
are presented and compared with the available data in the literature.
Chapter 5 presents the conclusions of the experimental study. The experimental

observations and the main findings of the study are presented for methane-air,
hydrogen-air and methane-hydrogen-air mixtures. Ignition energy and the associated
ignition probability for the flammable gases-air mixture are compared and discussed.
Chapter 6 utilises the experimental results to calculate ignition probability for methane
and hydrogen releases from the deterministic analysis of gas releases. The analysis is
based on the experimental studies carried out for natural gas releases (HSE, 1998).
Chapter 7 presents Issues for future work along with proposed modifications
suggested for the experimental facility. Installation of instrument like high speed
camera to the rig, will help determine flame speed, burning velocity, overpressure
wave etc. giving greater insight to the burning process.
24
Chapter 2
Literature Review
Ignition is a precursor to the fire or explosion process. It is of interest to researchers

because of the complex nature of the ignition process and the uncertainties
associated with it. These uncertainties (of ignition/non-ignition of a flammable gas-air
mixture) have been observed in practice during the investigation of incidents
involving gas release (fire and/or explosion event) and also during experimental
attempts to ignite flammable gas-air mixtures. To determine the level of risk
associated with the use of gas in domestic and industrial facilities requires
consideration of the likelihood of ignition. Often historical data is analysed to arrive at
the likelihood of ignition for a gas release. This section reviews available data and
methods for assessing the probability of ignition of a gas-air mixture.
2.1 Ignition theory

Ignition of a combustible gas or vapour - air mixture may occur in two ways. First, the
energy for ignition is supplied by a local source such as a spark or small flame at a
point within the mixture as presented in Figure 2.1 (a). Second, the bulk gas mixture
is heated up to its auto - ignition temperature as shown in Figure 2.1 (b).
Ignition source
(spark / .......,.. auto-ignition
small flame) Q temperature
lal (b)
Figure 2.1 : I.gnition of a flammable cloud (a) by spark (b) by auto-ignition
Combustion of a flammable gas-air mixture occurs, if the composition of the mixture

(cloud shown above) lies in the flammable range and the conditions for ignition exist.
25
The flammable range is bounded by the Lower Flammability Limit (LFL) and the
Upper Flammable Limit (UFL).
Ignition is a process whereby a material capable of reacting exothermally is brought

to a state of rapid combustion with oxygen (usually air). At ambient temperature and
pressure the mixture of flammable gas and air below auto-ignition temperature
(between LFL and UFL concentrations) will not ignite unless a source of energy is
provided. The source of energy could be heat, an electrical spark, another chemical
reaction or just compression (pressure). The external energy breaks the molecular
bonds of the fuel and oxygen, producing radicals. The fuel radical then combines with
oxygen radicals releasing more energy. If this energy is sufficient to break further
bonds then sustained combustion occurs. Although the minimum energy to break the
bonds is known, predicting when ignition will occur under practical circumstances is
difficult. Ignition has also been found to be sensitive to: temperature; fuel and oxygen
concentration; volume of flammable mixture; pressure; area of contact with a heat
source; time period of contact; geometry of mixture; and turbulence of the mixture.
Experiments with fuel-air mixtures give some indication of how easily a particular
mixture in a particular situation will ignite. The relevant characteristics which are
studied experimentally are:
• minimum volume
• minimum energy
• auto-ignition temperature
• ignition lag time
For a given mixture, at a particular temperature and pressure there will be minimum
volume of flammable mixture required to sustain ignition. This is because there is a
balance between the heat generated by the exothermic reaction of the specific
volume of the material burnt and the heat lost which is related to the surface area of
the flame. The minimum volume of the gas required to sustain ignition is where the
heat generated matches the heat lost. The minimum ignition energy is that required
to bring the minimum volume to a temperature that will allow combustion.
26
--------------------- -------- ---------- - ----
If the temperature of a flammable gas-air mixture is raised uniformly, it eventually

reaches a value at which combustion occurs in the bulk gas. Across the range of
flammable concentrations there is a mixture composition that has the lowest
temperature at which the ignition will occur. This is called the Auto Ignition
temperature as depicted in Figure 2.1 (b). Even though, a flammable mixture may be
raised to or above the temperature at which auto ignition occurs, ignition does not
occur instantaneously; there is finite delay. before ignition takes place. This time is
called the ignition delay Ilag time.
2.2 Historical data on ignition of gas releases
2.2.1 High pressure releases
Collated data is available from a number of sources relevant to gas escapes within
the oil and gas industry. These are summarised in Table 2.1 below. Both EGIG and
UKOPA data relate to high pressure transmission pipelines, whereas Offshore
hydrocarbon releases, UK is a database relevant to offshore operations in the UK.
Table 2.1 : Observed ignition incidents of released gases
Data Source Ignition incidents Percentage of Reference source

releases iQnited
UKOPA 9 out of 172 5.2 Advantica, 2005
releases -
Report no. R 8099
Offshore 46 out of 1526 3.0 Offshore hydrocarbon releases
hydrocarbon gas releases statistics and analysis, 2002
releases, UK
HSR 2002 002
EGIG 52 out of 1172 4.4 Report of the European Gas
releases pipeline Incident data Group,
2008
Doc.No.EGIG 08.R.0002
Accidental gas releases from high pressure pipelines (transmission lines) are ignited
either because the flammable gas cloud formed by the release comes into contact
with an ignition source, (sources near the location of release) or the release itself
may cause the generation of an ignition source (such as metal and rocks impacting
r
27
and generating sparks, Copper et. aI., 2004). The tabulated values presented in
Table 2.1, suggest that only a small proportion of high pressure gas releases get
ignited (max. - 5%). A breakdown of these ignited incidents based on the release
size, pinhole (0 to 2 cm), hole (more than 2 cm diameter hole) and rupture « 16 inch
and> 16 inch) was carried out as presented in Table 2.2. Ignition data, as analysed
by other researchers, is also presented in the same table. This shows that larger
releases are more likely to ignite.
Table 2.2 : Pipeline Ignition incidents based on release size
Data source Percentage of releases ignited
leaks 10
World - wide,
Townsend & Fearnehough (1986) ruptures 50
all sizes leaks 16

US Gas, Jones(1986)
ruptures 26
UKOPA pinholes (0 - 2 cm) 4

(Advantica, report no. R 8099, 2005)
holes> 4 cm & ruptures 1.2
pinholes / cracks 4
holes 2
European Gas Pipeline Incident Data
Group (2005) ruptures < 16 in. 10
ruptures> = 16 in. 33
The UKOPA data suggests the opposite observation (incidents involving pinhole
releases result in more ignitions than incidents involving ruptures). This is probably
due to the small number of cases reported in the database.
A review of world-wide pipeline failures carried out by Townsend & Fearnehough

(1986) suggests that the ignition for natural gas leaks from pipelines is approximately
10% for 'leaks and 50% for 'ruptures'. The US data analysed for incidents of Natural
gas transmission and gathering lines during the period 1970 to 1986 by Jones
suggest ignitions for 'leaks' as 16% and for 'ruptures' as 26% (The data for offshore
incidents shown in Table 2.1 have not been analysed for the effect of size as,
28
offshore releases are categorised differently to onshore pipeline releases. Moreover
the sources of ignition for offshore platforms cannot be compared with the onshore
ignition sources).
The variation in likelihood of ignition of pipeline releases arises as a result of the

various data sources used in their derivation. The data comprises of onshore
incidents from different pipe sizes, operating at different pressures, and with different
leak sizes, different weather conditions, different terrain and, most importantly,
variations in the distribution and type of ignition sources near the release at the time
of failure.
2.2.2 Low pressure releases
On a gas distribution network, releases from low pressure pipelines may occur near
buildings and result in gas ingress into the building. Similarly internal gas pipework
could leak and cause a gas accumulation within a building. In both cases, ignition
could lead to fire or explosion. The release rate is governed by the release size (the
size of the hole) and the pipeline pressure above the ambient atmospheric pressure.
Since the gas momentum is negligible (very small) the spreading and mixing are due
to the nature of the gas and ambient weather conditions (and hence ventilation)
prevailing at that time. A Gas In Building (GIB) event is defined (by HSE) as when the
flammable gas concentration has exceeded 20% of the Lower Flammability Limit
(LFL). Using data compiled by the HSE (HSE 2006) for the UK, the annual reported
GIB cases and the number of fire or explosion incidents can be correlated to arrive at
the percentage of GIB incidents which ignited.
29
Table 2.3 : Gas In Buildings events and ignition incidents for the UK
Number of Fire and Reported GIB Percent GIB

Duration
Explosion (F&E) incidents events resulted in F&E
1990-1991 43 1317 3.3
1991 -1992 50 1041 4.8
1992 -1993 35 889 3.9
1993 -1994 47 873 5.4
1994 -1995 35 738 4.7
1995 -1996 42 804 5.2
1996 -1997 40 1165 3.4
1997 -1998 45 973 4.6
1998 -1999 37 923 4
1999 - 2000 56 616 9.1
2000 - 2001 38 597 6.4
2001 - 2002 43 451 9.5
2002 - 2003 33 453 7.3
2003 - 2004 34 971 3.5
2004 - 2005 37 936 4
2005 - 2006 31 954 3.2

(HSE,2006)
The percentage of the cases which ignited is low, partly because only those escapes
which produce flammable concentration can ignite, whereas the GIB data includes
escapes which may produce accumulations with concentrations less than the LFL. It
is also possible that no suitable ignition source was available in some cases or
human intervention prevented further build-up of concentration/stopped the leak
and/or increase the ventilation. This is considered in the following section.
30
-------------
2.3 Sources of Ignition
Stoichiometric fuel-air mixtures of gases, can be ignited with a very small amount of
energy. The energy requirement is only a fraction of a milli-joule, so there are many
potential sources of ignition (see section 2.3.2) for stoichiometric fuel-air mixtures of
gases. Many hydrocarbons used as domestic fuels have minimum ignition energies
of about 0.30 mJ and less (methane 0.29 mJ, propane 0.26 mJ, butane 0.25 mJ).
These energies are extremely low when compared to sources of ignition from
common activities in daily life. An ordinary spark plug has a discharge energy of
25 mJ. The human body, with a capacitance of 200 pF, can be charged with static
electricity to a few kilovolts during normal routine operations, like walking across a
rug, sliding across a car seat, or removing a jacket. A person feels minor discomfort
with a "pin prick" sensation as a conductor is touched and the energy is discharged,
but no harm is done. Such human body discharges are more than 10 mJ. (Crowl
D.A and Louvar J.F. 1990). There are many other potential sources such as
electrical, mechanical, frictional etc. that generate similar energies (> 10 mJ). Hence
many sources routinely available are sufficiently energetic to act as an ignition source
for many stoichiometric mixtures of flammable gases, sensitive dust clouds or
aerosols and air (Eckhoff, 1997).
2.3.1 Common Ignition sources
There can be many potential sources for ignition of flammable gas clouds as
discussed above. These can be divided into various types: heat; compression;
chemical; and electrical. Ignition due to heat includes auto ignition, ignition by hot
gases, hot surfaces, mechanical sparks and thermal radiation. Ignition by open
flames is due to both chemistry and heat. Chemical sources include exothermic
reactions and catalysts. Electrical ignition is due to electrical sparks or arcs and
resistive heating of wires.
A non-exhaustive list of sources of ignition are presented below:
Naked flames, welding and cutting gear, flares, electrical sparks, electrostatic
discharges, exothermic reactions, friction sparks (impact between two similar metals
or dissimilar metals, such as aluminium on iron rust), lightning, hot surfaces, and
internal combustion engines.
31
-------
Crowl et. al (1990) have produced a study on ignition sources based on the analysis
of a data bank of national incidents provided by the Health and Safety Executive. The
study covers a one year period from April 1987 - March 1988 for process plants.
These included a total of 968 incidents presented in Table 2.4, covering a range of
fuels including solids and explosives and various offsite and onsite locations. These
data of ignition sources, is for process plants fire (and/or explosion) accidents and
not just for flammable gases.
Table 2.4 : Common Ignition sources for reported incidents in process plants
Ignition sources Number of incidents Percentage
Unknown 300 31
Flames 237 24.5
Hotwork 120 12.4
Electrical 70 7.2
Hot surfaces 48 - 4.9
Smoking 38 3.9
Friction 36 3.7
Spontaneous ignition 26 2.7
Autoignition 25 2.6
LPG fire equipments 24 2.5
Hot particles 20 2.1
Static electricity 19 2.0
Other 5 0.5
TOTAL 968 100
(Cox et al. 1990)
As seen from Table 2.4, the majority of the ignition sources are UNKNOWN (31%).
Most of the time during accident investigation the source of ignition can not be
positively identified, hence the large number in the Unknown category.
32
The past incidents of major fires in process plants as analysed by Crowl and Louvar
(1990) for'the known sources of ignition are presented in Table 2.5. The ignition
sources are based on 25,000 fires recorded in the Accident Prevention Manual for
Industrial Operations (1974). The sources of ignitions for past incidents are too
numerous to be individuality identified, hence the list of sources considered by Crowl
and Louvar are those with the greatest probability of causing ignitions.
Table 2.5 : Ignition sources for major fire incidents
Ignition sources Percentage
Electrical (wiring of motors) 23
Smoking 18
Friction (bearings on broken parts) 10
Overheating of materials (abnormally high temperatures) 8
Hot surfaces (heat from boilers, lamps etc.) 7
Burner flames (improper use of torches) 7
Combustion sparks (sparks and embers) 5
Spontaneous ignition 4
Cutting and welding (arcs, heat etc.) 4
Exposure (fires jumping in new areas) 3
Arson 3
Mechanical sparks (grinding, crushing etc.) 2
Molten substances (hot spots) 2
Chemical reactions (runaway reactions) 1
Static sparks 1
Lightning 1
Miscellaneous 1
(Crowl and Louvar 1990)
33
The data in Table 2.5 is for process facility fire incidents and not specifically for
flammable gases. The majority of these sources are relevant to flammable gas
releases in open atmosphere, but the relative proportions of each may not be
applicable.
2.3.2 Ignition sources for transmission pipeline release
In the event of gas releases from transmission pipelines, flammable gas may be
present at ground level up to few metres from the release point. Ignition sources with
a potential to cause ignition during such releases are discussed below. Many of the
ignition sources depend on the local conditions and the extent of human activity
nearby. Some ignition sources are closely associated with the cause of gas escape,
for example, lightning may cause a failure and also ignite the release. (Advantica, R
8249,2005)
• Road and rail vehicles
• Electrical sources
• Electrostatic sources
• Tools and machinery
• Lightning
• Flames and hot gases
• Impact generated sparks (rock-rock, rock-metal and metal-metal collision)
• Other sources (pyrophoric dusts, compression heating, electromagnetic

radiation, catalyst, etc.)
Road and Rail vehicles
The combustion engine in the vehicles, is the most significant ignition source.
Starting of the engine is an ignition hazard in any vehicle and may be a common
source as a result of people escaping the gas release. Stalled petrol vehicles in a
flammable gas atmosphere is a significant ignition hazard when the vehicle is
restarted. In diesel engines the strongest ignition source is due to 'over-speeding' of
the engine upon ingestion of the flammable gas-air mixture. Other sources include
exhaust gases, hot surfaces, mechanical sparks, and vehicle electric sparks. Faulty
vehicle wiring can also be an ignition source, if the faulty wiring generates a spark in
34
- - - - - - - -- ----
the flammable gas cloud atmosphere. Engines within excavators and heavy lifting
machines are much larger than in a car and a much greater power is required to start
the engine. The sparks generated from these machines are of greater energy and an
engine of this kind may be in use at an excavation site on or near a pipeline (Copper
et. aI., 2004 and Langan, M.J. & Darby, S. 1998). Indeed the EGIG database suggest
that about 50% of all transmission pipeline failures are due to third party interference,
that is, by digging operations damaging pipelines. If the damage results in immediate
pipeline failure, then the excavator can also provide an ignition source.
Rail transport produces similar hazards, as well as electric sparks from arcing.
Moreover the mechanical sparks from the contact between the wheel and the track
can also act as an ignition source. Sparks from a train was the ignition source of a
large gas cloud formed by a leaking LPG pipeline near UFA, Russia in 1989 and
resulted in a large number of fatalities (Mannan Sam, 2005).
Electrical sources
Street and rail electrical equipment produce sparks capable of gas ignition. These
sparks have higher probability of ignition with electromechanical flashing traffic lights,
when there are several bulbs (Cronin, P. and Smith, B.J., 2000). Overhead power
cables can have ignition potential, if damaged by debris following a pipeline release.
Security fences may also provide a high energy ignition source, if activated at the
same time as a gas release. Damaged telecommunications cables in the ground
near to the point of release may cause an ignition hazard (Copper et. aI., 2004).
Electrostatic sources
Electrostatic discharges are capable of igniting gas-air mixtures. Discharges from

clothing and the human body can act as an ignition source. It is possible that a static
discharge from a person working near a pipeline may ignite a gas release. Debris
(during an incident) impacting on an nearby electrically insulated object can build up
static charge. Spark discharge may also occur to the pipeline or other metal objects,
and so provide an ignition source for the released gas.
35
---------------------------------------------------------------------------------
Tools and machinery
The ignition sources relating to tools and machinery are friction-generated sparks,
hot spots, and hot surfaces. Mechanical sparks caused by friction or impact can have
a high ignition potential depending on the materials involved in their production.
Ignition can result from sparks (hot particles), or from high temperature (hot spots) at
the point of contact. This ignition mechanism should be considered for incidents
which occurred during work on an active pipeline - such as hot tapping or repairs
(Copper et. aI., 2004).
Lightning
Lightning is considered a likely cause of ignition, particularly if the gas release itself is
resulted from a lightning strike. EGIG data suggest a 50% ignition probability from
lightning; when the gas release was caused by lightning. The Ignition potential of
lightning is extremely high for gas ignition but the probability of a lightning strike
coinciding with a separate gas release is very small. Lightning strikes have also
caused punctures (mostly pinholes) in pipelines leading to gas release without any
ignition of the released material. Lightning is a potential ignition source for vent
stacks on a transmission system, because of the continuous nature of the gas
release.
Flames and hot gases
Open flames such as gas fired equipment, burn pits, furnaces and flares can ignite
released gases with certainty. However the likelihood of such a flame being near to a
gas release situation would need to be considered to calculate an ignition probability.
The most common occurrence of open flames is in the pilot lights of balanced flue
gas boilers, which draw in air directly from outdoors. Other, open flames include
bonfires, arson and accidental fires (Spencer, et. al. 1998).
Incident generated sparks
The gas release itself can produce sparks with the potential to cause ignition, these
are termed as release generated sources. Many transmission pipeline failures,
(especially ruptures) have ignited despite the fact that the incidents occurred in
remote rural locations, with no nearby sources of ignition (Copper et. aI., 2004).
,.
36
The tearing of the pipeline (during rupture) may cause sparks. Sparks can also be
generated from debris impacting on nearby objects or on each other. Sparks with the
potential to cause ignition can be generated from rocks impacting against other rocks
or metals, and metal fragments impacting on each other. The nature of rock
(Quartzitic), and the metal condition (rusted steel, aluminium, etc.) can affect the
ignition potential. There are several other metals with similar sparking properties to
aluminium (such as magnesium, titanium, etc.) which can result in spark generation.
It can be considered that the presence of these other metals at the time of pipeline
rupture is unlikely but cannot be discounted.
The high pressure in a pipeline can cause rock, soil and metal debris to be thrown
from the rupture site at very high velocities. The rock and soil type, as well as the
depth of cover is important in determining whether the high velocity debris has an
ability to cause incendive sparks. The controlling factors in impact ignition are the
materials involved, the temperature and the intensity of the sparks. These are
discussed below.
For impacts of metal on metal, pyrophoric sparks are the most likely to cause ignition.
Aluminium, magnesium, zirconium, titanium and cerium are the common pyrophoric
metals. The incendiary nature of the pyrophoric metals arises from their reaction with
atmospheric oxygen. The oxidation temperature generated in the reaction between
the metal and the oxygen causes the spark temperature to rise resulting in ignition of
the flammable gas.
Ignition by non-pyrophoric (carbon, tin, tungsten, zinc, platinum etc.) frictional sparks
is highly unlikely because the temperatures produced by these sparks are
considerably less than the pyrophoric sparks. (Langan, M.J. & Darby, S. 1998)
Aluminium based paints, when used to coat steel, and at temperatures above 150°C,
easily caused ignition when struck lightly with steel. Alloys of magnesium and
aluminium cause ignition on impact with steel, and this occurs most easily at
methane-air equivalence ratios near 0.7. Several other materials will cause ignition
with magnesium alloys, including lead oxide, silica and pyrites, and aluminium alloys
will ignite on impact with lead and ammonium nitrates (Copper et. aI., 2004).
37
"
An investigation of various past incidents of gas ignition relating to impact of debris
(rock and metal) has been made by the HSE, 1969 (then the UK Safety in Mines
Research Establishment). The HSE have studied these impacts further through
laboratory scale experiments. Experiments with rock on metal, metal on metal and
rock on rock, were carried out with rubbing, cutting and impact (Iow velocity, dropping
and ballistic impact) within a flammable methane air atmosphere. The results of the
investigations and experiments are presented in Table 2.6.
Table 2.6 : Summary of ignition between rocks and metal
Investigating material Ignitions in methane air

atmosphere
Sandstone with: -
Sandstone yes
carborundum (Silicon carbide) yes
corundum (aluminium oxide, alumina) yes
Limestone yes'
Shale no
Ironstone no
iron pyrites yes'
Bronze yes'
aluminium alloys yes'
Brass yes'
Steel with: -
sandstone yes'
iron pyrites yes'
iron carbonate yes'
carborundum (Silicon carbide) yes'
corundum (aluminium oxide, alumina) yes'
Carborundum (silicon carbide) with: -
carborundum (silicon carbide) yes
Corundum (aluminium oxide) with:-
corundum (aluminium oxide) I yes
High carbon / Tungsten steel with: -
iron pyrites yes'
Ironstone yes'
q uartzitic rocks yes'
Sandstone yes'
corundum (aluminium oxide, alumina) yes'
Tunqsten Carbide with: -
Steel yes
nickel and monel yes
copper / nickel alloys yes
Brass no
..
('Ignltlons did not occur easily)
38
Experiments with rock - rock impact, suggested that ignition could easily occur from
the rubbing of sandstone with sandstone, carborundum (silicon carbide) and
corundum (aluminium oxide, alumina). Thus for a transmission pipeline passing
through such soil type there can be a greater probability of a release of gas being
ignited.
Other sources (pyrophoric dusts, compression heating, electromagnetic

radiation, catalyst, etc.)
Pyrophoric iron sulphide scale Idusts can act as a potential ignition source for the
released gas (at the point of release) if the gas·stream is entrained with sulphide dust
or if the dust occurs as debris in the gas. In warm, dry conditions the sulphide scale
may glow red and act as a source of ignition. This situation can be of particular
danger in relation to vent stack releases, where it is one of the few sources requiring
consideration. As a safe practice, pyrophoric iron sulphide dust is damped down and
then removed from the equipment. No attempt is made to scrape it before it has been
damped. (Mannan Sam, 2005)
Adiabatic compression of flammable· gas and air results in a rise in temperature. The
resulting temperature rise may increase the mixture temperature to its auto-ignition
range. Ignition is unlikely just because of compression heating alone, but frequent
changes in pressure due to compression or shock waves can elevate the gas
temperature considerably. The heated gas can then be brought to the. auto-ignition.
temperature by other surrounding heat sources. (Mannan Sam, 2005)
Conductive materials (metal and machinery equipment pieces) in the radiation field of
high frequency waves can act as aerials, generating electric currents in the material
through electromagnetic induction. Sparks can be generated from these metal
objects discharging to earth. Thus large military transmitters, strong radio
transmitters or high frequency industrial generators can act as ignition sources ..
(Copper et. al. 2004)
Catalytic converters carry out the oxidation reactions (oxidation of CO and

hydrocarbons) in a vehicle exhaust system. A catalytic converter can serve as an
ignition source for a flammable gas, if it comes into contact with an exhaust system.
39
There are numerous other sources, which each have the potential to ignite a
flammable gas/air mixture. The sources can be very specific to the particular
situation, location or area to be addressed completely within this section. However
the possibility of a gas release being ignited due to these sources cannot be ruled
out. The industrial ignition database of process plant analysed by Crowl and Louvar,
1990 (presented in Table 2.4 and Table 2.5) includes a few of these other sources.
2.3.3 Ignition sources for Low pressure release
If the low pressure releases are outdoors, then many of the ignition sources
discussed for transmission pipelines are still relevant. Low pressure releases from
internal gas pipe work (carcass pipeline) needs consideration of indoor sources.
Similar releases from distribution and service pipes may track through the ground
and result in an accumulation within a building.
Indoor sources are also too numerous to be considered completely in the present
analysis, but the most common are discussed. Ignition sources inside residential
houses, includes pilot lights, cookers, doorbells, clothes dryers, light switches,
electrical tools, hair dryers, toasters, boilers (central and water heating) and other
electrical appliances. The sources can be continuous or intermittent and may require
human intervention (smoking, phones, light switching etc.). or automatic (refrigerator,
freezer, thermostats etc.) The sources are grouped under the following subsections
and discussed below:
• Pilot light / lamp
• Smoking
• Electrical appliances
• Gas boilers
• Static electricity
• Telephones / mobiles
• Cables
40
Pilot light /Iamp
A pilot lighUlamp is certain to cause ignition, if the gas release produces an

accumulation, which can come into contact with the flame (Spencer et. al., 1998).
Some appliances may contain a continuous pilot light whilst others are intermittent.
Gas fired room heating is considered in this category. A cooker having an open flame
is also considered in this category. Cooking equipment was the number one ignition
source for home fires (during the years 1995 to 2001) as per the Ontario office of the
Fire Marshal. (www.ofm.gov.on.ca). A leaking gas ingress inside a room/house may
increase the gas concentration to a flammable level inside a confined space. A
person subjected to the smell of gas for a prolonged period (especially an elderly
person) can become desensitised to the odorant and be unaware of the presence of
the gas accumulation. Operation of a cooker or other open fire sources under such
circumstances may lead to major fire and/or explosion in domestic properties.
Smoking
Smoking materials are one of the most common potential ignition sources. There are
two areas where smoking can pose an ignition threat, these being during the lighting
of the material, and the material when lit. Smouldering tobacco in any form is an
ignition source with a very low potential for a gas release. Tests by Jeffreys et al
(1982) demonstrated that the ignition source related to smoking is, in almost all
instances, due only to lighters or matches, and that burning Cigarettes produce a
negligible risk of ignition. Jeffrey's studies suggested that a lit cigarette cannot ignite
flammable gas, even during inhalation. Matches and lighters are open flames for the
duration of lighting the cigarette and at that time, have an ignition potential of unity.
The probability of a person deliberately lighting a match in a flammable gas
atmosphere is very slim, but cannot be totally ruled out as discussed in the
paragraph above. An electrical cigarette lighter (similar to a car cigarette lighter) is
not an ignition source ..
Electrical appliances
Ignition can be caused by electrical appliances being switched on or off by human

intervention or the switching can be automatic (Swaffield F., 1999). It is the
transmission of charge through a circuit which produces the spark. The possibility of
41
gas ignition from the operation of electrical equipment depends upon the type of
circuit (resistive, inductive and capacitive) involved and the electrical power input.
• For a resistive load, the rate of energy generation is E = 12R joules/s (where I
is the current in ampere and R is the resistance in ohm). This energy will heat
up the conducting material and the ambient atmosphere and thus ignition can
be caused by the hot surface. Ignition would require a long time period of the
hot wire in the presence of a flammable gas mixture, and the quantities which
determine the likelihood of ignition from this source are the total heating time,
the total energy flow (defining the current strength) and the core heating
temperature.
• For an inductive load, a low voltage inductive break spark energy is

2
E = O.5L1 , joules (where L is the inductance of the circuit in henry and I is the
current in ampere). The energy can be compared with the minimum ignition
energy of the flammable gas, to result in ignition.
• For a capacitive load, the spark energy is E = O.5CP·V2 joules, and discussed
in detail in Chapter 3, since the experimental apparatus used in the
experimental work described in this thesis is based on this principle.
Sparks may be caused by faults in a damaged electrical circuit, or occur in normally

working circuits containing components such as switches or relays. Sparks are likely
to occur with mains switches, whereas arcing will occur with dirty contacts or
damaged circuits.
Testing was undertaken by Advantica (Swaffield F., 1999 and Johnson D.M. & Wright
S.J., 1989), on various kinds of electrical equipment to determine their ability to ignite
a natural gas-air mixture. With some pieces of equipment, the electrical circuits within
them were broken down, so that the individual components most likely to cause an
incendive spark were tested. This was done at the higher end of the range of
voltage/current ratings used in that equipment.
Theoretically, both mains operated and battery operated equipment can produce
sparks of sufficient energy to be intrinsically hazardous, but testing suggested most
42
--------- ------ - - - -
devices were unable to cause ignition of a flammable gas air mixture (within the
flammability limits). The reason is, the spark locations are normally sealed from the
atmosphere preventing any ingress of the flammable gas-air mixture. Even if the gas
penetrates the seal, to the ignition source, there may not be a route out through
which the flame could propagate to the whole cloud.
Devices such .as vacuum cleaners, electrical drills, doorbells, automatic door entry
system and most motors (inductive loads) are capable of ignition since the spark
energy is greater than the minimum ignition energy of gas. Ignition of a gas cloud is
plausible, if the casing, or the circuitry itself, is damaged, although no records have
been found for testing of these items.
Automatic operation devices, (where mains power is being switched on and off,
without human intervention) like circulation pumps,. thermostats, timer switches,
Video recorders, refrigerators etc. have sufficient energy to act as ignition sources.
Inside buildings, previous incident ignitions have been attributed to thermostats,
timers and refrigerators or freezers with automatic switches.
Experiments with a refrigerator (Swaffield F., 1999) resulted in ignition within a small
number of cycles, and confirmed the findings of several explosion incident
investigations that found that the only likely electrical source of ignition was a
refrigerator. Newer designs are less likely to be a source, as the electrical elements
are encapsulated, whereas older designs may have open elements.
Burglar alarms and smoke alarms produce a very weak spark and will not result in
ignition because of the small current (Swaffield F., 1999). Standard non rechargeable
battery systems do not result in ignition even with a worst-case direct short-circuit
across the terminals on the highest common operating voltage.
Static electricity
Static electricity is the build up of electrical charge that is unable to flow to earth.
Static electricity has been attributed as the cause for many explosions where no
43
other cause was apparent. Static electricity discharges commonly occur from the
body and from clothing. The discharges are generally very weak, and do not cause
discomfort to human beings. Any form of rubbing activity or friction may charge the
human body or the tools and equipments used for pipe work. A charged body/object
has the potential to cause a spark discharge resulting in ignition (Copper et. al.,
2004). The energy of a normal human body spark is much stronger than the
minimum energy level of 0.29 mJ required for the ignition of a methane gas-air
mixture. Typical examples of household activities that result in a build-up of a static
charge are walking across a rug, placing different materials in a tumble dryer,
removing a sweater and combing hair. Clinging fabric, and sometimes audible sparks
are the result of the build-up of a static charge. Fabrics are also capable of producing
spark discharges with a likely potential for ignition under conditions of low humidity.
At higher humidity the discharge is unlikely since clothing will usually contain enough
moisture to be sufficiently conducting so as to prevent static accumulation.
Gas boiler
A boiler containing a strong ignition system is used for central/water heating in most
households with mains gas. Gas boilers use ambient air, and if there is ingress of a
gas - air mixture into the appliance, ignition is certain. Depending on the path of the
burning flame and the presence of a flame arrestor, etc., boilers can act as an
ignition source for the flammable releases (Spencer, et. al. 2004).
Telephones
Telephones, both mobile and fixed, are one of the most widespread pieces of
electrical equipment, and in the case of a gas leaks (smell), are often used by people
calling for help. Use of mobile telephones is forbidden at petrol/gas stations across
the world, because its widely believed that these can be a source of ignition.
However testing has shown that ignition from telephones in any kind of operation is
highly unlikely. Experiments with mobile phones, in the worst-case of direct short
circuit across the battery terminals have never resulted in ignition. (Johnson D.M. and
Wright S.J., 1989).
44
Cables
Damaged or faulty wiring can cause ignition of released gas. Increased electrical
load on a cable can also heat up the cable resulting in damage to the cable and
hence provide a potential source of ignition. Short-circuiting could also be an ignition
source for a gas release (Copper et. aI., 2004).
Ignition sources discussed above were further analysed for ranking, based on
experimental studies, observations and engineering judgement. Ranking can be
done using the characterisation of ignition sources described in the next section.
2.3.4 Analysis of Ignition sources
Ignition sources can be analysed further for the certainty with which the ignition of a
flammable gas is possible by that source. Brilton (1992) provides an example of the
ranking of ignition sources based on the consideration of their available energy in
relation to the minimum ignition energy required for various flammable gas or dust
clouds. During ignition all likely ignition sources are initially considered, which can
deliver energies greater than the minimum required for the ignition of a flammable
gas. A list of energies of various types of sources and the kind of flammable gasl
dust air mixture which could be ignited at that energy level is shown in Table 2.7.
45
Table 2.7 : Energies for various ignition sources
Source Energy (mJ) Typical substance which

can be ignited
flames, chemical sources, large hot 100 -1000 Methylene chloride,

spots Ammonia
human body spark, bulking brush 10 -100 Lycopodium

limit (discharge during powder filling -
silo filling)
brush discharge (single electrode 1 -10 Acetone

discharge with no sharp tip - in a
strong electric field)
mechanical sparks, stray current 0.1 -1 Methane, methanol

sparks, small hot spots
discharges from textiles, weak 0.01 - 0.1 Ethylene, hydrogen

inductive coupling, weak radio-
frequency pick -up
(Britton 1992)
A semi - qualitative approach was used by Jeffreys et. al (1982) to rank the sources
in terms of ignition potential. The ignition potential of strong, medium and weak were
defined as described below in Table 2.8. The ignition potential is based on the
experiments carried out for a 7% methane-air mixture. The ignition sources, such as
pilot lights, gas fired heaters, open flames etc. warrants a further category of
'certain', with an ignition potential of unity. Experimental studies can be further used
to define a category for those items with 'negligible' ignition potential like, radio
frequency waves. A few sources along with their ignition potential are presented in
Table 2.8.
46
Table 2.8 : Ignition potential for various sources
Source Ignition potential
Pilot lights I gas fired heaters 1
Cigarette lighter (open flame) 1
Switches (electrical) S
Electrical tools , S
Hair dryers M
Toasters M
Televisions W
Electrical Appliances (washing mic, iron, bulbs etc) W
Smouldering Ilit cigarette 0
Domestic and mobile phones 0
S = strong ignition potential with probability of ignition greater than 0.5

M = Medium ignition potential with probability of ignition between 0.05 and 0.5
W = Weak ignition potential with probability of ignition less than 0.05
2.4 Ignition model for gas releases
The strength of the ignition source (its delivered energy), spatial distribution of the
sources and the operation (continuous or intermittent) of the source have major
impact on the likelihood of ignition of the released gas. Accordingly, ignition models
for gas release developed by various researchers consider these parameters.
Ignition likelihood of the released gas, often referred to as ignition probability in the
literature is now described in the subsequent sections.
Expert judgement over the years has produced estimates for gas ignitions (ignition
probability) based on a single value or a few bands as presented in Table 2.9.
47
Table 2.9 : Estimates for Ignition probability by various researchers
Source Type of release Leak quantity / Location with/ Ignition

size of release Area Probability
Browning 'no' sources of

LPG massive 0.1
(1969) ignition
Kletz LPG 10 ton or more Onsite area 0.5

(1977)
Polyethylene 1 in
Small plant area
VCE 10,000
Hydrogen and
hydrocarbons mix small to major plant area 1 in 30
(hot, @ 250bar)
HSE LNG vapour general 'no' sources of 0.1

clouds ignition .
(1981 )
'very few' sources of 0.2
ignition
'few' sources of 0.5

ignition
'many' sources of 0.9

ignition
Blything LPG Large (gas Onsite area 0.7

& dispersion
Reeves distance to LFL
(1988) cone. is 60 m)
There is wide variation in the values of ignition probabilities given by the various
researchers tabulated above. Many of the values are case specific and may not be
directly applicable to general studies. A review was carried out for the proposed
ignition likelihood/ignition probability models in the open literature. The
correlations/models and ignition probability estimates discussed in subsequent
sections are
• Cox, Lees and Ang model
• PIPESAFE - Ignition probability model for transmission pipeline
• WS Atkins model
• Ignition probability estimates for low pressure releases
48
2.4.1 Cox, Lees and Ang model
This model (Cox A.W. Lees F.P. and Ang. M.L. 1990) is based on the historical data
for leak freqiJencies and fire and explosion frequencies. The approach is to make an
estimate of the frequency of leaks, with the standard plant equipment inventory. The
model has a simple relationship between the ignition probability and the release rate.
Cox et. al. (1990) reviewed a number of estimates of the probability of ignition of
various flammable releases, based on incident data as presented in Table 2.10.
Table 2.10 : Ignition probability estimates for flammable releases
Flammable gas release rate Ignition Probability
minor ( < 1 kg/s ) 0.01
major ( 1 - 50 kg/s ) 0.07
massive ( > 50 kg/s ) 0.3
Cox further assumed that the probability of ignition is proportional to a power of the
released gas mass flow rate. The correlation for the probability of ignition (IP) based
on the flammable releases analysed by Cox et al (1990) is
IP =a mb ... (2.1)
Where m is the mass flow rate of leaking gas in (kg/s)
a = 0.017 and b = 0.74
Cox also proposed different coefficients for equation (2.1) above, when there is some
control on ignition sources (a= 0.006, b =0.77), no control on ignition sources (a =
0.074, b =0.57) and also for self ignition (a =0.003, b =0.28).
The model is not intended for very high mass release rates, where it gives a
probability of greater than 1. The constant of proportionality and the power are
estimated from the observed historical data of ignition probability and the released
mass of the flammable gas under consideration. As seen the model is very general
49
and does not consider individual ignition source characteristics contributing to the
ignition process.
2.4.2 PIPESAFE - Ignition probability model for transmission pipeline
Advantica have produced an ignition probability model that is incorporated into their
PIPESAFE package (Robinson,.C. & Ronnie, S., 2005 and Advantica R 8224,2005).
The model is intended for transmission pipelines only and is based on the historical
experience of pipelines operations. A list of 68 transmission pipeline incidents has
been analysed by Advantica. Amongst the analysed incidents, where cause of failure
was reasonably ascertained; 15 were caused by external interference, 18 by a defect
in construction or repair, and 13 were due to corrosion, ground movement or other
natural causes. A further analysis based on the ignition delay time (time between the
gas release incident and the actual ignition event), suggested that the majority of the
releases ignited immediately (within one minute). The maximum time from the start of
the gas release to ignition recorded in this data was 90 minutes.
The model proposes a relationship of ignition probability values based on the pipeline
diameter and pressure. Considering the complexity of the processes involved in
pipeline release incidents, there are numerous other parameters affecting the ignition
probability, but the data analysed for rupture incidents by. Advantica suggest that
ignition probability increases with line pressure and pipe diameter. Accordingly the
equation utilised for calculation of ignition probability for ruptures in PIPESAFE
package is
IP (ruptures) = pd 2 ... (2.2)
Where p is the pipeline operating gauge pressure (bar)
d is the pipeline diameter (metre)
The relationship incorporated in the PIPESAFE package is used up to a pd 2 value of

48.2 bar m2 ; above this value a limiting ignition probability of 0.8 is used.
50
The same approach is used for puncture releases, except that the hole size is used
instead of pipeline diameter. The pd 2 value is halved since there is single source for
a puncture release. Unlike the two sources contributing to a gas release from a
rupture event. The equation utilised for puncture releases in the package is
IP (punctures) = O.5*pd h2 ... (2.3)
Where p is the pipeline operating gauge pressure (bar)
dh is the puncture hole diameter (metre)
2.4.3 WS Atkins model
The W,S Atkins model for determination of ignition probability for a drifting gas and/or
growing flammable gas cloud was developed by Spencer Hand Rew P.J. 1997 on
behalf of the Health and Safety Executive. (HSE eRR 146 / 1997).
The HSE report reviews ignition data and models for use in onshore risk assessment.
The model takes account of the likely ignition sources in the area and the duration
that the gas/vapour cloud, within flammable limits, interacts with these sources. Each
source is characterised with an ignition potential defined as the certainty with which
the source can ignite the flammable gas in an open atmosphere. Further, the
number of each type of ignition source per unit area, and its rate of activity
(continuous or intermittent) is also considered in the model.
The parameters that are used to define each ignition source in the model are
summarised below:
• Different ignition sources were identified and the density per unit area (f.i), of
each source is determined
• The probability of ignition or ignition potential (po), is the basic probability that
ignition will occur when the source is active and in contact with the gas, i.e.
source is surrounded by a flammable gas cloud. The ignition potential (po) value
has to be defined for each ignition source
51
• Ignition sources as per their operation, intermittent or continuous, are then
considered in the analysis. The period for which the source is 'active', ta, and the
period between each activation, ti, is specified. From these ta, and ti, values the
frequency, A, at which the source is able to ignite the gas, and the probability of
the source being active, a, is estimated
• Intermittent sources are considered as a special type of the generalised source

with ta =0, and thus a =O. Continuous sources are a special case with t; =0, and
thus a = 1 and A =0
Table 2.11 : Source parameters for the WS Atkins model
Source parameter Intermittent sources Continuous sources
Time period for which the source is

active (ta,) min
ta =0 ta =continuous
Time period between each activation of
source (ti,) min
t; =finite t; =0
Probability of the source being
a =0 a =1
active (a)
Frequency of activation of the source

(able to ignite the gas) A per min
A =finite A =0
In reality the sources can never be completely Continuous or Intermittent, hence all
real World sources have some finite ta, t;, a and A values. The overall ignition
probability after time 't' is calculated based on the following summation of the
probability
52
IP(t) = 1- Q(t) ... (2.4)
and
Q(t)= IeXp{A* ,uJI-ajPoj~-,!jPOjl -ID ... (2.5)

j=!
Where:
IP(t) overall probability of ignition from various sources at time t
Q(t) probability of non-ignition at time t
A the area of the region being in contact with the gas cloud
represents the 'j' th. ignition source ( when 'j = 1, 2, 3, ... to J' nos. of sources)
,uj ignition sources density per unit area of source 'j'
aj probability of the source 'j' being active
Aj frequency of activation of the source 'j' (per min.)
pOj ignition potential (probability of ignition) of single source 'j' in consideration
t the time duration the source has been in contact with the gas cloud (min.)
Simple correlations for the probability of ignition in terms of the cloud area for
different conditions (day, night, industrial, urban, rural etc.) can be developed using
the model.
The WS Atkins model presents an area based method for the ignition of flammable
gas clouds for offsite areas. The method does not include a model for the dispersion
of the gas. Separate applications of the method are needed for each vapour cloud
scenario! wind direction etc. The ignition source type, potential and its distribution is
difficult to estimate because of numerous ignition sources.
2.4.4 Ignition probability estimate for low pressure releases
Although originally developed to study large outdoor vapour clouds, the WS Atkins
model was utilised for estimating the ignition probability within a residential area
53
(house/room). Low pressure releases could occur from the internal pipe work
(carcass pipe). gas appliances or any external gas ingress into the building. It is
assumed that the gas leak has caused enough gas accumulation to exceed the LFL
concentration inside the building. Thus the probability of ignition is computed given
that an ignition source is in contact with the flammable gas cloud within the building.
A typical list of ten ignition sources with their source characteristics (as described in
the Atkins model) are presented in Table 2.12. The ignition sources include pilot light
(gas fire. candle. cooker etc.). gas lighter. electrical switch. fridge/freezer. gas boiler.
Microwave. Washing m/c. TV set. and Room heater. Some of the sources are human
operated and others are automatic as per the instrument settings. The values of
ignition potential (certainty with which the source can ignite the flammable gas) is
assumed based on engineering judgement.
Table 2.12 : Typical Ignition sources within homes
Source characteristic parameters
Sources No.s ta t A x 1000 a po
min min per min - -

Pilot light (candle.
cooker. etc.) 1 5 595 1.67 0.01 1
gas lighter /
cigarette lighter 1 1 479 2.08 0.002 1
gas boiler 1 10 110 8.33 0.08 1
Electrical switch 5 0 180 5.56 0 0.5
Electric room heater 1 30 570 1.67 0.09 0.2
Toaster 1 3 717 1.39 0.004 0.2
Fridge - freezer 1 20 100 8.33 0.13 0.05
Microwave 1 6 294 3.33 0.02 0.05
Washing mic 1 60 1380 0.69 0.04 0.05
TV set 1 180 600 1.28 0.23 0.05
54
The probability of ignition based on the activation of sources as presented above is
computed from Equation 2.5 presented by the WS Atkins model. The only
modification to the equation is instead of area (A) and the source density (f.J ) the
total number of similar ignition sources (N,) is used as follows.
N, = A * f.J
IP(t)=l- ±exp{N![(l-ajPO)-.!JPOjl-l~ ... (2.6)

j=1
Where:
IP(t) Overall probability of ignition from various sources at time t
N! Nos. of similar'j' th. sources (i.e. electrical switches = 5 nos.)
aj probability of the source 'j' being active
Aj frequency of activation of the source 'j' (per min.)
pOj ignition potential (probability of ignition) of single source 'j' in consideration
t the time duration the source is in contact with gas cloud (min.)
The ignition probability under the activation of ignition sources was estimated at
various time durations (t) and the variation as a function of time is presented in
Figure 2.2.
55
1.2
1 - - - - - - - - - - - - - -.~-~-::..--::..:..-~-
- ~- =-=--=-=-- ---------t
-~-
~ 0.8 ,
------ ----------------------------------------------------
:5
~Q. 0.6
!5
:2
" 0.4
E.>
O~--------~--------_r--------~----------~------~
o 100 200 300 400 500
TIme (min.)
Figure 2.2 : Ignition probability variation with time for gas leak inside house
As observed the ignition probability increases with time and attends a value of unity
after 300 min (5 hrs.) Le. Ignition is almost certain after 5 hrs. The computed ignition
probability·is based on activation of the 10 ignition sources and at the rates assumed
in Table 2.12.
2.4.5 Mode/limitation
The ignition energy of a flammable gas (methane, propane, butane, hydrogen, etc.)
and air mixture varies with its concentration. Therefore it is expected that the ignition
potential (as defined in the WS Atkins model) should also vary with concentration.
There is very limited information on the individual gas - air mixture ignition energies
at various concentrations covering the entire flammability limits. Most of the gas
ignition information gives the minimum ignition energy (without associated·
probability) at stoichiometric or near stoichiometric concentrations. Also, due to local
fluctuations in concentration within a flammable gas cloud or plume, an ignition
source within a cloud may experience a range of concentrations, hence the ignition
potential (po) of each source will vary with position within the flammable region. The
ignition potential of individual sources is the key for ignition of the gas cloud. The
ignition probability thus will vary greatly with variation in ignition potential of each
56
source. Hence experiments to determine the ignition energy (and associated
probability values) were planned within this project not only for the stoichiometric air-
gas mixture but also over the entire flammable ranges of methane (5% to 15%),
hydrogen (4% to 75%) and methane-hydrogen mixtures. The ignition of flammable
gas-air mixtures is a function of the minimum ignition energy of the gas. Hence the
issues considered in the next section is the MIE values for the gases methane and
hydrogen.
2.5 The Minimum Ignition Energy of gases
For a flammable gas (within its flammability limits) to be ignited the most important
criteria which must be met is the requirement of the minimum energy to start and
~ustain the combustion of the gas. It therefore is reasonable to describe the
ignitability of the mixture in terms of the limiting value of the energy, which just ignites
the mixture. The energy liberated by the source must be sufficiently large to stimulate
a propagating flame reaction into the gas volume, in spite of the energy losses to its
surroundings (unburnt gases).
The energy requirements are different for different gases and it is also different for
the same gas if the gas concentration is different. For a given gas, if the composition
of the gas-air mixture is varied, the curve of the limiting energy necessary for ignition
exhibits a distinct minimum. This minimum value is designated the Minimum Ignition
Energy (M lE). Gases often exhibit their MIE very close to the stiochometric
concentration. Hence MIE for most of the gases are reported at or near the
stiochometric concentration. The energy required for ignition of gases is
progressively larger on either side of the concentration at which the ignition energy is
a minimum until the Lower and Upper Flammability Limits are reached. (Lewis & Von
Elbe 1987 and Esseghir M. & Polymeropoulos C.E., 1988). The requirement of
ignition energy becomes infinitely large at the limiting concentration (LFL or UFL) of
gases.
The energy delivered by a source and the energy received by the gas mixture during
the ignition process can be different quantities. Hence the MIE is defined as the
minimum value of the energy (stored energy delivered by the source), which on
discharge, just ignites the mixture of quiescent gas (gas at rest) in the most ignitable
57
composition. MIE is an extremely important property for safety standards as well as
for the fundamental understanding of the ignition process of combustible mixtures.
Data on MIE's are necessary for fuel ignition systems and safety standards in relation
to possible explosion hazards in fuel handling and in industrial situations.
MIE's for most of the flammable hydrocarbon gases are very small (0.3 mJ and less).
The measurement of such a small magnitude of energy is difficult. Spark, Laser and
energy delivered by hot wire (by passing electric current through wire) are a few of
the techniques used to provide such a low magnitude of energy (Martin Hattwing,
2004 and Lewis 1987). A few experiments of igniting flammable gases (methane,
hydrogen etc.) in air have been carried out using frictional sparks from light alloys
(Komai et. aI., 1994). The aim of these experiments was to examine the incendivity
of the sparks and to understand the safety aspect of handling flammable and
explosive gases. Spark ignition through electrical discharge has been the preferred
and standard method for determining the MIE, ASTM E 582 1988.
2.5.1 MIE determination
The electrical discharge used during an experiment can be measured to estimate the
quantum of energy delivered. Since the amount of energy delivered by these sources
is often very small, measurement of these energies is only possible with advances in
electrical measurement devices (Huang et. aI., 2007 and Ono et. al 2007). The
electrical devices either measure the voltage or the amount of total charge
transferred during the discharge to estimate the energy delivered by the specified
spark process.
Electrical spark discharges can be deliberately generated in a flammable mixture and

the various parameters measured to arrive at the energy discharged and thus enable
the MIE to be estimated. The standard way to characterize capacitive discharge
sparks is in terms of the quantity of stored electrical energy, measured in (mJ). The
actual amount of energy that is deposited in the gas by the discharge is lower and is
some fraction of the stored electrical energy measured through electrical devices.
58
Beginning with the work of the US Bureau of Mines in the 1940s and continued by
Blanc et al. (1947), an extensive series of spark tests (Lewis and van Elbe, 1961)
were carried out to determine the ignition energy in hydrocarbon-air vapours.
Extensive experiments were carried using methane, the major constituent of natural
gas. Ignition energy was found to be a function of the concentration of methane in the
methane-air mixture. The ignition energy was found to be a U-shaped function of
concentration with the vertical portions of the "U" occurring at the flammability limits
and the bottom of the "u" at some intermediate composition. The ignition energy for
other hydrocarbon fuels also follows a similar trend to that of methane gas. The U.S.
Bureau of Mines tests were aimed at understanding the ignition characteristics of
methane and its energy. The objective was to prevent any chance of methane
explosion occurring due to sparks generated during the cutting of coal with mining
equipment.
The experimental apparatus for the determination of the MIE essentially consisted of
electrodes with a spark gap, a charging capacitor, a charging resistor and a high
voltage source. The spark was produced in a vessel filled with a known concentration
of gas. Once the gas mixture was prepared inside the vessel, the voltage was slowly
raised across the electrode gap through the charging resistor. At the breakdown of
the dielectric field strength of the gases between the electrode tips, a spark is
observed. Theoretically, a spark discharge in a gas occurs when the free electrons
that are present in gases are accelerated in an electric field to such a high kinetic
energy that the gas molecule is ionised in the collision. (Martin Glor I Martin Hattwing,
2004). After this ionisation an avalanche of ions or electrons is created, which then
depending on the field strength and, in accordance with the laws of physics, leads to
a gas discharge. The dielectric field strength depends on the type of gas and the gas
pressure. The dielectric field strength increases proportionally with the gas pressure,
starting from the initial value of about 3 kV/mm at normal ambient pressure.
Lewis and van Elbe 1987, (including Blanc, 1947 and the US Bureau of Mines, 1940)
have used similar experimental setups for spark ignition studies. The apparatus was
designed to produce sparks with a capacitance of 100 pF or more. The gases were
admitted to a test bomb which was made of stainless steel and had an inside
diameter of 5 inches. The spark electrodes were mounted in the centre of the bomb.
The electrodes were adjusted through a micrometer to precisely measure the spark
59
- - - - - -
gap distance. The electrodes were connected to a system of fixed and variable
condensers and a variable range calibrated voltmeter. The electrical circuit was
designed to minimise the electrical resistance. The aggregate capacitance was
varied between about 100 pF to 5000 pF. The exact value of the capacitance was
measured through a Wein bridge. A high voltage was supplied via a resistor to the
electrode - capacitor system. A rotary charger with a small metal sphere mounted on
either end of hard rubber alternately touched the tenminals of the power source and
the spark electrode transferring charge during each contaci of the metal spheres to
the electrodes. Thus the spark circuit was effectively isolated from the power source.
The voltage at which the spark occurred was observed and the actual capacitance in
the circuit was noted. If the mixture did not ignite, the capacitance was increased and
the experiment repeated. The value of voltage and capacitance at the ignition
threshold was used to calculate the ignition energy as.E = 0.5*CP*V2 (where
capacitance, CP (F) and Voltage, V (V)). The lowest ignition energy value obtained
through this setup at the most ignitable composition was reported as the MIE.
Similar setups have been used by various researchers to arrive at the MIE of various
gases.
2.5.2 MIE and quenching distance
Lewis and von Elbe 1987 have carried out experiments with glass flanged electrodes.
The spark electrodes during the experiments were tipped metal rods, and in addition,
were also flanged by glass plates. Glass flanges were used so that the spark occurs
at the centre (between metal rods). Several series of experiments were carried out in
which the composition, pressure and temperature of the flammable gases were held
constant and the length of the spark gap was varied with a micrometer assembly.
Lewis observed that the minimum ignition energy with glass flanged electrodes takes
a sharp vertical turn (increases very rapidly) when the distance between electrodes is
reduced below a certain critical value. The glass flanges have the effect of
suppressing the ignition (initiated at the centre by metal electrodes) altogether, when
the electrodes were within a critical distance. This critical distance was named the
quenching distance. The quenching distance was observed to be independent of the
mode of ignition. The same or nearly the same value of quenching distance was
observed from experiments in which the flammable gas was enclosed in a
rectangular channel bounded by two plane parallel plates, and ignited at one point by
a pilot flame. The materials used for the walls also did not have any effect on the
60
quenching distance. Glass and metals were equally effective as heat sinks, even
though the heat conductivity of metals far exceeds that of glass. The reason being
the heat conductivity of solids exceeds heat conductivity of gases by an order of
magnitude. The use of flanges made from an electric non-conductor (glass) assured
that the sparks remained centred between the electrodes. A plot of minimum ignition
energy and the electrode (flanged electrodes) distance, showed a distinct minimum
before a very steep rise in minimum ignition energy at very small electrode distances.
At electrode distances larger than the quenching distance the size and shape of the
electrodes did not affect the value of the minimum ignition energy. (Lewis and von
Elbe 1987). At the bottom of this minimum ignition energy versus electrode distance
curve, the energy values depend primarily on the mixture variables, notably the
composition and the pressure. The quenching distance at the stoichiometric
concentration along with the MIE values of methane and hydrogen is presented in
Table 2.13.
Table 2.13 : MIE and quenching distance for flammable gases
Gases Gas concentration (% by vol.) MIE (mJ) Quenching distance (mm)
Methane 9.5% 0.33 2.16
Hydrogen 30% 0.019 0.64
The above values suggest that there may be a relationship between the minimum
ignition energy and the quenching distance. Kondo et. al 2003 has suggested two
theoretical expressions for calculating the minimum ignition energy of gases from the
'quenching distance; one is based on the amount of energy that the minimal incipient
flame contains, and the other is based on the heat loss from the surface of the
minimal flame.
In the first equation Kondo et. al 2003 assumed a certain minimum energy at a spot
in time is required to ignite a flammable gas mixture. The amount of energy was
equated with that required to increase a certain minimal sphere of the mixture to the
flame temperature. The diameter of the sphere was approximated to be the
quenching distance and the following equation was obtained.
61
... (2.7)
where,
qd quenching distance (cm)
Pb molar density (mol/cm 3 ) of the burnt gas at the temperature Tb
Tb flame temperature (K)
To unburnt gas temperature (room temperature).
SPH" represents the molar heat capacity (J/moI/K) at constant pressure averaged
for the temperature range To to Tb
SPH,,(Tb - TJ represents an amount of enthalpy required to heat up one mole of

gas from To to Tb
In the second equation, Kondo et. al 2003 assumed, the outward growth of flame
depended on the balance between the heat of reaction and the heat loss from the
surface of the sphere. The minimum ignition energy was estimated from the heat
losses from the surface of the sphere, the average laminar burning velocity, and the
temperatures To and Tb as follows
MIE =7r qd' A,,(Tb - To) ... (2.8)

Sa,
where,
qd quenching distance (cm)
A" heat conductivity (W/cm/K) averaged for the temperature rangeTo to Tb
Tb flame temperature (K)
To unburnt gas temperature (room temperature).
S" laminar burning velocity (cm/s) averaged for temperature range To to Tb
Since there is a large temperature difference between both sides of the flame front Tu
and Tb, the value of heat conductivity (A,,) was averaged by Kondo et. ai, for the
temperature range. On the other hand, laminar burning velocity (S,,) was used to
derive the time for the gas to pass through the width of flame front to complete the
combustion reaction. The averaged value for the temperature range between To to
Tb is also used here. The data for the quenching distance and burning velocity at
--62
room temperature used for the calculation as well as the data of minimum ignition
energy was taken from NACA (now NASA) report 1957. The flame temperature was
calculated using the adiabatic flame temperatures for gases. The heat capacity and
coefficient of heat conductivity were from the thermo-physical properties of gases
and used to obtain equations to estimate approximate values at higher temperatures.
MIE as estimated by Kondo et. al 2003 through the proposed equations are
presented in Table 2.14, against the actual values of the MIE as observed through
experiments.
Table 2.14 : Theoretical estimates of MIE for gases
Gases Gas conc. MIE observed MIE calculated MIE calculated

(% by vol.) (experiments) mJ (eq. 2.7) mJ (eq. 2.8) mJ
Methane 9.5% 0.33 3.33 2.18
Hydrogen 30% 0.019 0.04 0.04
The expressions (equations 2.7 and 2.8) of minimum ignition energy calculation by
Kondo et. ai, 2003 from the quenching distance do not compare well with the actual
MIE values as observed through experiments especially for methane. This suggests
that the burning process is so complex that simple theories for the prediction of
ignition energy can not quantify the MIE. Hence researchers have relied on
experimental values of MIE for flammable gases.
2.5.3 Historical MIE data
A literature search has been undertaken on the reported minimum ignition energies
(MIE) of the flammable gases under consideration, that is, methane and hydrogen.
Since the majority of accidental gas releases and subsequent ignition happen in the
open atmosphere, the literature search was restricted to MIE values reported under
ambient conditions of temperature and pressure and for quiescent mixtures.
63
Table 2.15, lists the MIE data as found in reports, standards and as given by various
researchers that focused on the subject of fire and explosion. The reported values
are based on the ignitions carried out using electrical sparks generated across an
electrode gap charged to a high voltage. The stored capacitor energy at the time of
the spark is considered as the MIE for the gas-air mixture. The experimentally
observed values are at the stoichiometric fuel-air ratio, which can be, practically,
considered as the minimum value of ignition energy.
Table 2.15: Minimum spark ignition energy data from various sources
Reference source MIE (CH.) mJ MIE (H 2 ) mJ
Toriyama, Y. and Saito, S., 1942 0.67 -0.58 -

Calcote et. aI., 1952 0.47 0.028
NACA report, 1957 (now NASA) 0.33 0.02
US Bureau of Mines explosive research report, 1966 0.30 0.017
Sayers et. aI., 1970 0.55 -
Moorhouse, J., 1974 0.63· -

Lewis and von Elbe, 1987 0.29 0.02
Lees F.P., 1996 0.29 0.019
Pratt T.H., 1997 0.21 0.016
• Ignition probability of 80 % is assigned to this energy
The historical MIE data, were determined using measured values of Capacitance and
Voltage at the time of the spark. Also, the technology of electrical measurement has
vastly improved over the period, hence the reduction in the MIE values for more
recent studies as shown in Table 2.15. The criteria of MIE by Moorhouse, 1974 is for
the 80% ignition probability of methane, hence it is likely to be greater than the MIE
reported by other researchers/sources.
64
2.5.4 Limitations of MIE apparatus
MIE determination (historically and to date) has been carried out through the
capacitive spark discharge. The charge built up was with a gradual transfer of high
voltage to the electrode-capacitor system. A rotary charger with small metal spheres
mounted on either end of hard rubber is used to slowly increase the charge by
alternately touching the terminals of the power source and the spark electrode,
transferring charge during each contact of the metal spheres to the electrodes.
This mechanism allows high voltage to be present on the electrodes for a

considerable longer duration of time prior to the discharge. The long duration of high
voltage permits the condition of a corona discharge. The high voltage prior to the
spark can also ionise the gas, disturbing the actual dielectric strength of the gas
mixture. This in turn affects the voltage at which a spark occurs and has a direct
impact on the calculation of the energy. The spark inside the gas chamber occurred
after a certain time lag, which was observed by Lewis to be considerably long.
Radium capsules of various strengths were placed in the bomb to reduce the time lag
(Lewis et. aI1987). The effect of radium capsules on the MIE values are unknown. In
later work the apparatus was modified to replace the rotary charger with a resistor
rod, of the order of 10" 0 which served the same purpose of slowly transferring
charge from the power unit to the spark electrodes. The resistor rod was moistened
with glycerine or conversely (by rubbing the rod with a dry cloth), to accelerate or
retard the rate of charge transfer (leakage). These activities helped in adjusting the
charging process but required a lot of trial and error. Hence this method cannot be
termed as a standard method for the determination of MIE.
2.6 Ignition energy and ignition probability for gases

Historically when the MIE's for gases were evaluated the primary goal was in relation
to safety standards applicable to the industries. The standards were based on
threshold energy limits for released gases, or for the coal mining industry. The aim
was to prevent fire and explosion. Hence the MIE was equated to the minimum
observed energy value resulting from a series of ignition tests carried out in a test
apparatus.
65
Most researchers only report a single value of MIE for a particular gas. However,
due to local fluctuations within a nominally uniform mixture, ignition of a gas mixture
with a given energy level is not guaranteed even if the energy exceeds the MIE.
Therefore, there is an associated probability of ignition for a given energy level and
given gas concentration. In Table 2.15, Moorhouse, 1974 provides details of the
associated probability (80%). Lewis and von Elbe reported MIE values when one
ignition occurs in a hundred tests for the specified energy level. (i.e. MIE with an
ignition probability of 1%). Komai et. aI., (1994) has used a criteria of 50 % ignition
probability for the experiments investigating the ignition of a methane - air mixture by
frictional sparks. Kono et al. (1977) has determined MIE for three lean concentrations
of propane in air. The minimum ignition energies were based on a 50 percent ignition
frequency criterion. Parker (1985), using pulsed sparks, determined MIE, for 2.7
percent propane in air with an ignition probability of 10 percent. So the MIE values
are reported at different values of ignition probability. To have uniformity in the
energy values, it is necessary to associate ignition probability values to the ignition
energy values.
Although data is available in the literature for the MIE of methane and for hydrogen,
only a limited range of concentrations are normally examined and reported. There is
strong variation in ignition energy as the lean limit is approached (Shepherd, J.E., et.
al 2000). Moreover the ignition energies of methane-hydrogen-air gas mixtures are
not available. Hence there is a need to determine these ignition energies and
associated ignition probabilities experimentally.
Accordingly, an experimental apparatus was designed, primarily based on the

capacitive spark discharge. The apparatus is similar to the Lewis and von Elbe
apparatus with certain modifications to overcome the limitations outlined in section
2.5.4. The electric circuit was based on the spark ignition of dust clouds as proposed
by Randeberg 2006. The entire experimental setup, its design and the operating
procedure is presented in the next chapter.
66
Chapter 3
Design and Operation of the Experimental facility
3.1 Introduction
The experiments reported in this thesis to study the ignition of the methane-
hydrogen-air system were carried out in a facility, designed and built at the
Department of Chemical Engineering, Loughborough University. Prior to actual
experimental runs, the equipment was thoroughly checked. All the gas pipelines, gas
regulators, manifold assembly and cylindrical vessel were pressure tested with soap
solution at each joint, to eliminate leaks. A pressure holding test (with dry air) at
200 kPa pressure was carried out. The entire experimental setup was also vacuum
tested at the lowest achievable vacuum (using a vacuum pump) to demonstrate that
there was no ingress of any outside gases during experimental runs.
3.2 Conceptual design

The experimental facility comprised of: a cylindrical combustion chamber; a gas
delivery system; an electrical circuit for generating a spark to ignite the gas-air
mixture; and a data acquisition system. The cylindrical combustion chamber was
equipped with instruments to measure the pressure and temperature within the
chamber and also the potential difference across the electrodes at the time the spark
was generated (i.e. the electrode voltage). A schematic of the experimental facility is
presented in Figure 3.1.
The key components of the experimental setup are
• Gas supply, delivery and evacuation system
• Cylindrical combustion chamber
• Electrical circuit for ignition of the gas -air mixture
• Instrumentation
• Data acquisition system
• Safety systems
67
The experimental programme involved numerous explosions of flammable gas - air
mixtures formed inside the combustion chamber. The flammable gas consisted of
methane, hydrogen and mixture of methane and hydrogen. Fire and explosion
hazards were present as a consequence of the highly flammable nature of the gases.
In addition, the generation of a very high voltage (15 kV) across the spark electrodes
made the experimental setup vulnerable to electrical hazards. Hence, safety
interlocks were provided to mitigate the risks and these are discussed separately in
Section 3.6 (Safety systems).
3.3 Gas supply, delivery and evacuation system
The gas supply and delivery system for the experimental facility comprised of three
gas cylinders (methane, hydrogen and dry air) each at an initial gauge pressure of
20 MPa (200 bar - when the cylinders were full). These gas cylinders were secured
in position with belts to prevent them being disturbed. Speciality gases two-stage
regulator valves were fitted to each gas cylinder to deliver the gases to the cylindrical
combustion chamber. Gases from BOC (Brin's Oxygen Company, Ltd.) were utilized
in this work. The Cylindrical combustion chamber was evacuated with an Edwards
vacuum pump.
The gas cylinder regulator valve was opened and the delivery pressures for the
flammable gases were set at a gauge pressure of approximately 50 kPa (- 7.5 psi).
Limiting the maximum pressure upstream of the combustion chamber gave better
control when introducing the flammable gases, since the gas introduction process
was manually controlled through the operation of manifold valves (V1 to V3 -
Figure 3.1). The delivery pressure for the dry air cylinder was set at a gauge pressure
just above 100 kPa (- 15 psi), since air was the last gas to be introduced into the
combustion chamber and the final pressure within the combustion chamber prior to
ignition was required to be approximately 1 atm.
Firstly, the vessel was evacuated using a vacuum pump. Once the gas cylinder
delivery pressure was set to the desired values the gases were admitted one at a
time and the desired mixture obtained through the method of partial pressures. The
flammable gas (methane and/or hydrogen) was fed to the vessel by opening the
appropriate manifold valve. Each manifold valve was completely closed once the
68
desired amount of methane (or hydrogen) was added to the vessel. The first gas fed
to the chamber occupied the entire volume of the vessel and the actual amount of the
gas was computed by the data acquisition system. For experiments with a mixture of
flammable gases (methane and hydrogen), the desired quantity of the second
flammable gas (either methane or hydrogen) was injected by opening the second
manifold valve. The amount of the second flammable gas added to the vessel was
also computed by the data acquisition system. The air inlet valve on the manifold was
slowly opened, so as to increase the pressure inside the vessel to atmospheric. The
pressure transducer reading P1 attached to the manifold was recorded to determine
the vacuum achieved and after the addition of each of the component gases
(including dry air), until near atmospheric pressure conditions were achieved. At the
end of the gas filling process, the gas filler pipe was removed from the vessel by
disconnecting the coupling (CV4) so that the explosion vessel was isolated from the
feed gas system. This action also isolated the pressure transducer (P1) from the
overpressure wave generated after gas ignition, thus preventing it from being
damaged.
After the introduction of each gas and air (the last component), a settling period of 1
to 2 minutes was always provided before the data acquisition system was initiated to
record the values of the parameters. The turbulence (eddies) generated due to gases
introduction dissipates within this time, as observed through the stabilised pressure
transducer readings (P1 and P2) and a quiescent well mixed gas-air mixture was
achieved. Thus the gases conditions inside the experimental vessel were essentially
quiescent and the experimental results reported relate to the ignition of a quiescent
fuel gas-air mixture formed within the cylindrical combustion chamber.
69
Data
------------------------------,
,..---- acquisition
system ------------~ T1
-$-
Electronic Extra High
Cylindrical , Spark
Compon- Tension
combustion ent Unit (EHT) pulse
I
electrodes
chamber (ECU) generator
CV4 I ~ , -e-
P1 Manifold
V1~ , vi
'L
_L
LL
r-
,
V~
--
7
~
I
I
Two way
Valve
vs',
,
~
\15
..
-
Purged
gases to
atmos.
D
Storage
Oscilloscope
1 2 3 vacuum}
pump
1 - Dry air cylinder
Vent to
2 - Hydrogen cylinder
atmosphere
'-- '-- '-- 3 - Methane cylinder
Figure 3.1 : Schematics of the experimental setup

3.4 Cylindrical combustion chamber
The cylindrical combustion chamber was fabricated from a 152 mm (six inch)
diameter. seamless, (schedule 40) stainless steel pipe of length 232 mm.
(Figure 3.2). Flanges were welded on either side of the pipe and polycarbonate
plates were used as windows to observe the onset of ignition and the subsequent
burning process. Stainless steel flanges covering either side of the pipe were
specially fabricated (see Appendix C, Figure C.1) to hold the polycarbonate plate in
place with minimum compressive stress. The windows retaining flange was secured
to the vessel flange using bolts. The calculations on the required thickness of the
vessel shell and windows (incorporating a safety factor) for the expected pressure
rise inside the vessel is presented in Appendix C.
transducer
Temperature
thermocouple
Electrodes
Feed gases Exhaust gases
Retaining (Dimensions are given in

flanges Appendix C, Figure C.1)
Figure 3.2 : Schematic representation of the cylindrical combustion chamber
The cylindrical vessel had a total of six tappings, one each for gas feed and exhaust,
two spark electrode insertion tappings diametrically opposite, one tapping for a
thermocouple insertion and one tapping for a pressure transducer. As described in
Section 3.2, the gases were introduced to the vessel through a manifold device after
evacuation, using the vacuum pump to the lowest steady state pressure value
achievable. The pressure transducer attached on the gas manifold (P1 - Figure 3.1)
displayed the dynamic/transient pressure of the introduced gas. Steady state
71
pressure values were recorded at vacuum condition and after the introduction of
each gas component to the vessel.
A second pressure transducer (P2 - Figure 3.1) mounted on the vessel was used for
recording the pressure rise inside the vessel after the ignition. The transducer had a
pressure range from near complete vacuum conditions to 930 kPa (- 135 psi). The
sole purpose of having this pressure transducer in place was to check and ascertain
the ignition of gases inside the vessel. A fast response time of 2 ms for this
transducer made sure that the smallest overpressure wave generated through the
burning process was captured, even though at times it was not possible to observe a
flame through the viewing window.
A thermocouple was used to measure and record the gas temperature. The gas
mixture inside the cylindrical vessel was prepared at ambient temperature. The
temperature recording was initiated only after ignition of the vessel contents (after the
onset of the spark). The dynamic temperature rise of the thermocouple sensor was
recorded. The only purpose for temperature measuremenVrecording was to
determine the occurrence of ignition inside the vessel. This was an additional
instrument to the pressure transducer (P2 - Figure 3.1) mentioned above to ascertain
ignition of the vessel contents.
The spark electrodes were 2 mm diameter copper rods. The copper rods were tightly
housed in a solid teflon rod 10 mm outside diameter to provide the necessary support
to the copper electrodes and also to act as electrical insulators from the stainless
steel vessel. Details of the electrode assembly inside the vessel is shown in
Figure 3.3. Approximately 10 mm of bare copper rod protruded from the teflon
housing near the centre of the vessel. The tips of the copper rods were produced,
initially with a 60° taper angle, but in order to reduce the generation of a corona, the
electrode end tips were slightly rounded off. The electrode gap was set at 2 mm
during the ignition energy measurement experiments performed inside the vessel.
After every 4 to 5 successful ignition attempts of the vessel contents, the electrode
tips (copper tips) were cleaned. This action prevented the deposition of
combustion/oxidation products on the copper electrodes. The condition of the
electrode tips was also observed through the polycarbonate windows and
accordingly the decision to clean the electrodes was initiated.
72
4--- 155 mm Dia. --~
10 mm Dia.
2 mm
Gap
Figure 3.3 : The electrode assembly
3.5 The Capacitive spark discharge circuit

The electrical circuit for the capacitive spark generation and energy measurement is
shown in Figure 3.4. The spark was generated by the release of electrical charge
stored in a capacitor when the voltage across the electrode tips overcame the
dielectric strength of the gaseous atmosphere. The stored energy for a capacitor with
capacitance GP, charged to voltage V is given by
E = 0.5 • GP • V 2 ..........•..•..•••••••..•...•.•..•..•.•.••............ (3.1)
Where GP Capacitance (F)

V Charged voltage of capacitor (V)
E Stored energy in the capacitor (J)
Since the energy required for gas ignition is very small, the value of E is reported as
mJ rather than J. In the present study the stored energy in the capacitor was used to
create an electrical discharge which defines the ignition energy. The ignition energy
tests follows the practice started by Lewis and von Elbe (1987) of reporting the stored
energy rather than measuring the energy discharged into the spark. The circuit for
measuring ignition energy for the gases, is based on standards BS EN 13821 : 2002
(Potentially explosive atmospheres - Explosion prevention and protection -
Determination of minimum ignition energy of dusUair mixtures) and ASTM E582
1988 (Standard Test Method for Minimum Ignition Energy and Quenching Distance in
Gaseous Mixtures)
73
Discharge Capacitor
Primary HV Charging
.--_c_a_pa_c,ito_r-t ~former resistor Electrodes
Vessel
HV probe
Curren
t sense
resistor
Oscilloscope
TDS - 3034B
Rectified Trigger Pulse

AC mains
CH1 CH2
Figure 3.4 : Simplified electrical circuit for Capacitive spark discharge
The ignition energy for gases (methane or hydrogen) at stoichiometric conditions is

very low, hence an electrical circuit capable of producing sparks of very low energies
was utilised in the present study. A similar circuit was utilised by Randeberg et. at
(2006), to study the minimum ignition energy of (Iow energy) dust clouds, in the
below 1 mJ region. Accordingly, the present spark generator was designed based on
the feed of 230 V AC mains. This spark generator circuit is similar to the ordinary car
induction coil, where instead of the 12 V DC (battery) supply, the circuit was modified
to have a rectified 230 V AC mains signal (see Appendix F, Figure F.1 for the rectifier
circuit). Safety interlocks were added to this circuit as a consequence of the
flammable nature (explosion potential) of the gases and the handling high voltage,
and to protect the low pressure transducer.
The circuit operates in two steps
1. Generation of Extra High Tension (EHT) I high voltage pulse through the
EHT pulse generator.
2. Capacitor discharge to the spark in the gas air mixture, charged by this EHT
pulse.
A trigger pulse, as shown in Figure 3.4 was fed to a thyristor. The triggering of the
thyristor discharges the primary capacitor of 1 IJF (initially charged to the rectified AC
74
mains voltage) through the primary coil of the HV transformer. This induces a high
voltage pulse with amplitude of around 15 kV on the secondary coil side. The high
voltage/EHT pulse was fed through a charging resistor, causing a voltage build-up on
the discharge capacitor, placed downstream of the charging resistor. Discharge of
the discharge capacitor occurred when the voltage across the electrode gap (2 mm)
reached the breakdown voltage (6 kV for a 2 mm electrode gap in air). Specific
combinations of the resistor and discharge capacitor, resulted in different energies
being produced as presented in Table 3.1. Hence, different spark energies could be
studied starting from the lowest energy (smallest capacitor) to the maximum energy
achievable (largest capacitor) from the electrical circuit.
Table 3.1 : Energy delivered by Capacitor
Discharge Discharge Charging Nominal Energy

Capacitor Resistance time (liS) Breakdown (milli Joules)
(pFarads) (k Ohms) Voltage(kVolts)
10 1000 10 6.0 0.18
50 100 5 6.0 0.9
100 100 10 6.0 1.8
150 100 15 6.0 2.7
200 100 20 6.0 3.6
300 100 30 6.0 5.4
The choice of charging resistor depends on the size of the discharge capacitor, the
. aim being to avoid recharging of the discharge capacitor during the duration of the
discharge. This was achieved by choosing a charging resistor that gave a theoretical
time. constant R*CP (where R is the charging resistance in ohms and CP the
discharge capacitance in farads) of at least 1 lis, which ensured insignificant
recharging of the discharge capacitor during the spark discharge (normally nano
seconds). During experiments, the time constant R*CP was always kept above the
theoretical value of 1 IJs. However, if the charging resistor was too large (i.e. large
time constant and long charging time), the spark gap voltage would not reach
breakdown during the lifetime of the feed from the EHT pulse generator.
75
A system for the measurement of the spark voltage through a high voltage probe (HV
probe) was included in the discharge circuit, enabling measurement of the spark
energy according to Equation 3.1. The high voltage probe had a capacitance of
3.0 pF. This 3.0 pF capacitance adds to the discharge capacitance in parallel, and
was taken into account when estimating the total capacitance involved in the
discharge and also during the estimation of the spark energy. The voltage signal was
captured, and the maximum voltage observed was recorded as the peak breakdown
voltage. The signal was captured on the oscilloscope for a total duration of 400 ns by
adjusting the scale of the oscilloscope.
The fact that the current flow through the charging resistor was negligible during the
duration of the discharge is very important for evaluating the spark energy. If the
charging resistance (R ) is too small, the energy input to the spark process continues
to increase until there is appreciable voltage on the secondary side of the HV
transformer. Hence, during the experiments it was always assured that the charging
time of the discharge capacitor (R*CP) was at least 5 ~s (> 1 ~s.). This assured that
the assumption of a "pure" capacitive discharge from the electrical circuit was
reasonably correct.
The capacitive spark discharge circuit is very compact, and without switches which
can introduce stray capacitance and additional energy. Moreover, the time of spark
discharge was much more precisely determined because of the initiation of a trigger
pulse. Lewis and von Elbe (1987) used a static high-voltage source to slowly raise
the capacitor voltage until breakdown and this required the use of very large resistors
during the charging process. In addition, the gas mixture was exposed to high
voltage for a much longer time until the actual spark occurred. This provided ample
time for corona discharge through the gas medium.
3.6 Instrumentation
3.6.1 Pressure Transducers
Two pressure transducers (PX 219 - 30V15G5V and PX 219 - 30V135G5V, Omega
Solid state pressure transducers) were used to measure the gas pressure at two
different locations in the present study. The first pressure transducer P1 (PX 219 -
76
30V15G5V) generates an output from 0 to 5 V DC for a total gauge pressure range of
2.048 MPa (-14.7 psig to 15 psig). The second pressure transducer P2 (PX 219 -
30V135G5V) also generates a similar output from 0 to 5 V DC for a total gauge
pressure range 10.321 MPa (-14.7 psig to 135 psig). The calibration data from the
National Institute of Standards Technology, United States for these transducer was
utilised to convert the output voltage (0 - 5 V DC) to the pressure (kPa) value. The
calibration graphs for both these transducers are presented in Appendix G, Figure
G.1 and G.2.
The first transducer (P1 -. Figure 3.1) was fixed to the manifold to measure the
pressure from vacuum to near atmospheric (101 kPa) pressure. The pressure
recordings from this transducer (P1) was utilised to estimate the gas composition
(concentration of flammable gas in the gas-air mixture). The second pressure
transducer P2 (PX 219 - 30V135G5V) was used to check/ascertain the ignition of the
gas-air mixture.
3.6.2 Temperature thermocouple
A Chromel/Alumel (Ni Cr +ve and Ni AI-ve), K - type, RS Thermocouple (maximum

continuous temperature range of 11 OOoK) was used to record the gas temperature
inside the cylindrical combustion chamber after ignition or non-ignition. The
temperature recording was initiated only after the onset of the spark. The only
purpose for the temperature measuremenUrecording was to check and ensure the
occurrence of ignition inside the vessel. This was an additional instrument to the
pressure transducer (P2) mentioned above. This double check on the buming
process was useful during experiments with lean concentrations of hydrogen, since
the flame was invisible.
3.6.3 Oscilloscope
A Tektronix Oscilloscope model no. TDS 3034 S, was used for data acquisition. The
oscilloscope has a bandwidth of 300 MHz, a maximum sampling rate of 2.5 giga
samples per second (GS/s) and four channels. Channel 1 was used to record the
breakdown voltage by means of high voltage Tektronix probe (P 6015 A). Channel 2
was used to record the current during the spark process across the current sense
77
resistor by means of probes available with the oscilloscope. (Figure 3.4). The
parameters were recorded using the single shot acquisition. (single sequence
button). The scope was set to capture the spark voltage for a total duration of 400 ns.
The spark discharge process was complete during this time as observed from the
voltage trace signal shown in Figure 3.5.
8
Max. Voltage:: 6.2 kV
~
ll> 0
J!o
>
-4 ------- ---------------------------------------
-8+---------~----------._--------_r--------~
o 100 200 300 Time (ns) 400
Figure 3.5 : Voltage trace for the spark discharge captured on the scope
3.7 Data acquisition (LabVIEW interface)

National Instruments LabVIEW version 8.0 was used as the programming software
tool for measurement (partial pressure of gases), automation of the spark generation,
and simultaneous recording of transient temperature and pressure as shown in
Figure 3.6. The experimental procedure was semi-automatic where the ignition
process and subsequent data recordings were automatic and the preparation of the
gas-air mixture (valve operation) was done manually. Human intervention was
required during, vacuum creation, introduction of the gases (valves operation) and
combustion products expulsion. The LabVIEW program was used for the automation
process and data recording. Programming was done to record the dynamic pressure
and temperature values at a click of the control button. Data recording could be either
continuous, from the time of the click of the button, or just recording a single value
parameter on the click of the button. Accordingly the control buttons were given
specific modes as required for the recording of each particular parameter.
78
The pressure value from transducer P1 was utilised to record the initial pressure
(vacuum reading) and also after addition of each of the component gases (methane
and/or hydrogen and dry air). The signal from pressure transducer P1 was generated
at an interval of 0.001 s. At this rate, there was constant fluctuation in the pressure
value, and hence an arithmetic mean of 1000 readil)gs was recorded as the pressure
value. The control button (AM to File) was selected such that only one Arithmetic
Mean pressure value was recorded at the click of the control button "AM to File". All
the pressure values - transducer P1 readings (vacuum reading and after addition of
each component gases including air) were recorded to the LabVIEW file PP.LMV.
The readings from file PP.LMV were used to calculate the flammable gas
concentration during each of the experimental runs.
Another control button "Fire & Record" was used to perform three tasks
simultaneously. The electrical circuit required a small trigger pulse to activate spark
generation (generating high voltage across the electrodes inside the gas chamber).
It was observed that the electrical trigger was required for a stipulated duration of
1.5 s., to activate spark generation. The use of a timing function (in LabVIEW
programming) coupled with the control button "Fire & Record" helped to deliver the
trigger pulse for 1.5 s as required for the electrical circuit. The duration of this
electrical trigger is shown as "Preset for firing" in the Figure 3.6.
Simultaneously, on activation of the trigger pulse, the temperature and pressure

recordings were started. The temperature (thermocouple T1) and pressure
(transducer P2) readings were recorded in separate files (ouUemp.LMV and
out_press.LMV respectively) for a specified duration of 5 s. through the use of
another timing function coupled with the same control button "Fire & Record". The
duration of 5 s. was chosen since the thermocouple (T1) and the pressure transducer
(P2) were able to record the peak values during this stipulated time. The signals from
the thermocouple T1 were at an interval of 0.0833 s. Hence temperature values were
recorded in "ouUemp.LMV" file at this time interval, starting from the click of "Fire &
Record" button for a total time duration of 5 s.
The rate of data acquisition for pressure transducer P2 was at an interval of 0.001 s.
and was recorded in file "outyress.LMV'. The total duration for pressure recording
was also equal to 5 s (same as the temperature recording) but the number of data
79
points obtained for the pressure value was far greater than for temperature. It was
found that the pressure readings defined the success of ignition more clearly than the
temperature readings. Appropriate digital values from LabVIEW files generated
during each experimental run were recorded in a separate MS excel sheet and then
all LabVIEW files were deleted, so that new files generated during tlie execution of
the LabVIEW program carried the same name.
All the LabVIEW files and the control buttons are shown in
Figure 3.6. The block diagrams for the LabVIEW program are illustrated in,
Appendix F, Figure F.2.
80
I , 1[011.>:;1
l':II1LJ. ~5:S5
19/12/200715:55
19/1Z/2OO115:S5
1(,1 . ...•., . ! I>

'l (';' I mzj
1101.11'1,
11 F~) -~, t~. 'i:"'!'W ~sert ~", ~005. ~~ 1'1"""'~.

_ B xl
:
. F3' -..:,,:n<\ ;
A 8 ( 0 E F GIH[
101.19 De,. ' ...... 1. I .•
,~OP~~.i
I.p~ n<~:\lc07~ ,or:, ,F ir:ms' 1&0::v~~I~(:l~~1~:."
Q, . . '. d.1M,.,' ! .• ro-'98!;! . •
1118.104j sa.! [ioo"K-'1; "'7' [1601"
1101.967:
-vac;/press.trots)
IIII-'''-I---J·_·_--!- -" "r-I-
?,f)CIZ.<B" 11I"!~-I···········:··II:" i···· ;."""

l. -~ -)J.-;, •..'-. ".'.'.'.'.•'•.'.
1---- --." .. ············1····
i;! ........ " ' , '
. .···i
..
. . . . . ·· i·············
! ......
_L ..........
I······ i········. · ..····
o 5
_ ..'P'
Figure 3,6 : LabVIEW (front panel) with data file and readings
3.8 Safety Systems
The experiments involved handling of flammable gases (methane, hydrogen and

mixtures of methane and hydrogen) and carrying out confined explosions inside the
vessel. A Risk Assessment (RA) of the entire experiment was carried out. As a result
of the RA study, various safety systems were included during the design of the rig.
These systems are further discussed below.
Accidental release of flammable gases from the rig cannot be ruled out, hence Xgard
(type 5) gas detectors (manufactured by Crowcon) for each of the flammable gases
(methane and hydrogen) was installed just above the gas cylinders. An alarm
sounded if a gas concentration of 20% of the lower flammability limit (LFL) was
detected. If the flammable gas concentration exceeded 40% LFL then air ventilation
for the laboratory area was activated.
The filling process to the vessel was manual, through a manifold system. The gas
filler pipe feeding gas to the vessel acted as a link between the ignition vessel and
the gas cylinders and was attached by means of a coupling (CV4 -Figure 3.1). This
coupling CV4 was always detached and locked in a stored position before attempting
any ignition. An interlock prevented use of the spark· generation circuit if the coupling
was not locked in the stored position. This action also isolated the pressure
transducer (P1) thereby protecting it from the overpressure wave generated after
ignition.
Prior to experimental runs commencing, all the gas pipelines, and connecting
equipments were pressure tested so that leaks were eliminated. A pressure holding
test at 200 kPa pressure was also carried out to identify any minor leaks. The entire
experimental setup was also vacuum tested to detect ingress of gases.
The electrical circuit was provided with a safety interlock 'key switch'. The electrical
circuit could only produce a high voltage (spark) after inserting the key and turning it
ON. An interlock was also provided on the Electronic Component Unit (ECU) box,
housing the capacitor and resistors. The electrical circuit was designed such that
82
opening of the ECU box discharged any voltage on the secondary side of the HV
transformer. A spark could not be achieved, if this box was not closed and latched. A
"Monitor Point" and an earth probe was provided to discharge any residual charge
manually before replacing/changing electrical components.
Prior to actual experimental runs commencing, a written Risk Assessment Record,

Electrical Assessment Record, and COSHH forms were submitted to the competent
authorities. All these risk assessment forms are presented in Appendix 0 of this
thesis. A written detailed operation and shutdown procedure was approved and
followed during the experiments. Experiments were performed strictly following the
laid down procedure.
3,9 Experimental procedure
Prior to conducting a series of experiments the distance between electrode tips was
adjusted to 2 mm, so that the spark was able to jump across the electrode gap and
the generator could produce sufficient voltage to break the dielectric strength
(3 kV/mm) of the air. A confirmatory test was performed to obtain consistent sparks
for the combination of largest capacitors (200 pF and 300 pF) with a charging
resistance of 100 kO in the circuit. This was necessary, since larger capacitors have
higher time constant 'R-CP' (Table 3.1) and hence requires more time to reach the
breakdown voltage. The generator provides the high voltage pulse only for a
specified duration. The experimental procedure is based on the flowchart presented
in Figure 3.7.
83
I .
Start I
Ensure following systems are ON
1. Gas supply to manifold
2. Spark generation (EHT) system
.
:.
~, ,
Evacuate vessel completely using vacuum pump
I
I •
Record absolute rig pressure
I
•
Introduce desired quantity of flammable gas
and record steady state pressure
•
Introduce dry air and record steady state pressure
(around atmospheric). Estimate gas concentration
•
Detach the gas filler pipe from the vessel, and lock it in position
.~
I
Initiate spark (& record energy) by activating EHT generator I
y
Gas Ignited
Increase spark, energy using larger capacitor
N~ y
,. ~
I Stop (End series of experiments)
Figure 3. T: Flowchart for the experiment
84
3.9.1 Gas mixture preparation
During the initial start-up for an experimental run, the combustion chamber was
purged with dry air for at least 120 s. to drive out water vapourslflammable gases
accumulated from the last experiment. For each of the subsequent experiments in
the run, air purging was done with atmospheric ambient air, after evacuation of the
vessel, so that combustion products (C0 2 , CO, unburned flammables etc.) did not
accumulate within the vessel chamber.
The regulators (two stage regulators - valves) on the gas cylinders required for the
experiments were opened sufficiently to deliver gas to the cylindrical combustion
chamber. The flammable gases were set at a delivery pressure of around 50 kPa
(maximum 10 psig) and the dry air was set at a delivery pressure of around 100 kPa
(maximum 15 psig). This action gave good control for the introduction of flammable
gases into the cylindrical combustion chamber. This was necessary because the gas
introduction process was manually controlled through the operation of the manifold
valves.
Data acquisition (pressure and temperature recordings) and triggering of the spark
process (EHT pulse generation) was carried out using the LabVIEW program
specially prepared for the operation of the rig. The major steps for the operation of
the vessel are outlined with an explanation for the necessary relevant actions.
The cylindrical combustion chamber was evacuated completely using the Edwards
vacuum pump (HP no. G1099 - 80024) to achieve a condition of near complete
vacuum (absolute pressure below 2 kPa). The vacuum pump was required to be run
for 3 to 5 minutes. A steady state pressure reading as displayed by the pressure
transducer (P1) was recorded and stored as a digital value (file : PP.LMV) on
actuation of the LabVIEW front panel display button "AM to File" through a click of the
mouse. This recording was done after switching off the vacuum pump and isolating
the vessel by closing all inlet and outlet valves.
The flammable gas (methane and/or hydrogen) was fed to the cylindrical combustion
chamber through the method of partial pressures by opening the appropriate
manifold valve. The first gas fed to the chamber occupied the entire volume of the
85
vessel and the amount of the gas was computed from the difference between the
displayed pressure reading and the vacuum pressure reading as monitored by the
pressure transducer and recorded by the LabVIEW program. The manifold valve was
completely closed once the desired amount of methane (or hydrogen) was added to
the vessel. A steady state pressure value was then recorded with a click of "AM to
File". If the experiments were carried out for a mixture of flammable gases (methane
and hydrogen), then the desired quantity of the second flammable gas (either
methane or hydrogen) was injected by opening the appropriate manifold valve. A
steady state pressure value, after closing the manifold valve, was then recorded with
a click of "AM to File".
The air inlet valve on the manifold was opened slowly so as to increase the pressure
inside the rig to around 101 kPa (pressure transducer P1 reading). The introduction
of air in a jet creates turbulence, mixes the gases and results in a homogeneous
mixture inside the vessel. The air inlet valve was completely closed after addition of
dry air and the steady state pressure transducer reading recorded, (in the same file
PP.LMV) with a click of "AM to File".
The gas filler pipe was then removed from the vessel (by disconnecting the coupling
CV4) and locked in the stored position. This action isolated the flammable gas
cylinders from the combustion vessel. The transducer P1 with a maximum pressure
range of 2.048 MPa was also protected from exposure to the overpressure wave.
3.9.2 Gas Ignition
An appropriate value of discharge capacitor and charging resistor, as presented in

Table 3.1, were connected into the electrical circuit (starting with the combination of
R and GP delivering lowest ignition energy required for the experiment to be
conducted).
A trigger pulse to the electrical circuit was actuated with a click of the mouse on the
LabVIEW interface button "Fire & Record". This trigger induced a high voltage across
the electrode tips inside the, gas chamber and initiated a spark. The oscilloscope
screen displayed the voltage as a function of time (ns) over the entire duration of the
86
spark process. The breakdown voltage i.e. maximum voltage achieved during the
time duration was manually recorded (in a MS Excel spreadsheet) as shown in
Figure 3.6 for the specified flammable gas composition prepared. The energy
delivered by the spark was then calculated based on the discharge capacitor in the
circuit and the breakdown voltage recorded by the oscilloscope. The energy was
recorded irrespective of the success, or the failure of the ignition process. If the first
ignition attempt was unsuccessful, then the spark was initiated with a different
combination of resistance and capacitance values giving a higher value of energy for
the same gas composition. A total of four such attempts were made to ignite the gas,
increasing the spark energy incrementally, utilising higher capaCitors during each
attempt. In the case of ignition failure, a very small amount of mixture burnt by the
spark near the electrode tips, can contaminate the un-burnt gases and influence the
ignition energy of the mixture. The spark process also generates free radicals, which
can modify the ignition characteristics. Hence the maximum number of successive
unsuccessful ignition attempts was restricted to four with the same gas-air mixture. If
all the four attempts were unsuccessful then the vessel was again evacuated and
filled with fresh gases to give the same overall composition.
If the spark, ignited the flammable mixture inside the cylindrical combustion chamber,
as observed through the visual inspection, sound, recorded. pressure or temperature
rise, then the spark energy was recorded as a successful ignition. The cylindrical
combustion chamber was evacuated to remove the combustion products, purged
with ambient atmospheric air and then again evacuated to start up the process of
filling with fresh gases to give the same composition. Attempts were made to prepare
exactly the same flammable gas concentration during each planned set of tests and
the actual composition achieved was reported along with the spark energies for
successful and un-successful ignitions in the MS Excel spread sheet.
The ignition of methane can be visually observed since the vessel had transparent
polycarbonate plate flanges. The flame initiation, flame growth and the flame
quenching on the walls of the vessels, can be distinctly observed because of the
nature of the methane - air combustion process. Hydrogen combustion was often
invisible (especially near the lean limit) to naked eyes, and hence the criteria for
ascertaining the success of ignition was based on values of pressure and/or
temperature recorded after each attempt of spark initiation. The gases within the
87
vessel were considered to have been ignited, if the pressure increase within the
chamber was more than 7% of the initial pressure value. Cashdollar, et. al. (2000)
initially suggested this 7% pressure rise, and this was confirmed through the
observations during this work. The temperature rise for lean hydrogen burning, as
recorded during these experiments, was negligible, possibly because of the high
thermal capacitance of the vessel.
A written operating procedure for the experiments leading to an estimation of the

ignition energy and the ignition probability for a quiescent flammable gas
concentration was formulated. This is given in Appendix E. The human actions
performed during the operation of various valves are based on the valves
nomenclature described earlier and shown in Figure 3.1.
3.10 Experimental programme
An experimental programme was formulated to study the entire range of flammable

gases concentrations for methane-air, hydrogen-air and methane-hydrogen-air
mixtures. The main purpose of this work was to study methane-hydrogen-air
mixtures, since this had not been done before and was required for the Naturalhy
project. Measurements of MIE for methane-air and hydrogen-air had been made by
previous workers and these served as a good check on the experimental method
used for the work described in this thesis. As the experiments involved two
flammable gases, methane and hydrogen, experiments were initially proposed in
which the concentration of each of these gases was studied separately over each of
their flammable ranges. Then three different flammable gas mixtures of methane and
hydrogen were studied which were 25%, 50% and 75 % hydrogen in methane. Thus
a total of five different gases; pure methane, 25% hydrogen in methane, 50%
hydrogen in methane, 75 % hydrogen in methane and pure hydrogen were mixed
with air and the ignition of these mixtures studied over their flammable ranges. The
composition of the fuel-air mixture used during each experiment was identified using
the Equivalence Ratio (ER). The Equivalence Ratio is. defined in Section 3.10.1
below and an estimation of the range of flammable concentrations studied during the
experiments is given in Section 3.10.2.
88
3.10.1 Equivalence Ratio
The Equivalence Ratio (ER) is often used to define a fuel-air mixture undergoing a
combustion process and is defined as the ratio of fuel to air compared to the
stiochometric fuel to air ratio. A stoichiometric fuel-air mixture is one that leads to
complete combustion since the air provides just enough oxygen to achieve the final
combustion products of carbon dioxide and water vapour. For example, a methane-
air stoichiometric mixture contains 9.5% by volume methane in air. Similarly for
hydrogen, a stoichiometric mixture contains 29.5% by volume hydrogen in air.
Experimentally, it is very easy to ignite a mixture which is close to the stoichiometric
ratio and often the ignition energies for stoichiometric flammable gases are reported
in the literature. The deviation of an actual gas composition from the stoichiometric
composition is often reported using· the equivalence ratio. Mathematically
Equivalence Ratio (ER) is defined as
Ratio of number of moles of jlamm. gas to air in actual mixture composition

ER
Ratio of number of moles of jlamm. gas to air in stoichiometric composition
A 'lean' mixture will have ER < 1, due to the excessive dilution by unconsumed
oxygen and its associated nitrogen. ER > 1 defines the condition of a 'rich' fuel-air
mixture. Equivalence Ratios for methane, hydrogen and methane-hydrogen mixtures
can be worked out from the Stoichiometric combustion reaction as follows
Combustion reaction for methane Combustion reaction for hydrogen
Moles of CH, =1 Nos. of moles of H2 = 1
Moles of O 2 = 2 Nos. of moles of O 2 = 0.5
Moles of Air = 2' (100/21) Nos. of moles of Air = 0.5' (100/21)
Stoi. molar ratio CH, : Air = 1 : (200/21) Stoi. molar ratio H2 : Air = 1 : (50/21)
Stoi. molar ratio CH, : Air = (21 1 200) Stoi. molar ratio H2 : Air = (21/ 50)
(molar ratio CH4 : Air)ll(:lualcompositim (molar ratio H z : Air)aetualCQmpositim

EReH. = (mo
I ar ratlD
. CH4: A·) ER H, = .. .
zr sloichiomClric (mo/ar ratzo H z . Alr)sloichiomdric
ER = :..(m_o_IQ_r_r_Q_ti_O--,C.,..H":'_:-,A_ir..::)"""'=~_~""'=itim'!!. ER = (molar ratio H z : Air)actua!compositim

CH. (211200) H, (21150)
,EReH = 200 _(mOles of CH.) ER = 50 _( moles of H2 )

• 21 moles of Air ._ .. , ..
......u .. composlllln
H, 21 moles of Air ._ .., ..
...".a, composlU[II
89
For a two component (methane and hydrogen) mixture the equivalence ratio can be
worked out as follows
aCH, + bH 2 + (2a+0.5b )0 2 -------->
where a is the moles of CH, in the mixture

b is the moles of H2 in the mixture
Ratio of fuel (CH, + H2 mixture) to air at stoichiometric concentration will be
FUel] (a + b)
[
Air "oiohiomdrio (2a +0.5b)*('~n
Equivalence ratio for the mixture will be
ERmix,urc
(a +b) ] .
[ moles 0 Air ..
if actual composllllll
ERmix,urc
(a+b) 1
[ 2a + 0.5b * ('~n
(2a +O.5b)*('~nl
ERmixture = .
[
moles of Air .",,'oompo,"irn
ER.
mlxtUTe
= 200
21
*( a) moles of Air actu
" compoS1Um
..
+ 50
21
*( b )
moles of Air actuaI composrtun
..
Hence for the two component (methane and hydrogen) mixture, the equivalence
ratio can be worked out as follows
Therefore the equation used for estimating ER of two component mixture is
ER. lec::.
= -'C(2:.:0:.:0-,-/.::2-,,1)c..*-,(c.:.mc:.0 :.: s -",ofc...::
C:..;ll",,!..)+",::-"(5:.:0,-::/-",2,,,1)c..*-,(",m",o.:..;le;:.s.::,of,-,,-H'-".C2)
mixture (moles of Air)
90
Experimentally the gases mixture were prepared on a volume percentage basis.
Since for gases, the volumetric percentage is equivalent to the molar percentage,
accordingly the equation for the equivalence ratio is modified as :.
ER. = (200/21)' (volumetric percentage oJ CH,) + (50 /21)' (volumetric percentage oJ H,)
ml)([ure (volumetric percentage.of Air)
3.10.2 Flammability limits
The flammability limits LFL and UFL for Methane were taken as 5% and 15%
respectively. For hydrogen the LFL and the UFL values were taken as 4% and 75%
respectively (Lewis and von Elbe, 1987). For the two component methane hydrogen
mixtures (25%, 50% and 75% hydrogen in methane) the flammability limits was
estimated from the Le Chatelier's equation (Lees, F.P. 1980) as follows:
LFL . = -~-=---,---==--=-----,--=---,---.,.,-
m" mole fraction CH. mole fraction H,
+
LF4,
mole fraction CH. mole fraction H,

+
UF4 ,
This Le Chatelier's equation is empirically derived equation and is not universally

applicable. In the present experiments, the equation was utilised to get some idea
about the limiting ranges for the mixtures, so that gas concentrations could be
determined. The LFL and UFL values thus computed from the Le Chatelier's
equation for the gas mixtures are presented in Table 3.2.
Table 3.2: Flammability limits of gas mixture
Gas mixture LFL UFL
25 % Hydrogen in Methane 4.7 21.1
91
3.10.3 Targeted gas concentrations
Methane has flammability limits of 5% to 15% (vol. percentage) at ambient

atmospheric conditions in air. Even though the flammable limits for methane are well
known it is difficult to ignite mixtures close to the limiting conditions. Hence the
extreme end values of the flammability range were discarded from the proposed
experimental concentrations. The ignitions of the lean mixtures is more important
from a safety perspective than rich mixtures. Any accidental flammable gas release
within a confined space, will result in a concentration build-up first crossing the lean
limit.
For the first two gases (methane and 25% hydrogen in methane) four different
concentrations groups were studied. The next two gases (50% hydrogen in methane
and 75% hydrogen in methane) five different concentrations were targeted.
Hydrogen has the widest flammability limits (4% to 75%) and hence was subdivided
into six concentrations so that the entire flammable range could be covered during
the experimental programme. The target concentrations for the 5 different flammable
gases/gas mixtures are shown in Table 3.3.
Table 3.3 : Target gas concentrations for the experimental programme
Gas LFL-UFL (vol.%) Target conc. of total flammable gas (vol.%)
CH. 5-15 6 8 10 14
75% CH. + 25% H2 4.7-21.1 8 12 16 20
50% CH. + 50% H2 4.4 - 27.7 6 10 16 20 26
25% CH. + 75% H2 4.2 -40.5 8 14 21 28 38
H2 4-75 6 10 20 30 45 60
The targeted concentrations of flammable gases were based on the theoretical

calculations of the flammability ranges. These targeted concentrations were initially
prepared inside the vessel, considering the practicality aspect of achieving a
consistent repeatable concentration during each run. The flammable mixtures were
prepared from pure gases methane and hydrogen directly inside the vessel through
the method of partial pressures.
92
Chapter 4
Results, Data analysis and Discussions
A total of approximately 2000 ignitions were attempted on the flammable gas-air

mixtures. These covered the flammable gases methane, hydrogen, and three
different mixtures of methane and hydrogen. Different targeted concentrations for
each of these gases were given in Table 3.3. In each case a range of ignition
energies was used by varying the capacitor-resistor combination. Both ignition and
non-ignition of the flammable gas-air mixtures was recorded. The basic results giving
percentage flammable gas concentration, (methane and/or hydrogen mixture), total
flammable gas concentration, air concentration, spark energy and success of ignition
are presented in Appendix H, from Tables H.1 to H.5. In this section the experimental
data is analysed to arrive at the results from the experimental observations. Firstly
the data analysis of the measured parameters is considered in Section 4.1. In
Section 4.2 the errors in the measurements and derived values are assessed. The
'results obtained for methane, hydrogen and methane-hydrogen mixtures
respectively are presented in Section 4.3 to 4.5. The overall results are discussed in
Section 4.6.
4.1 Data analysis
During the gas mixture preparation, spark initiation and subsequent ignition or non-
ignition of the flammable gas-air mixtures, various parameters were measured and
recorded. The measured parameters were analysed to estimate the gas
concentration, spark energy and the ignition probability from the series of
experiments. The actual calculations. performed during the estimation of the
parameters are discussed in more detail in subsequent sections.
4.1.1 Concentration estimation
Experimental runs were targeted for various gas concentrations across the
flammable range. The gas concentration (CgM ) was estimated from the values of
recorded pressure (kPa) sensed by the pressure transducer (P1 - Figure 3.1).
93
Pressure recordings were made at each stage of the gas filling process (Figure 4.1)
and used to calculate gas concentration as follows.
p - p
C == gaJ vac X 100 ... (4.1)
,~ P
gal-air
Where
P,ac pressure when vessel is under vacuum (kPa)
Pg", pressure when flammable gas is introduced into the vessel (kPa)
Pg",-a;, total pressure of flammable gas and air inside the vessel (kPa)
.::: Pgas-alr
::::
P gas
Pvac
Figure 4.1 : Pressure recording during gases introduction
For two a component system compnslng of a methane-hydrogen mixture, when

methane was the first flammable gas injected during the gas filling process (see
Figure 4.2) the concentration was estimated as follows
94
Ptotal gas·alr
. .........
............. PH2+CH4
. . ... . . . .
. . . . . . . . ........ . P CH4
Pvac
Figure 4.2 : Pressure recording during methane and hydrogen introduction
The Equivalence Ratio (ER) for the flammable gas-air mixture (for two components)
was also calculated as described in Section 3.8.1.
ER. = (200/21) * (volumetric percentage oJ CH,) + (50 /21) * (volumetric percentage oJ H,)
mlJIIUfC (volumetric percentage of Air)
When only one flammable gas (either methane or hydrogen) was mixed with air,
then ER, was calculated as
ERe = 200 * (vOlumetric percentage oJ CH,)

For Methane H, 21 volumetric percentageoJ Air
ER = 50 • (volumetric percentage oJ H, )
For Hydrogen Ht 21 volumetric percentage of Air
4.1.2 Energy estimation
Ignition assurance
In a closed vessel the success of ignition for invisible flames was difficult to establish.
In order to determine the ignition energy, a criteria for successful ignition is required.
. .
Most tests resulted in visually observed ignitions but for the borderline cases, the
pressure transducer was relied on to define ignition. The overpressure wave as
95
observed for methane (6.7% and 8%) and hydrogen (30%) ignition are presented in
Figure 4.3.
800
CH4 (6.7%)
CH4 (8%)
600 H2 (30%)
ill
·c
e
"
~
.
~ 400 ------------------------------------------------------ -----
E
c .r\
'.
~ \.,
"
11. "
200 - -.- -- -- ---~-----------------------------------------------
o ,:
,
" ~~.-------
- . . .-=:---- '- ------
-
o 2
Time (s)
3 4
Figure 4.3 : Overpressures during methane and hydrogen ignitions
When the pressure rise was very small, at lean ignitions of hydrogen (Figure 4.4),
assurance was required that the pressure rise was due to gas combustion. It was
initially confirmed that there was no pressure rise, associated with the spark by
recording the pressure for test runs with air. To distinguish ignitions from non-
ignitions for lean concentrations of hydrogen, the combustion reaction should result
in a quantifiable pressure rise within the vessel. The smallest pressure rise observed
during the combustion process was just over 7%.
96
--------------------------------------------------------_.---
10
- - 112(6%) I
8 ----------------------------------------------------------------
."2!
I!! 6
"::l
I!!
a.
"
Cl
J!lc: 4
~
"
Q.
2 - --------------------------
o J
o 2 Time (s)
3 4 5
Figure 4.4 : Overpressure for lean hydrogen ignitions
Energy determination
Electrical sparks have been used over a number of years to measure the Minimum
ignition Energy (M lE) of gases. Electric sparks are a very hot and fast acting mode of
ignition. The discharge time of an electric spark is very short (of the order of 10-8 to
10-7 s), the energy that is imparted to the gas at the end of the discharge period is
highly concentrated, so that a very steep temperature profile with a very high
temperature at the centre is established. (Lewis and von Elbe 1987). In the initial
stage of the flame development, the combustion process is insufficient to maintain
such a steep temperature profile, so the profile broadens and the temperature at the
centre decreases. Depending on the physical and chemical properties of the gas and
the discharge energy (if sufficient to establish sustained combustion) the flame
continues to propagate as a steady state wave. The temperature at the centre settles
down to the value of the flame temperature.
Spark discharges are typically two electrode discharges. They occur between
conductors (electrodes) with different electric potentials. A spark is a discrete
discharge that bridges the gap between the two conductors in the form of an
ionisation path, in which the stored energy is quickly transferred. The spark is
97
triggered when the breakthrough field strength is reached at a certain pOint in the gap
between the electrodes, normally at the surface of the electrodes. The field strength
between the electrode space must be sufficiently high, that the discharge can travel
through that space. For a homogeneous electric field between the electrodes, spark
discharge will occur when the ratio of potential difference and the gap between the
electrodes reaches about 3 kV/mm (dielectric strength of air). For an electrode
spacing of 2 mm a potential of 6 kV would be required for a spark to occur. Although
the dielectric strength of air is 3 kV/mm, the dielectric strength of hydrogen is only
1.75 kV/mm. (Astbury and Hawksworth, 2007). The variation of dielectric strength
with concentration of hydrogen in air mixtures is unknown.
Studies of nanosecond duration spark discharges suggest that the process proceeds
in several stages: (i) deposition of electrical energy to create an ionized channel of
hot gas between the electrodes; (ii) expansion of the hot gases and creation of a
shock wave; (iii) recirculation of fresh gases into the channel; (iv) turbulent mixing of
hot and cold gases (Kono et. al 1989). As a result of these processes, the kernel of
hot gas created in the discharge region initiates a reaction in the surrounding cooler
gas, buoyancy may also come into play. If the ignition energy is sufficiently large,
then a propagating flame emerges from the vicinity of the kernel of hot gas. Several
ideas (Lewis and von Elbe 1987, Kondo et. al 2003) have been proposed to correlate
ignition energy with other combustion parameters. Because the process is very
complex, there was limited success in the theoretical prediction of ignition energy.
The standard way to characterize capacitive discharge sparks is in terms of the

quantity of stored electrical energy, measured as millijoule (mJ). Precise
measurement or estimation of stored electrical energies is essential when
determining the ignition energy. The actual amount of energy that is deposited in the
gas by the discharge is lower than the actual spark energy, particular for short
duration sparks. A distinction can be made between measurement of gross capacitor
energy prior to breakdown and integration of spark power versus time. In the former
case, the energy is simply assumed equal to the difference between stored capacitor
energy before and after the discharge:
98
... (4.2)
where E is the calculated energy (J)

since E, is very small it is reported as mJ in the entire thesis
GP is the capacitance (F)
Vbefore is the capacitor voltage before discharge (V)
Vaffe, is the voltage after discharge (V)
Usually, Vbefore » Vafter , and thus the energy can be approximated by
... (4.3)
Whether the spark energy can be accurately estimated by this simple expression
depends on the characteristics of the discharge circuit. The discharge circuit
inevitably contains some resistive elements in series with the spark, where some of
the energy is lost, and not delivered to the spark. The energy lost in the electric
circuit, to the electrodes, through a corona and radiation is disregarded when stating
the spark energy. When estimating the spark energy purely based on stored
capacitor energy, information about the spark duration, voltage and current
waveforms are not obtained.
Randeberg E. (2006) nieasured the circuit variables (voltage and current trace
waveforms) and calculated energy from the integral of the power, i.e. the product of
voltage 'V' and current 'I' of the spark, over the duration of the discharge.
E= Iv I dt ... (4.4)
Where V is the voltage on capacitor at a point in time (during the duration of spark)
Thus, only the energy delivered by the spark was found, and circuit capacitance and
losses are implicit. A major challenge to this approach is the fact that the different
spark phases (e.g. creation of ionized channel of hot gas, breakdown, and spark etc.)
have durations differing by orders of magnitude, and the currents and voltages are
varying by orders of magnitude in the different phases. (Maly R. and Vagel M. 1979).
99
The actual energy delivered to the spark can be estimated by subtracting the
resistive losses from the spark energy over the duration of the discharge using:
E= fVldt-fRI'dt ... (4.5)
where R - is the circuit resistance in series with the spark gap (0)
I - is the current flowing through the spark (A)
V - is the voltage measured as a function of time during the spark (V)
However the resistive losses are small, for the test rig used for this work, so the
second term in the Equation 4.5 can be neglected. Moreover, to measure the spark
energy as per Equation 4.5, it was important to measure parameters 'V' and 'I' as
functions of time. (nanosecond discharge over the spark duration). The current 'I'
was found to oscillate over the spark duration. The metallic stainless steel vessel (a
very good electrical conductor) might have impacted upon the frequency· and
amplitude of this current trace. These difficulties with energy measurement prevented
the use of Equation 4.5 to calculate the energy. Moreover most of the earlier work
(including those of Lewis and von Elbe) utilised Equation 4.3, to calculate the ignition
energy. Accordingly the stored capacitor energy and the maximum value of the
voltage trace waveform (captured by the oscilloscope) were used to derive the
ignition energy throughout this work.
E=O.5CP*V ' ... (4.6)
4,1.3 Ignition probability computation
The result presented in Appendix H, Tables H.1 to H.5, show that for higher spark
energy levels, more successful ignitions were achieved, indicating an increase in the
probability of ignition with increasing energy. The ignition energy results for each gas
concentration for each flammable gas were analysed to derive the associated ignition
probabilities as follows:
The entire energy range (zero to maximum value) was sub divided into bands of
2 mJ (except for test results of hydrogen). In each band there would be number of
ignition and non-ignition results. However, if a mixture had been successfully ignited
by an energy within a low energy band, then clearly this mixture would also have
100
--------- -------
been ignited had the energy been greater. Therefore, in calculating the probability of
ignition associated with a given energy band, successful ignitions from all lower
energy bands were also included. That is :
Assume the energy range used for a series of experiments at a particular

concentration is divided into 'M' bands and that for a given band, m. (1 ,;;m ,;;M)
there were IG msuccessful ignitions and Nlm non - successful ignitions.
Then the probability of ignition (IP m) for the energy level associated with band 'm' is
given by
... (4.7)
1'=\
The ignition probability for each of these individual bands was determined
accordingly and then assigned to the average energy value within this band.
The first band was lowest value of non-ignition energy to 2 mJ. Below the lowest
successful ignition energy value within this first band, the ignition probability has to
be considerably less than the ignition probability for the band. Accordingly, a value of
zero ignition probability was assumed for the average energy value, calculated from
lowest non-ignition energy value and lowest ignition energy value, within this first
energy band.
The ignition probability (IP) was observed to be an exponential function of the ignition
energy (E). Hence an equation of the following form was assumed to relate the
ignition probability to the energy E.
JP = \- exp[- b(E - E min )] ... (4.8)
Where' E min ' is the energy when IP equals zero.
The above equation tends to 'unity' for higher values of 'E' and is 'zero' at ' Emin "
and well describes the observed variation in the ignition probability with energy.
101
--------
Since standard Microsoft excel worksheet does not correlate this type of curve fitting,
Equation 4.8 was rearranged as follows:
Ln(I-IP)=bE.
mm -bE ... (4.9)
A best fit plot of . Ln(l- IP) . against energy' E' provides a line with:
slope = -b
Intercept =bEmin
Accordingly the parameters . b 'and . Emin' was derived for each of the target
concentrations for each of the flammable gas. The energy for 1% ignition probability
(often reported by many authors) was then calculated from Equation 4.8 as:
E/p=I% = E min - b1 * Ln(l- 0.0 I) ... (4.10)
Where E /P=I% is the energy for 1% ignition probability
E min is the energy when ignition probability is zero
4.2 Error Analysis

The quantifiable parameters of pressure, temperature, capacitance and voltage were
measured by various instruments during the experiments. Measurement of these
parameters can not be 100% perfectly accurate or exact. The gas concentration in air
of the mixture produced is determined from measurements of pressure during the
filling process. The accuracy of the concentrations determined is considered in
Section 4.2.1. The energy delivered by the spark is determined from the voltage and
capacitance as given in Equation 4.6. The accuracy of the spark energy is assessed
in Section 4.2.2.
102
4.2.1 Error analysis for concentration
The absolute error in the measured parameter and accordingly the maximum and the
minimum measured values are presented in Table 4.1.
Table 4.1 : Error in pressure measurement
Measured Measurement Maximum value Minimum value

parameter error
Pressure ±M P"", = P+M p.

mm =P-M
Concentration ±~e (e gas )"", =(egJ+~e (egJ mi
" = (eg..J-~e
Referring to specification of the pressure transducer (PX 219-30V15G5V) the error

(± M ) in pressure measurement is ± 0.25%.
i.e. ±M = ± 0.0025 * P . .. (4.11)
The error in concentration of the gas (e gas ) was determined from the pressure
transducer error (± M) used during gas mixture preparation stage. The

concentration was measured as per Equation 4.1 discussed earlier.
p - p
C = gas \lQC' xtOO
,~ p
gtu-iljr
The gas concentration (e gas ) is the volume percentage concentration of the
flammable gas in the gas-air mixture.
The basic equation for error in the concentration measurement of the gas is
... (4.12)
The maximum value of concentration will be
(c,~L (p,w +6P,J- (p,., -6P~J

100 (P,ar-air - OPgas_air)
103
(C,m)~ (p,m +0.0025P,J- (P=-0.0025P=)
100 (p,m.,;, - 0.0025Pgm .,,,)
(C,m)mu (Pgm-p,J+0.0025(Pg~ +P,,J

lOO P,~.,,,(1-0.0025)
Similarly the minimum value of concentration will be
(C,~L _ (Pgm -p.J-0.0025(Pgm +P~)

lOO - Pg".,;, (1+ 0.0025)
2' l>e (p•• - p.J+ 0.0025(P,•• + p.J (p._ - p.J-0.0025(P,. + p_l

JiiO - P.__ .JI - 0.0025) P,_-.JI + 0.0025)
(I +0.0025)[(Pgm - P~J+ 0.0025(Pg~ + p~J- (1- 0.0025)[(Pg~ - P~J - 0.0025(Pgm + p.J

Pg~.",(I- 0.0025)(1 + 0.0025)
Pg" -P~, +0.0025(Pg~ +P~J+0.0025(Pg" -p~J+0.0025'(Pg~ +P~J

-(Pg" -p~J+0.0025(Pg~ +P~J+0.0025(P,,, -p~J-0.0025'(P,,, +P~J
P,,,.,.(1-0.0025' )
2 • 6C _ 2' 0.002S(Pg~ + P~, + P,~ - p~J

100 - p,~.", (I - 0.0025')
!!.C 0.0025(2Pg",)
... (4.13)
100 = Pg",.a;, (1- 0.0025 2 )
Considering (1 - 0.0025 2 ) = 1 and that P.a, is small so Pg", = Pg", - P.a,
... (4.14)
104
~o~ ;: 0.0025 * 2Cgas
D.C ;: 0.5 or ... (4.15)

C gas
The percentage error in concentration is 0.5% of the measured gas concentration

. (Cgas ) or the error in concentration is twice the error in pressure measurement
(2M'). Even if Equation 4.13 is utilised to calculate the percentage error in

concentration, the maximum error was 0.65% for the range of pressure values used
during the experiments. Accordingly Equation 4.13 and Equation 4.15 give a similar
error.
The error estimated for the gas concentration is very small (D.C = 0.5%) as compared
to the range of the gas concentrations studied (6.5% to 10.5%). The error for the
methane gas concentration as shown by the horizontal spread of the concentration
value is presented in Figure 4.5. Two data sets of 6.5% and 8% methane
concentrations only are plotted to show the horizontal spread clearly. Since the error
is very small the concentration parameters are displayed as a single value for all
results presented hereafter.
Error estimation for the methane-hydrogen mixture was estimated from the individual
errors of methane and hydrogen concentrations. The error in ratio of methane to
hydrogen was estimated as follows
R (max) = C eH , (max) and R (min) = CeH , (min)

CH2 (min) CH2 (max)
2D.R = R (max) - R (min)
2D.R = CeH , (max) _ CeH , (min)

CH2 (min) C H2 (max)
105
CCH 4 -I'1CCH 4
CH' + I'1CH,
C [1-
CH4 I'1C
C
CH4 ]
CH4
21'1R
CH' [1- CI'1C H , ]
H'
Since R = CCH4 and

CH'
21'1R = R(I + 0.005) R(I-0.005)

(1- 0.005) (I +0.005)
2 I'1R = (1 + 2 *0.005 + 0.005') - (1- 2 * 0.005+ 0.005')

R 1- 0.005'
2 I'1R = 4 * 0.005
R 1-0.005'
Considering (1 - 0.005 2 ) = 1
I'1R : 2 *0.005
R
I'1R I'1R
-:0.01 or -:(2I'1C) ... (4.16)
R R
Accordingly the error in the ratio of the flammable gases is 1% (twice the gas
concentration error). The error is still small compared to the gas concentration ratios
considered in the experiments for the methane-hydrogen mixtures.
106
4.2.2 Error analysis for energy
The specifications of capacitors suggest that the capacitance values for each
capacitor have around 10-15 % error. The actual capacitance of each capacitor was
measured using RLC (Resistance Inductance and Capacitance) meter to arrive at the
correct capacitance value. Table 4.2 presents the nominal and actual measured
values of capacitors used in experiments.
Table 4.2 : Nominal and actual capacitance of the capacitors
Capacitor (nominal capacitance) Actual capacitance measured by RLC meter

(pF) (pF)
10 10.8
47 55.8
100 111.3
150 160
200 206
300 336
A Tektronix Oscilloscope model no. TDS 3034 B was used for measuring the voltage.
The specifications of the voltage measuring instrument (oscilloscope associated with
the high voltage probe) suggest that the error (± L'lV) is 0.018% of the measured
voltage.
Actual error in voltage measurement = ± L'lV
= ± 0.00018 * V volt
A similar exercise was carried out for the errors in voltage (Table 4.3) and energy
measurement, as was carried out during error analysis of ,gas concentration
measurement. The results are shown in Table 4.3.
Table 4.3 : Error in voltage measurment
Measured Measurement Maximum I Upper value Minimum I Lower value

parameter error
Voltage ±iw V""x = V +L'lV Vmm =V-L'lV
Energy ±M Eupw=E+M E,o_, =E-M
107
Energy E = 0.5 Cp·V2
Error in the energy can be evaluated from upper and lower values:
Accordingly the upper and lower values of energy are
E,,!,,,, = O.SCP* V";,,
Z * !lE = O.SCP * V";,, - O.5CP * V~;n
Z * !lE = O.SCP* (V";', - V~;n)
Z * !lE = O.SCP * (Vm" - Vm;n ) * (Vnw, + Vm;n)
Z * !lE = O.SCP * (Z~V) * (ZV)
Z * !lE = O.5CP* V2(Z * 0.00018 * V) *(ZV)
Z * !lE = O.SCP * V2 (4 * 0.00018)
!lE=0.SCP*V 2 *(Z*0.00018)
!lE
-=0.00036 or !lE =(nV) ... (4.17)
E E
Hence, the error for the energy measurement was calculated as twice the voltage
error. i.e. 0.036% of the measured energy value. The error estimated for the ignition
energy is very small as compared to the range of the energy values observed during
experiments. The typical measurement errors for the methane test runs are shown in
the Figure 4.5, by the vertical spread of the energy values plotted against the gas
concentration. Indeed, the error was too small to be clearly seen on the plots, hence
only a single value for the ignition energy is used for all results presented hereafter.
108
10,------------~~--------------------------------------~
~ -i-ftI'i
I : ::5 I
t+1t+1~
t+1~
I~~ ~
~~
t+1
$ t+i
t+1 ~~ ~ ~
--- --- -- --- - - ---- -~- -- --------- - - -- - - -- -- - - - - - - - - - ....
t+i
~ - - - - --
~ t+i
t+i ~
;~~
t+i ~
0.1 +----------r----------~--------_r----------~--------~
6 6.5 7 7.5 8 8.5
Methane Concentration (%)
Figure 4.5 : Errors analysis for experiments with 6.5% & 8% methane concentrations
4.3 Results for methane
Experimental test runs with methane gas at the nominally lean (5%) and rich (15%)
flammability limits never resulted in any successful ignitions. The concentration range
of methane was then narrowed and ignition attempted with the highest spark energy.
No ignitions were observed even after a hundred attempts outside the range 6% to
12% methane concentration. Experiments were then conducted with the target
concentrations shown in Table 4.4. The corresponding equivalence ratio is also given
in the same table.
Table 4.4 : Experimental conditions for methane tests
Parameter Concentration of methane (% by volume)
Observed range outside of

6 <---------------------------------> 12
which there were no ignitions
Target concentrations for

methane experiments
-- 7 8 g 11 --
Equivalence Ratio (ER) 0.61 0.72 0.83 0.94 1.06 1.30
109
Twenty different gas mixtures for each target concentration were prepared and
ignition attempted up to 4 times (maximum) for each. The actual gas concentrations
were calculated from the pressure transducer readings. A plot of the actual gas
concentration prepared in each run and the observed ignitions (Yes) and non-
ignitions (No) for a range of energies is presented in Figure 4.6.
14
c
o IoNol
10 - - -- -.-
• c •
_Q - - - - - - - - - - - -- - - - - - - - - - - - - - - - - - - - - - - - - - -
c ~
-.0.---- - - - ---
-,
.§.
~6 !,.. ------- ----- ---- --------------- --------_!.t -- -- --- -- ---
- - -- ..
~ , tf
w ea
< eO 0
~.
•
"I:~ ------------
c 0 ,.,
2 - - - fJ-?:-l- ----- -1-

c-~ ----;l----0 - - - - - - - - - - - - - - ~
• .,. .",. ~~ i~ • ij
-2~-------- __------~--------~---------r-------- __ --------~

6 7 8 9 10 11 12
Methane Concentration (%)
Figure 4.6 : Results of ignition experiments for methane
The data is summarised in Table 4.5. An arithmetic average of the different nominal
concentrations for the specified set is also presented. If the first ignition attempt was
uns\lccessful, the energy was progressively increased (by introducing a larger
capacitor into the circuit), until the vessel contents were ignited or a maximum of four
attempts had been made. The range of ignition energies; lowest to the largest, for the
specified nominal concentration set, as observed during test runs is also presented.
Nominal concentrations at the near stoichiometric (8% and 9%) compositions were
almost all ignited with only a very small increase in energy value. Hence energies
were not increased further for the 8% and 9% composition mixtures. Concentrations
of lean (7%) and rich (11%) mixtures for the twenty samples (each) could not be
ignited for the largest energy that could be generated inside the rig, even though a
few ignitions were observed at lower energies. The range of ignition energies (lowest
to largest) as observed for methane is presented in Table 4.5. Ignition probabilities
calculate from this range of observed energies is discussed in Section 4.3.2.
110
There is a large variability in the ignition energy for 7% and 11 % methane
concentrations. The reasons for non-ignitions can be due to many reasons, including:
the condition of electrode tips (covered with combustion products) may prevent the
combustion, even though the sparks are visible across the electrodes. Similarly minor
changes in temperature and humidity can also play an important role for deciding the
success or failure of ignition. Moorhouse J. et. al. (1974) , Esseghir, M. and
Polymeropoulos, C.E. (1988), Shepherd, J.E. et. al. (2000) and Randeberg E. (2006)
observed similar variability in the ignition energy when working with their
experimental rigs. All these parameters suggest a probabilistic element to the ignition
energy as discussed in section 4.3.2.
Table 4.5 : Observed Ignition energies for various methane concentrations

.
Target concentration (%) 7 8 9 11
Actual Concentration (%) 6.69 7.96 8.95 10.63

(Mean ± std. dev) ±0.15 ±0.17 ±0.20 ±0.10
Actual Equivalence ratio (ER) 0.68 0.82 0.94 1.13

±0.015 ±0.018 ±0.022 ±0.011
(Mean ± std. dev)
Lowest observed Ignition 1.82 0.24 0.22 0.83

energy for gases (mJ)
Range of observed Ignition 1.82 0.24 0.22 0.83

to to to to
energies for gases (mJ)
13.13 3.75 2.27 10.85
4.3.1 Comparison with previous data

Previous work on the ignition of quiescent methane-air mixtures at atmospheric
pressures was examined. The ignition energy results of the present work are
compared with Lewis & von Elbe, 1987, Bjerketvedt et. al. (1997) and calculated MIE
data of Kim et. al. (2004) for methane in Figure 4.7 which shows the lowest observed
ignition energy plotted against mean methane concentration. Kim et. aI., 2004 have
solved one dimensional unsteady state conservation equation numerically to arrive at
the MIE values.
111
100
....... Lew is & van Sbe data -- Bjerketvedt et.al data
:::;- - - - - Kim et. al data - e - - this work

~
'"E!'
"c 10
"c
.~
.
~
.
.2'
"~ .
..
"on 1 ---- -- --- --- - - - -. . . - .-
'.''.,....
.Q
0
.
1ii
"0~
...J
-.-
~~.-'..... -'-. ...,,-'
0.1
5 6 7 8 9 10 11 12 13
Mean Methane concentration ( %)
Figure 4.7 : Comparision of methane ignition energy results with those from the
literature
The lowest observed ignition energies agree reasonably well with the reported data,
except at the leanest mean concentration. The experimental method used by earlier
researchers allows high voltage to be present on the electrodes for a considerable
duration. Long duration of high voltage permits the condition of a corona discharge.
High voltage before the spark can ionise the gas, reducing the dielectric strength of
gas-air mixture. This in turn reduces the actual voltage at which spark occurs directly
affecting the calculation of energy (E = 0.5 CP*V2 ). Thus the energy required for
sparking process is reduced resulting in reporting of lower MIE value at the lean
concentration. Lewis and von Elbe (1987) further introduced radium capsules of
various strengths inside the bomb (experimental vessel) to reduce the time lag
between the charging process (voltage elevation) and the onset of spark. The
introduction of radium capsules may have lowered the MIE values. Ignitions at near
stoichiometric are always guaranteed and hence these effects may not have
impacted in lowering of MIE, which is in good agreement with this work.
4.3.2 Ignition probability

The success of ignitions quantified through the experiments was reported as ignition
probability at the specified concentration of methane in air. Ignition probability in each
of the 2 mJ energy bands was determined as described in Section 4.1.3 and the
112
computed values are presented in Table 4.6. The maximum ignition probability value
was restricted to 99% to get finite value for the term' Ln(l- IP) '. There was an
observed increase in the ignition probability from lower to higher energies, with
higher energies resulting in more successful ignitions.
Table 4.6 : Ignition probability for various methane concentrations
Mean Percentage ignition probability for energy bands

methane
0-2 mJ 2 - 4 mJ 4 - 6 mJ 6 - 8 mJ 8 - 10 mJ 10 - 12 mJ
concentration
CH. = 6.7% 5 20 59 69 89 89
CH. = 8% 29 82 99 99 99 99
CH. = 9% 49 99 99 99 99 99
CH. = 10.6% 15 47 75 68 88 94
The coefficients within the exponential function, Equation 4.8 (values' b 'and' E min ')
. and the calculated ignition energy at 1% ignition probability using the Equation 4.10
is presented in Table 4.7. The lowest experimentally observed ignition energy is also
shown for comparison with the E 1P=1% value.
Table 4.7: Ignition energy at 1% ignition probability for methane
Mean methane Lowest observed Value Value Energy

. concentration ignition energy (mJ) 'b' Emin (mJ) E1P=1% (mJ)
CH. = 6.7% 1.8163 0.2285 1.1803 1.2243
CH. = 8% 0.2402 0.6245 0.3339 0.3499
CH. = 9% 0.2242 1.7017 0.3874 0.3933
CH. = 10.6% 0.8349 0.2535 0.5538 0.5935
The value of E min defines the energy level at a theoretical ignition probability of
'zero'. Whereas the more significant lowest value (often reported by various
researchers) is at an ignition probability of 1% i.e. E 1P=1%' The lowest observed
ignition energy is based on the limited number of experiments carried out, whereas
E 1P=1% (theoretically calculated), is based on one in a hundred chance of ignition.
113
A plot of ignition probability against energy as observed for methane-air is presented
in Figure 4.8. Also shown in Figure 4.8 is the line representing the Equation 4.8 fitted
to each data set.
0.8
.1P1 (6.7%)
• 1P2 (8%)
AIP3 (9%)
x 1P4 (10.6%)
0.2
-2 o 2 4 6 8 10 12 14 16 18 20
Energy (mJ)
for methane
4.4 Results for hydrogen

Similar to methane-air experiments, test runs with hydrogen-air at the lean (4%) limit
never resulted in any successful ignitions. The concentration of hydrogen was
increased and ignition attempted with the highest spark energy. The highest lean
concentration below which no ignitions were observed even after a hundred attempts
is presented in Table 4.8. The target concentration prepared for the experimental
runs are also depicted in the Table 4.8. Since the concentration range examined for
hydrogen was very wide seven different concentrations were targeted. The
corresponding equivalence ratios are also given.
114
Table 4.8 : Experimental conditions for hydrogen
Parameter Concentration of hydrogen (% by volume)
Observed lean limit below

5 <------------------------------------->
which there were no ignition
Target concentrations for

-- 6 10 20 25 30 45 60
hydrogen experiments
Equivalence ratio (ER) 0.13 0.15 0.26 0.60 0.79 1.02 1.95 3.57
Experimental runs for different target concentrations for hydrogen were conducted
similar to that for methane. A plot of the gas concentration prepared in each run and
the observed ignitions (Yes) and non-ignitions (No) for a range of energies is
14
•
\ I ::, I
10 ----~-----------------------------------------------------------
-. .
•
••
----,.-----------------------------------------------------------
•
, 0. 0
'"
0
.~
• ij 0
•• ·0 0
2 - - - -~. - - I!L - - - - - -- - - - - -- - - - --- - -- - - - - - - - - - -- -. - - - - - - - - - -- - - -..:I' - - - --

ffl\ 0 '0 • ..,
A~ ~ !/1I11 aPiII> ~ I·~go:'·
·2~------ __- - - - - -__- - - - - - - -__- - - - - -__- - - - - -__- - - - - -__ ------~
o 10 20 30 40 50 60 70
Hydrogen concentration {~
Figure 4.9 : Ignition energies for a range of hydrogen concentrations
An arithmetic average of the different nominal concentrations was calculated and

presented along with a mean and a standard deviation value in Table 4.9. The range
of ignition energies as observed, is also presented in the same table. The range of
ignition energies studied are narrower at the 20%, 25% and 30% concentrations as
these mixtures are easily ignited at low energy levels (near stoichiometric
concentration of hydrogen-air mixtures).
115
Table 4.9 : Observed Ignition energies for various hydrogen concentration
Nominal concentration (%) 6 10 20 25 30 45 60
Actual concentration (%) 6.32 10.65 20.96 25.47 30.38 45.68 62.11
(Mean ± std. dev) ±0.44 ±0.89 ±0.74 ±0.64 ±0.94 ±1.19 ±1.83
Equivalence ratio (ER) 0.16 0.28 0.63 0.81 1.04 2.00 3.90
(Mean ± std. dev) ±0.01 ±0.03 ±0.03 ±0.03 ±0.05 ±0.10 ±0.30
Lowest observed Ignition

0.76 0.10 0.04 0.03 0.02 0.06 0.22
energy for gases (mJ)
Range of observed' Ignition 0.76 0.10 0.04 0.03 0.02 0.06 0.22
energies for gases (mJ) to to to to to to to
11.96 4.63 0.87 0.36 0.24 0.56 4.96
4.4.1 Comparison with Previous data
Previous work on the ignition of quiescent hydrogen-air mixtures at atmospheric

pressures was examined. The ignition energy results of the present work are
compared with Lewis & von Elbe, 1987 and D.W.v. (German Hydrogen Association,
2002) for hydrogen. Further the MIE values were also compared with experimental
work of Ono et. al. (2007) and Kim et. al. (2004)
The experimental results of Ono et. al were aimed at understanding the influence of
spark gap length on the MIE values. The lowest values of MIE were obtained with
spark gap 'of 0.5 mm. MIE increases with increasing distance between spark gap.
The minimum spark gap delivers the lowest MIE which are compared with the
present work.
Kim et. aI., 2004 have solved one dimensional unsteady state equation numerically
for hydrogen-air mixture to arrive at the MIE values. Detail chemical kinetic
mechanisms are used to predict the MIE for hydrogen-air mixtures. The results
showed that MIE is a sensitive function of the radius of ignition energy source
considered as half of quenching distance and the duration of supplied energy as 0.52
I.ls. The MIE values at ambient pressure as reported by Kim et. al are compared with
the present work.
116
100
Law is & vcn 8be data D.w.V data
Ono et. al.data Kim et. al.data
---e- this work
.'
.... _-----
0.01 .j..:...--~---_---~--~---_---~--~---_I
o 10 20 30 40 50 60 70 80
Mean Hydrogen concentration ( %)
Figure 4.10 : Comparision of hydrogen igntion energy results with those from the
literature
The lowest ignition energies for hydrogen have a 'U' shape dependence on
concentration (similar to that of methane). The energy increases asymptotically at the
lean and rich limits. The lowest observed values of the ignition energy agree
reasonably well with the reported data.
The success of ignitions quantified through the experiments was used to derive
ignition probabilities for each concentration of hydrogen gas. Different energy
bandwidths were selected for the different concentrations. At the leanest
concentration (6%), there were not many ignitions in each bandwidth, therefore a 4
mJ bandwidth was selected and the results are presented in the Table 4.10. At near
stoichiometric (20 - 30%) concentration and also at 45%, ignition was achieved
easily at low energy levels, hence a 0.5 mJ bandwidth was selected as presented in
Table 4.11. For other concentrations 10% and 60% an energy bandwidth of 1 mJ
was considered appropriate as shown in Table 4.12. Ignition probability in each of
these individual bands was determined as described in Section 4.1.3. The results of
the 20%, 25% and 30% were grouped together when drawing the exponential curve
fit line.
117
Table 4.10 : Ignition probability for very lean (6%) hydrogen concentration
Mean Hydrogen Percentage ignition probability for energy bands
concentration 0-4 mJ 4-8 mJ 8 -12 mJ
H2 = 6.32% 3 10 15
Table 4.11 : Ignition probability for 20% to 45% hydrogen concentration
concentration 0- 0.5 mJ 0.5 -1 mJ
H2 = 20%, 25% and 30 % 75 99
H2 = 45.68% 61 99
Table 4.12: Ignition probability for lean (10%) and rich (60%) H2 concentration
concentration 0-1 mJ 1-2 mJ 2-3 mJ 3-4mJ 4 - 5 mJ
H2 = 10.65% 22 62 79 89 95
H2 = 62.11% 17 42 77 84 91
The coefficients of the exponential function as described in equation 4.8 (values

'b 'and' E min ') and the calculated ignition energy at 1% ignition probability using the
Equation 4.10 is presented in Table 4.13. The lowest experimentally observed

ignition energy is also shown for comparison with the EIP=I'!. value.
118
Table 4.13 : Ignition energy at 1% ignition probability for hydrogen
Mean Lowest observed Value Value Energy

Concentration ignition energy (mJ) 'b' E min (mJ) E/P=l% (mJ)
H2 =6.32% 0.7607 0.0173 0.3815 0.9624
H2 = 10.65% 0.1049 0.6744 0.1481 0.1630
H2 =20% - 30% 0.0244 6.3469 0.0287 0.0302
H2 =45.68% 0.0576 6.7521 0.0841 0.0856
H2 =62.11% 0.2242 0.5632 0.2317 0.2450
The lowest observed ignition energy is based on the limited number of experiments
carried out, whereas E/P=l%' is the one in hundred chance of ignition. A constant
ignition energy value (E/P=l% ) for 20%, 25% and 30% is in agreement of D.W.v data
A plot of ignition probability against energy as observed for hydrogen gas is

presented in the Figure 4.11. Also shown in Figure 4.11 is the line representing
Equation 4.8 fitted to each data set.
0.8
,
- - - - -- - ,- - - - - - - - -- - - -- ----------------------
• P1 (6.32%)
• P2 (10.65%)
• P3 (20-30%)
X P4 (45.68%)
., o 2 3 4 5 6 7 8 9
Energy (mJ)
Figure 4.11 : Ignition probability at various energies along with exponential trend
lines for hydrogen
119
4.5 Results for methane hydrogen mixtures
4.5.1 Gas with 75 % me'thane and 25 % hydrof/en
The flammable range for the gas mixture (methane with 25% hydrogen) was initially
determined using the Le Chatelier's equation (see Table 3.2). The lowest lean
concentration practically achievable was 4% flammables in air (3% methane and 1%
hydrogen). It was quite evident that, since this concentration was below the LFL
given in Table 3.2 (4.7%), ignition was not possible. Experiments performed with the
next practically achievable concentration, 8% flammables in air (6% methane and 2%
hydrogen) resulted in ignitions, hence this was considered as the lowest practical
lean concentration. The rich limit where no ignitions were observed even after a
hundred attempts and the target concentration are presented in the Table 4.14.
Table 4.14: Experimental conditions for 75% CH, + 25% H2 mixture
Parameter Concentration of flammable gas mixture (% vol.)
Gases CH, H, CH, H, CH, H, CH, H, CH, H, CH, H,
End limits
3 1 15 5
(no ignition) <--------------------------------------->
Total (CH, + H,) 4 20
Target
-- -- 6 2 8 3 10 3 12 4 -- --
concentration
Total (CH. + H,) --- 8 11 13 16 ---

Equivalence
0.44 0.67 0.94 1.18 1.5 1.6
ratio (ER)
Gas mixtures at the target concentrations were prepared and the actual gas
concentrations determined from the partial pressures of methane and hydrogen as
described in Section 4.1.1. A plot of the observed energies, ignitions (Yes) and non-
ignitions (No) against the concentration are presented in Figure 4.12.
120
14
o.
•• 0 ~r-I:J
0 Yes
10 - ----- ----- ---- -- --- -- -------- ---- - -0- C- - - - - - - - - -er- --------------
cP. •
...,
.§.
o
o
. • .,. •
~ 6 -------------------------------- c-rq,-.- - - - - - - - -...• , .....
.'a - _. -- - - - - -- --
I! 0 jl ....'lo
W
o
[J oD*
.CI. • • •-
.
2
·0
- - - - - - _0_ - - - - - - - - - - -0 - - -
~: IJ [J
•
~ - - - - - - - ~- ~
·ce
-I -.--------~! -.fC -" ------------
0 •
, •
o '!' ~/!m1 e,~ .. ~.fIJI' •
-2~---- __------~------__------~----________~______~
6 8 10 12 14 16 18 20
Totalflammables conc. (%) of mixture (75%CH4 + 25%H2)
Figure 4.12 : Ignition energies for 75% CH. + 25% H2 mixture
An arithmetic average of the total flammable concentration (75 % CH4 + 25 % H2

mixture) was calculated for each nominal specified data set, and presented in
Table 4.15. The lowest ignition energy and the range of ignition energies observed
for each concentration are also presented in the same table.
Table 4.15 : Observed Ignition energies for 75% CH. + 25% H2 mixture
Gases CH, H, CH, H, CH, H, CH, H,
Targeted concentration (%) 6 2 8 3 10 3 12 4
Actual concentration of
6.01 2.01 8.03 2.64 10.1 3.39 12.42 4.28
flammable gas mixture (%)
Ratio (CH. : H2) 3.02 3.07 2.99 2.91

Mean ± std. dev. ±0.31 ±0.34 ±0.25 ±0.19
Total cone. (CH. + H2) 8.02 10.67 13.49 16.70

Mean ± std. dev. ±0.32 ±0.39 ±0.42 ±0.39
Equivalence ratio (ER) 0.67 0.93 1.20 1.54

Mean ± std. dev. ±0.31 ±0.34 ±0.25 ±0.19
Lowest Ignition energy for

0.58 0.10 0.52 5.19
gas mixture (mJ)
Range of Ignition energies 0.58 0.10 0.52 5.19

for gas mixture (mJ) 105.05 102.32 109.63 1011.74
121
4.5.2 Ignition probability of 75% CH4 + 25% H2 mixture
Ignition probabilities for the flammable gas mixtures were calculated in a similar way
to that used for the methane tests. The entire energy range (Minimum to Maximum)
was divided into bands of 2 mJ each. The ignition probability in each of these
individual bands was determined as described in Section 4.1.3 and the computed
values are presented in Table 4.16.
Table 4.16 : Ignition probability for 75% CH. + 25% H2 mixture
Total flammables Percentage ignition probability for energy bands
and CH.: H2 0·2 mJ 2·4 mJ 4·6 mJ 6·8 mJ 8·10 mJ 10·12 mJ
8.01 (6.01 :2.00) 5 94 99 99 99 99
10.67 (8.03: 2.64) 73 99 99 99 99 99
13.49 (10.10:3.39) 10 56 79 94 99 99
16.70 (12.42:4.28) 0 0 7 15 40 55
The coefficients of the exponential function (values 'b 'and 'Emin ') as described in
Equation 4.8 and the calculated ignition energy at 1% ignition probability using the
ignition energy is also shown for comparison with Efp=,% value.
Table 4.17 : Energy at 1% ignition probability for 75% CH. + 25% H2 mixture
Concentration Lowest observed Value Value Energy

ignition energy (mJ) 'b' E min (mJ) Efp=,% (mJ)
Total (CH.: H2)
8.02 (6.01: 2.01) 0.5840 1.0115 0.3499 0.2368
10.67 (8.03 : 2.64) 0.1049 1.603 0.1591 0.1591
13.49 (10.10 : 3.39) 0.5160 0.5135 1.0365 1.1608
16.70 (12.42: 4.28) 5.1900 0.1260 4.973 5.05278
The lowest experimentally observed ignition energies (column 2) are comparable

with the Etp=,% values, which are based on one in hundred chance of ignition.
122
0.8 ---- ------ -- ---- --- - - .. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
:a~ 0.6
~
o
ac
o • 1P1 (8.02%)
:2 0.4
c
!l' • 1P2 (10.67%)
,,1P3 (13.49%)
0.2 - - - - - --
x x 1P4 (16.70%)
-2 o 2 4 6 8 10 12 14 16
Energy (mJ)
for 75% CH, + 25% H2 mixture
A plot of ignition probability against ignition energy as observed for the gas mixture is
presented in the Figure 4.13 together with the exponential curves fitted to the data.
Similar to the methane tests there was an observed increase in the ignition
probability from lower to higher energies, with higher energies resulting in more
successful ignitions.
4.5.3 Gas with 50% methane and 50% hydrogen
flammability limit practically achievable was 4% f1ammables in air (2% methane and
2% hydrogen). It was quite evident that, since this concentration was below the LFL
given in Table 3.2 (4.4 %), ignition was not possible. Experiments performed with the
next practically achievable concentration, 6% f1ammables in air (3% methane and 3%
hydrogen) resulted in ignitions, hence this was considered as the lowest practical
lean concentration. The rich limit where no ignition were observed even after a
hundred attempts and the target concentration are presented in Table 4.18. Since the
flammability range was wider than for 75% CH, + 25% H2 gas mixture, five different
gas concentrations were targeted within this range.
123
Table 4.18 : Experimental conditions for 50% CH. + 50% H2 mixture
Concentration of flammable gas mixture
Gases CH, H, CH, H, CH, H, CH, H, CH, H, CH, H, CH, H,
End limits
2 2 12 12
(no ignition)
<---------------------------------------------->
Total
4 24
(CH, + H,)
Targeted
conc.
-- -- 3 3 5 5 7 7 8 8 10 10 -- --
Total
(CH, + H,)
--- 6 10 14 16 20 ---
ER 0.25 0.38 0.66 0.97 1.13 1.49 1.88
described in Section 4.1.1. A plot of observed energies, ignitions (Yes) and non-
ignitions (No) against concentration are presented in Figure 4.14.
1.
c.
c. ,
r!!. • A
.~
..
cc.
10 --------------------------------------------------.--------------
'\
: q, • 0
• •• ••.g
c . .~
-----.-.---------------------------------------------~-----------
o •• •
~. ~\
.~ ••• e
• 0 • •
.",.P • Cl Co • •• •
o rF
2 - - - .... - ..,- - - - - - -- - ~~-
0 ••
- - - -- --- - - - - - - - - - - -- - - - - - - - .... !""- ----------
._~_~ tr.. o cc ...... .
~ ~~ ~c ~~c '9Y
-2~--------~----------~--------~---------- __--------~
• 8 12 16
Total flammables cone. (%) of mixture (SO%CH4 + 50%H2)
20 2.
124
An arithmetic average of the total flammable concentration (50% CH. + 50% H2
mixture) was calculated for each nominal specified data set. and presented in
for each concentration are also presented in Table 4.19.
Table 4.19 : Observed Ignition energies for 50% CH. + 50% H2 mixture
. Concentration of flammable gas mixture
Gases CH. H2 CH. H2 CH. H2 CH. H2 CH. H2
Targeted conc. 3 3 5 5 7 7 8 8 10 10
Actual conc. of
3.01 3.13 4.99 5.04 6.80 6.68 8.09 8.07 10.10 10.00
flammable gases ..
Ratio (CH. : H2) 0.97 0.99 1.02 1.00 1.01

Mean ± Std. dev ±0.10 ±0.09 ±0.03 ±0.04 ±0.05
Total conc. 6.14 10.03 13.48 16.16 20.10

Mean ± Std. dev ±0.10 ±0.09 ±0.03 ±0.04 ±0.05
ER 0.39 0.69 0.93 1.15 1.50

Mean ± Std. dev ±0.03 +0.03 +0.03 +0.03 +0.06
Lowest Ignition
energy for gas 2.72 0.16 0.09 0.10 4.11
mixture (mJ)
Range of Ignition 2.72 0.16 0.09 0.10 4.11

energies for gas to to to to to
mixture (mJ) 13.34 3.16 2.31 2.98 11.40
4.5.4 Ignition probability of 50% CH4 + 50% H2 mixture
125
Table 4.20 : Ignition probability for 50% CH, + 50% H2 mixture
and CH,: H2 0-2 mJ 2 -4 mJ 4 - 6 mJ 6 - 8 mJ 8 - 10 mJ 10-12mJ
6.14 (3.01: 3.13) 0 11 29 43 54 71
10.03 (4.99: 5.04) 67 99 99 99 99 99
13.48 (6.80 : 6.68) 72 99 99 99 99 99
16.16 (8.09: 8.07) 65 99 99 99 99 99
20.10 (10.1 : 10.0) 0 0 11 18 38 50
The coefficients of the exponential function as described in Equation 4.8 (values

, b 'and' Em', ') and the calculated ignition energy at 1% ignition probability using the
equation 4.10 is presented in Table 4.21. The lowest experimentally observed
ignition energy is also shown for comparison with EIP=I% value. As can be seen they
are almost of the same magnitude.
Table 4.21 : Energy at 1% ignition probability for 50% CH, + 50% H2 mixture

ignition energy (mJ) 'b' Em', (mJ) E'P=I% (mJ)
Total (CH,: H2 )
6.14 (3.01: 3.13) 2.7209 0.1456 2.6353 2.7043
10.03 (4.99: 5.04) 0.1590 1.6527 0.2630 0.2690
13.48 (6.80 : 6.68) 0.0894 1.606 0.1633 0.1695
16.16 (8.09 : 8.07) 0.1044 1.6221 0.2190 0.2252
20.10 (10.1: 10.0) 4.1084 0.0975 4.1733 4.2764
A plot of ignition probability again energies as observed for the gas mixture is
presented in the Figure 4.15 together with the exponential curves fitted to the data.
126
O.B - -- -- - -- - - - - - - - - - - --- - - - - - - - - -- - - - - - - - - - - - - - --- - - - - - +-
..:s
>-
0.6 --------- ---------------------------- •----------7 --------
ie
.2
0/
______________________________• ________/ - _ _ _ _ .IP1 (6.14%)
~O.4
E' olP2 (10.03%)
.IP3 (13.48%)
0.2 ------- - - - - - - - - - -0- - - - - - - - - - - - - - - - - - -
x IP4 (16.16%)
• o o IP5 (20.10%)
-2 o 2 4 6 8 10 12 14 16
energy (mJ)
for 50% CH 4 + 50% H2 mixture
4.5.5 Gas with 25% methane and 75% hydrogen
flammability concentration practically achievable was 4% flammables in air (1%
methane and 3% hydrogen) as was the case with 75% CH, + 25% H2 mixture
described in Section 4.5.1. It was quite evident that, since this concentration was
below the LFL given in Table 3.2 (4.2%), ignition was not possible. Experiments
performed with the next practically achievable concentration, 8% flammables in air
(2% methane and 6% hydrogen) resulted in ignitions, hence this was considered as
the lowest practical lean concentration. The rich limit where no ignitions were
observed even after a hundred attempts and the target concentration as prepared for
the experimental runs is presented in Table 4.22. Five different gas concentrations
were targeted within this range, similar to the 50% CH 4 + 50% H2 mixture.
127
Table 4.22 : Experimental conditions for 25% CH. + 75% H2 mixture
Parameter Concentration of flammable gas mixture
CH, H, CH, H, CH, H, CH, H, CH, H, CH,

.
H, CH, H,
End limits
1 3 9 27
(no ignition)
<----------------------------------------------->
Total
4 38
(CH. + H2 )
Targeted
conc.
-- -- 2 6 3 10 4 13 5 16 7 20 -- --
Total
(CH. + H2 )
--- 8 13 17 21 28 ---
ER 0.17 0.36 0.60 0.83 1.08 1.57 2.34
described in Section 4.1.1. A plot of observed energies, ignitions (Yes) and non-
ignitions (No) against concentration are presented in Figure 4.16.
14
10 - - - - - -~
~~1.- ---------------------------------------------
·0
I ::. I
-~
00
0
. . . . -----
~ c ••
0 0 ·
-------~.-------------------------------------------~.--~-----
• Cl ~
o • •
i.s. -'2l
A-:
.....
• •• Cl Cl Cl Cl •
2 ---- --• _0 __ - - - _______ - ____ - - - - _ -______ - - - - -D- - - - - - - - - - - - - - - - - - - - - -
• 1:1 Cl Cl
-.r.
'. JJa ~·dJ
-2+-----------r---------~----------~----------~----------~
5 10 15 20 25 30
Total flammables cone. (%) of mixture (25%CH4 + 75% H2)
128
An arithmetic average of the total flammable concentration (25%· CH, + 75% H,
mixture) was calculated from each nominally specified data set and presented in
for each concentration are also presented in Table 4.23.
Table 4.23 : Observed Ignition energies for 25% CH, + 75% H, mixture
Parameter Concentration of flammable gas mixture
CH, H, CH, H, CH, H, CH, . H, CH, H,
Targeted conc. 2 6 3 10 4 13 5 16 7 20
Actual conc. of
2.27 6.16 3.S 10.14 4.S6 13.34 S.S9 1S.84 6.38 20.39
flammable gases
Ratio (CH, : H,) 8.43 13.64 17.90 21.43 27.22

Mean ± Std. dev. to.04 to.03 to.02 to.02 to.01
Total conc. 8.43 13.64 17.90 21.43 27.22

Mean ± Std. dev. to.32 to.41 ±0.47 to.49 to.S3
ER 0.40 0.67 0.92 1.16 1.S6

Mean ± Std. dev. to.03 to.03 ±0.04 to.OS ±O.OS
Lowest ignition
energy for gas 2.06 0.10 0.07 0.11 1.24
mixture (mJ)
Range of Ignition 2.06 0.10 0.07 0.11 1.24

energies for gas to to to to to
mixture (mJ) 12.42 2.S9 1.73 2.64 11.74
4.5.6 Ignition probability of 25% CH4 + 75% H, mixture
129
Table 4.24 : Ignition probability for 25% CH, + 75% H2 mixture

and CH,: H2 0-2 mJ 2 - 4 mJ 4-6 mJ 6 - 8 mJ 8 - 10 mJ 10 - 12 mJ
8.43 (2.27 : 6.16) 0 8 38 63 69 75
13.63 (3.50: 10.14) 69 99 99 99 99 99
17.90 (4.56: 13.34) 74 99 99 99 99 99
21.43 (5.59 : 15.84) 65 99 99 99 99 99
27.22 (6.84 : 20.39) 5 29 55 75 85 88
The coefficients of exponential function (values 'b 'and 'Em" ') as described in
Equation 4.8 and the calculated ignition energy at 1% ignition probability using
ignition energy is also shown for comparison with Etp:1%' As can be seen they are
almost of the same magnitude.
Table 4.25 : Energy at 1% Ignition probability for 25% CH, + 75% H2 mixture

ignition energy (mJ) 'b' Emin (mJ) E/p : 1% (mJ)
Total (CH,: H2)
8.43 (2.27 : 6.16) 2.0639 0.1531 1.5336 1.5993
13.63 (3.50 : 10.14) 0.1049 1.6131 0.1882 0.1944
17.90 (4.56: 13.34) 0.0663 1.5844 0.1161 0.1224
21.43 (5.59 : 15.84) 0.1115 1.6221 0.2190 0.2252
27.22 (6.84 : 20.39) 1.2358 0.2240 1.2063 1.2511
Similar to the methane tests there was an observed increase in the ignition
probability from lower to higher energies, as presented in the Figure 4.17. Higher
magnitude of energies resulting in more successful ignitions. The exponential curves
fitted to the data are also shown in Figure 4.17.
130
o
o
0.8 -- - -- --- --- --
o
~
:E 0.6 -------------------
III
.c
o • IP1 (8.43%)
a.
5 ----------------- • 1P2 (13.63%) _
E 0.4
c::
~
-----
•
.1'3(17.90%)
0.2 -- - --- --- -- ---- -- x 1'4 (21.43%)
o 1'5 (27.22%)
o •
-2 o 2 4 6 8 10 12
Energy (mJ)
for 25% CH 4 + 75% H2 mixture
4.6 Overall results

Sections 4.3 to 4,5 shows that an exponential function, Equation 4.8, can be used to
rep~esent the variation in ignition probability with energy for all the gas-air mixtures
studied. Values of band E mi , were identified for the gas-air mixture and for each set
of flammable gas concentration. The value of Emi , represents the hypothetical energy
for which there is no possibility of ignition (Ignition probability has theoretically zero
value). Any small increase in the ignition energy above E mi, gives a small probability
of ignition. Therefore, the values of Emin represent the hypothetical minimum ignition
energies for the fuel-air mixtures studied. The values of b, in combination with the
corresponding values of E mi, , define the shapes of the ignition probability-energy
curves.
The consistency of the experimentally obtained ignition probability was compared

with the calculated value of ignition probability using Equation 4.8. The values of
131
band E mi, utilised in Equation 4.8 are derived for each fuel-air mixture. This
comparison is shown in Figure 4.18.
•
0.8 - - - - -- - - - - - -- - - - -- - - - - --- -- --- - -- - - -- - - - - -- - - ----6- ----.--------
".
• ex
___________________________________ ~_O _______ Q ________ ________ _
6 x A •
;e
c
JP
A
• • CH4
"0 • 0
~ 0.4 ------------------------
x • 75CH4+25H2
o
1l x •
o A. 50CH4+50H2
0.2 ----------- -i:o - - - -- ........ - - - - - - -- - - - -- - - ---- - - - - - - --

X 25CH4+ 75H2
• •
~ o H2
x
O~~--~------_ _- - - -_ _----~====~
o 0.2 0.4 0.6 0.8
Calculated Ignition probablity
Figure 4.18 : Observed Ignition probability plotted against the calculated value
Figure 4.18 shows that for the range of gas-air mixtures studied, the ignition
probability is well represented by Equation 4.8 and that this equation could be used
to determine the ignition probability at a particular ignition energy if the relevant
values of band E mi , are known. This is further discussed in the next section.
4.6.1 Ignition energy
The overall experimental results for the variation of lowest ignition energy with
equivalence ratio for the range of flammable gases studied (100% CH 4 , 75% CH 4 +
25% H2 , 50% CH 4 + 50% H2 , 25% CH 4 + 75% H2 , and 100% H2 ) are presented in
Figure 4.19.
132
10
---0- Methane
-·0· - 75CH4 + 25H2
- -6- - SOCH4 + SOH2
.. ·x··· 2SCH4 + 75H2
-0- Hydrogen
0.01 .f---~--_--_---_--_--~--_--_----I
o O.S 1.5 2 2.5 3 3.5 4 4.

Equivalence Ratio
Figure 4.19 : Lowest ignition energy at various ER of flammable gases
The nature of the curve for all the gases indicates that, it may be possible to collapse
the data into a single curve. Accordingly the Equivalence Ratio (ER) and the ignition
energy for lean and rich flammable gas-air mixtures were normalised to form
dimension less parameters as follows:
Lean mixture
. ER-ER ... (4.18)

Normaltsed ER (NER) = MIE (when ER < ERM1E )
ERMIE - ERLFL
Rich mixture
ER-ERMlE ... (4.19)

Normalised ER (NER) (when ER > ERM1E )
ERUFL - ERMIE
Where
ERMIE Equivalence Ratio of the gas at the MIE value
ERLFL Equivalence Ratio at the Lower Flammability Limit
ERUFL ~quivalence Ratio at the Upper Flammability Limit
133
The Ignition Energy (lE) for each gas was normalised by using the MIE value as
follows:
. lE ... (4.20)
Normalised lE = -
MIE
The lean and the rich Equivalence Ratios with their corresponding ignition energy
values are considered separately. Logarithmic values of normalised lE are plotted
against the normalised ER (for lean values) for all the gases under consideration and
is presented in the Figure 4.20. All the test results of methane, hydrogen and the
methane-hydrogen mixtures (25%, 50% and 75% hydrogen in methane) were
collated and the data points' were plotted from the most lean mixture (ER < ERM1E ) to
ER equals ER M1E , i.e. NER between the range -1 to O. Mirror images of the lean
normalised ER values (i.e. for NER between the range 0 to 1) were then plotted
along with the normalised ignition energy. Curve fitting was done for the entire set of
data points (including the mirror image points) with an intercept at the origin. This
type of curve fitting achieves a zero slope at the origin. The lowest point on the curve
is described by 'NER = 0' and the corresponding ordinate of the lowest point i.e.
normalised ignition energy = 1 describes the MIE for the gas under consideration.
Ignition energy data at various ER values for Ethane and Propane gas (Lewis and
von Elbe, 1987) were also analysed in similar way. The normalised lE for Ethane and
Propane are also plotted and presented in the same Figure 4.20. The data analysis
for all these gases (CH., H2, CH. + H2 mixtures, C2He and C3Ha) suggest that the
ignition energy for flammable gases can be collapsed through such dimensionless
parameters and the data correlates well for the lean compositions.
134
--------
5
Lean gas mixtures o CH4, H2 & their mixtures
• Ethane & Propane
4 ---------------------------------------------------------------
y = 3.2342x' + 2E-17x
s o R'= 0.6729 o s
3 -------------------------------------------------------------
W 0
~2 _____ 9 _____________________________________ ____ _ ~
C • 0 0 •
.J
00 0
o __ o
_____ 0 _______________ _
o ---------------_0 ____ _ ~_. -,..,,-::~.-
-1~-----~--------- __
----------~----------~
-1 -0.5 o 0.5
Norm alis e d ER
Figure 4.20 : Normalised ignition energy vs normalised ER for lean gas composition
A second degree polynomial equation fitted for the lean values gives.
Ln(Normalised lE) = 3.2342 * NER * NER + 2 x 10-17 * NER
Normalised lE = exp (3.2342 * NER * NER + 2 x \0-17 * NER)
lE = MIE*exp (3.2342* NER* NER) ... (4.21)
A similar exercise was carried out for the rich mixtures and the variation of
normalised Ignition Energy
, against normalised ER is presented in the Figure 4.21.
The rich mixtures were plotted initially and then mirror images for the normalised ER
values were plotted. The best fit second degree polynomial equation for the rich gas
mixture is
Ln(Normalised lE) = 10.119* NER * NER +5 x 10-17 * NER
Normalised lE = exp (10.119 * NER * NER + 5 x \0-17 * NER)
lE = MIE*exp (10.119* NER* NER) ... (4.22)
The curve fitting for the rich values also gives a zero slope at the origin, similar to the
lean values. The ordinate of the lowest point (nomnalised ignition energy = 1)
describes the MIE for the gas under consideration.
135
----------
5
Rich gas mixtures
4 - - - - - - - ,,- -
y= 10.119.'+ SE-17.
R' = 0.8194
3 ---0------------------------0---
c c
'. -------------------- c c
c
:-.........r::__________________________ _
o
-, -0.5 o
Normalised ER
0.5
Figure 4.21 : Normalised Ignition energy vs normalised ER for rich gas composition
The trend line for the entire range (lean to rich mixture composition) is presented in
the Figure 4.22. Data points for Ethane and Propane gas are also plotted on the
same graph, suggesting good agreement with the correlated equation of the trend
line.
5
o Lean gas mix. .. Ethane & Propane
Cl Rich gas mix. - - Trend line
4 ------------------------------------------------ --a--------
o
3 ----------------------------------------LT---
o
o - - - - - - - - - <>_ - - - _ 0 _____ <f!'~_ __<J:cik'- ____ 51_ - - - _- - ___________ - --
-,~---------~----------~---------------------~
-, -0.5 o 0.5
Normalised ER
Figure 4.22 : Normalised Ignition energy vs normalised ER for all gases
136
A correlation to determine the minimum ignition energy in Equations 4.21 and 4.22
can also be obtained in terms of the concentration of hydrogen in the methane-
hydrogen mixture as presented in Figure 4.23.
0.3
o Experifrental tv1IE
y = O.2433e"o.0221x
R2 = 0.9507
" Reported MIE (a-14 )
0.2
o Reported MlE (H2)
~1 ------------------Q-----
o~------~------~------~------ ------~------~ __
-25 o 25 50 75 100 , 125
Hydrogen percentage In gas mixture (%)
Figure 4.23 : MIE variation for various concentration of hydrogen in mixture
The reported values of MIE for methane are from Lewis & Von Elbe, 1987 and
Bjerketvedt et. aI., 1997. An average value of the literature data is presented as
single value in the in Figure 4.23. Similarly for hydrogen the average value of MIE
from Lewis and D.W.V (German hydrogen association) data is presented. The
exponential curve fit line suggests an equation for MIE as follows:
MIE =0.2433 * exp (- 0.221 * CH') ... (4.23)
Where CH' is the percentage of hydrogen in the methane-hydrogen mixture.
A generalised correlation to estimate the ignition energy for a methane-hydrogen-air

mixture was derived from the test results and the range of observed data. The
equation for calculating ignition energy for a methane-hydrogen-air mixture (including
those of pure methane (when CH2 = 0) and pure hydrogen (when C H2 = 100) obtained
through the correlation is:
For lean gas mixtures
lE = MIE * exp(3.2342 * NER * NER) (when ER < ERMJE )
137
lE = 0.2433* [exp(- 0.221* CH ,)]* exp(3.2342 * NER* NER) ... (4.24)
For rich gas mixtures
lE = MIE*exp (10.119* NER * NER) (when ER > ER MlE )
lE = 0.2433*[exp(-0.221*CH ,)]*exp (10.119* NER* NER) ... (4.25)
Equations 4.24 and 4.25, above presents the generalised correlation based on the
experimental observations for the methane-hydrogen gas mixture. The equations can
also be used to calculate the ignition energy for pure gases methane and hydrogen
at various concentrations. However the equations are restricted to methane,
hydrogen and mixtures of methane and hydrogen.
The equivalence ratio at MIE for the flammable gases (methane, hydrogen, and
methane-hydrogen mixtures) was then correlated to the molecular weight of the
flammable gas. A plot of the Equivalence Ratio at MIE (ERM1E) against Molecular
Weight (MW) is presented in Figure 4.24 for the methane-hydrogen system studied.
y = 0.0014x' - 0.0307x + 1.081

R' = 0.8436
1.5
---- -~;-----........;..;~:...:..:..:..4.-.:..:..:.-
o 0
---:.:..:.:..:..:..:..:..:..-~-
--:..:..:..:-
A
--------------
e
0.5
o +-----------~------------~----------~-----------4
o 5 10 15 20
Molecular Weight
138
The value of ER at the MIE for the gases (methane, hydrogen and methane-
hydrogen mixtures) can be calculated from the Molecular Weight (MW) of the
flammable gas using the following equation.
ERMIE = 0.0014 * MW * MW - 0.0307 * MW + 1.081 . .. (4.26)
Equations 4.24, 4.25 and 4.26 are restricted to the methane hydrogen system.
Figure 4.25 presents analysis of the data considering ethane and propane along with
methane-hydrogen system.
y = O.0004x' - 0.012x + 1.0338

R' = 0.816
1.5 ------------------------------.-------.--------------.-----
~ g.~-_:;_'=ir.:..;;..:..:.:-r---:..:-:.,;;..;;;;..:;..-----,.-:~~~ -- ----
0::
w
1 -
~oQ6.
OB ------------------------------.-----------------.-----.----
O~------~--------~--------~------~------~
o 10 20 30 40 50
Molecular Weight
including ethane and propane gas
Based on Figure 4.25 the coefficients of Equation 4.26 modify to the following:
ERMIE =0.0004* MW* MW -0.012* MW + 1.0338 ... (4.27)
The analysis was extended further, to establish relationship between the minimum
ignition energy and the laminar burning velocity. The variation in MIE for gases were
plotted against the reported values of laminar burning velocity for hydrogen, methane
(Lewis and von Elbe, 1987) and predicted values (based on a correlation proposed
by Huanga, et. al. 2006) for methane-hydrogen mixtures. The reported values of
laminar burning velocity (Su) of fuel-air mixtures near the stoichiometric composition
139
were utilised. The best fit curve through the data points is presented in Figure 4.26.
Accordingly the equation to calculate MIE for these gases is
MIE = 0.0693 * SU-J.1028 ... (4.28)
0.4
o 0i4, H2 and their rrix.
o Other hydrocarbons
0.3 --- -0- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
D
'.E."
w
0.2
i
y = O.0693x·'·3028
0.1 --------------------------------R 2
-';' ---- -O~94i
D
O~----------~--------~----------~----------~
o 2 3 4
Laminar burning velocity (rnls)
Figure 4.26 : MIE variation with Laminar burning velocity
Other hydrocarbons ethane, propane, n-butane, pentane, hexane acetylene,

ethylene and propylene also satisfy the trend line equation as shown in the
Figure 4.26.
A generalised equation to calculate ignition energy for gas mixtures can be obtained
from Equation 4.21 and Equation 4.22 as follows:
lE = 0.0693* SU-J.3028 *exp (3.2342* NER* NER) (when ER < ER'flE) ... (4.29)
lE = 0.0693 * SU-1.3028 • exp (10.119 * NER * NER) (when ER > ER M,E ) ... (4.30)
A plot of ignition energy as calculated through the equations above and the actual
observations is presented in Figure 4.27. The actual observation agrees well with the
correlated equations for all the gases methane, hydrogen and methane - hydrogen
mixtures. Good agreement is also obtained for ethane and propane. The lines are
plotted as per the equations for ignition energy (Equation 4.29 and Equation 4.30)
and the points are the actual experimental observations.
140
1000 --~thane
, - . _. - 75%CH4+25%H2
i I - - - - 50%CH4+50%H2
100 - - - - -
i: J
r-;-- -:.,----------------.---.-------. -----.... _.. -. 25%CH4+75%H2
j I :' A --H~r0gen
;:,
I
:' I
.• --Ethane
. i
,I i - ~ • Propane
/ /
0' CH4
o 75%CH4+25%H2
o 50%CH4+50%H2
• x 2S%CH4+75%H2
• • H2
• Lewis (CH4)
0.1
• Gexcon (CH4)
• lewis (H2)
At. OW.V(H2)
0.01 +---_---_---_--_---_---_--_-1 + C2H6
o. 2 3 4 5 6 7
equivalence Ratio x C3H8
Figure 4.27 : Ignition energy for gases at various Equivalence Ratios
The consistency of the approach taken to determine the minimum ignition energy and
the variation in lowest ignition energy with equivalence ratio is illustrated by plotting
the observed and predicted ignition energy. This comparison is shown in Figure 4.28.
10
o
o
:;- -----~---------------
.§.
!!!
"~
"
B _0 __ 0 ___________________________________ _
0 0.1
0.01 -JL.--------~--------~--------_1
0.01 0.1 10
Precdicted lE (mJ)
Figure 4.28 : Observed and predicted values of Ignition energy
141
4.6.2 Prediction of Ignition probability
After the success of the use of the correlations obtained from the experimental
results to predict the ignition energy illustrated above, attention was given to the
ignition probability. The relationship between b, Emin (energy at zero ignition
probability) and the lowest ignition energy (as referred in Equation 4.8) was
examined.
3
y = 0.3209><,,·1235
R' = 0.9182
- --~---------------------------------------------------
o
__
o1-~ ~O~=v~::~:~v::::~::::::::~!::::=:=~~~____~
2 3 4 5 6
Lowest Ignition energy (mJ)
Figure 4.29 : Variation of parameter 'b' with Lowest Ignition energy
As shown in Figure 4.29, the relationship between the value of 'b' and the lowest
ignition energy is given by:
b = 0.3209 * IE(-O·7235) ... (4.31)
6
y = 0.95x
5 _______________________________ ~_ ~9·1I~!!8_ - -- ---- ___ ~ _____ --
4 --------------------------------------
o
-----------------------------------------------
o~~----~----~------~------~------ __------~I
o 2 3 4 5 6
Lowest Ignition energy (mJ)
Figure 4.30 : Variation of parameter 'Emin' against Lowest Ignition energy
142
The relationship between Em;n and the lowest ignition energy (as shown in
Figure 4.30) is:
Emm = 0.95 * lE ... (4.32)
The Ignition probability (IP) equation after substitution of parameters 'b' and 'Em;n' is
IP = 1- expl- 0.3209 * IE-<J7215 * (E - 0.95 * lE)]
Referring to Equation 4.21 and 4.22 ignition probability can be calculated as
For lean mixtures (ER < ERM1E )
lE = MIE * exp (3.2342 * NER * NER)
Hence
IP = I-ex [- 0.3209* {MIE* exp (3.2342 * NER * NER)}-<J7215] ... (4.33)

p * (E - 0.95 * {MIE * exp (3.2342 * NER * NER)})
Where NER = ER - ERMIE

ERMIE - ERL'L
ERMIE = 0.0014* MW* MW -0.0307* MW +1.081

MIE = 0.0693* SU-I.J02'
For Rich mixtures (ER> ER M1E )
lE = MIE*exp (10.119* NER* NER)
Hence
IP = 1- ex [- 0.3209 * {MIE * exp (10.119* NER * NER)}-<J7215]

p * (E -0.95 * {MIE* exp (10.119* NER * NER)}) ... (4.34)
143
ERuFL - ERMIE
ERUFL Equivalence Ratio at the Upper Flammability Limit
MIE and ERMIE has same meaning as referred for lean mixtures.
As per Equation 4.33 and 4.34 above, IP is a function of Energy (E in mJ) and
Equivalence Ratio (ER) at lean and rich flammable gas concentrations.
Accordingly in this section a generalised correlation for the ignition energy (Equation
4.21 and Equation 4.22) and Ignition probability (referred above) was developed from
the test results of the flammable gases. The generalised ignition energy equation
was also found useful for other flammable gases like ethane and propane as
presented in the Figure 4.22. Ignition probability can not be compared for other
gases, since data for other gases are not available. The overall result of the ignition
energy tests are further discussed based on ignition probability at various energies
and the impact of hydrogen on MIE of gases.
4.6.3 Ignition probability computation
The Ignition probability calculated from the Equation 4.8 for fixed energy levels (2 mJ,
5 mJ and 10 mJ) and at a three different equivalence ratios (lean, near stoichiometric
and rich) for the flammable gases are presented in Table 4.26.
144
Table 4.26 : Ignition probability of flammable gases at various energies
Ignition probability at energies of

Gases Equivalence ratio
2mJ SmJ 10 mJ
Methane 0.17 0.58 0.87
75% CH, + 25% H2 ~

0.81 0.99 1.00
50% CH, + 50% H2 ER = 0.67 0.94 1.00 1.00
25% CH, + 75% H2 0.95 1.00 1.00
Hydrogen 1.00 1.00 1.00
Methane 0.94 1.00 1.00
75% CH, + 25% H2 0.95 1.00 1.00
50% CH, + 50% H2 ER = 0.93 0.95 1.00 1.00
25% CH, + 75% H2 0.95 1.00 1.00
Hydrogen 1.00 1.00 1.00
Methane 0.31 0.68 0.91
75% CH, + 25% H2 0.40 0.87 0.99
50% CH, + 50% H2 ER = 1.2 0.95 1.00 1.00
25% CH, + 75% H2 0.95 1.00 1.00
Hydrogen 1.00 1.00 1.00
The ignition probability for all gases increases with energy across all the lean
(ER = 0.67) to rich (ER = 1.2) limits. Methane ignition probability varies a large
amount from 17 % to 87 % (70 % variation) for ·energies from 2 mJ to 10 mJ
respectively. This observation justifies that sources with energies around 2 mJ can
result in only few ignitions of the released methane gas, but with higher energy
sources around 10 mJ, ignition is virtually guaranteed as shown in Figure 4.31.
145
1.2
¥ . . ___ .-.- "'" ::::.-:.::::-::~-B-"---" .... ....-.-.....-------.... ------.----..... -11(

/'
;
... 0.8 .... ~'
=
:s
'"
.c
0~ ER= 0,67
Co
C
0
:2 --+-ch4
c
~ 0.4
- ·.·-7SC+2SH
- ...... -SOC+SOH
.. ·x··· 2SC+7SH
0
0 2 4 6 8 10 12
Energy (mJ)
Figure 4.31 : Ignition probability variation for flammable gases at ER = 0.67
Hydrogen is excluded from Figure 4.31, Figure 4.32 and Figure 4.33 since its ignition
probability (at energy levels 2,5 & 10 mJ) is unity. (see Table 4.26.)
The wide range in ignition probability is narrowed down to a very small percentage
change of 5 to 6% (94% to 100 % for methane and 95% to 100% for methane -
hydrogen mixtures) when the gases approach the stoichiometric concentration.
Ignition probability at close to a stoichiometric mixture (ER = 0.93) increases rapidly
for all gases under consideration as compared to lean mixtures. (Figure 4.32)·
146
1.2
-s' H H
IS'~-
=0.8
~
:E
co
.c
~ ER= 0.93
Q.
I:
.,
0 ~ch4
]'0.4 -
- ' .·-75C+25H
- --..- - 50C+50H
-- -x- -- 25C+75H
0
0 2 4 6 8 10 12
Energy (mJ)
Figure 4.32 : Ignition probability variation for flammable gases at ER =0.93

Ignition probability for rich mixtures (ER = 1.2) again expands to a range of about
60% wide variation (31% to 91% for methane, 40% to 99% for 75% CH, + 25% H2
mixture) for energies of 2 mJ to 10 mJ. The observations are presented in
Figure 4.33. The variation in ignition probability is only 5% (not significant) for the
other two gases (50% CH, + 50% H2 mixture and 25% CH, + 75% H2 mixture).
1.2
~O.8
~
~
~
ER =1.2
a.
.
/
6 -+-ch4
:e
c
.!!' 0.4 . . .. _._.... .. -- -_.. __ . .
- .•. -75C+25H
- --..- - 50C+50H
... x.. - 25C+75H
O+---__--~------~--------~--__--~------~--____~
o 2 4 6 8 10 12
Energy (mJ)
Figure 4.33 : Ignition probability variation for flammable gases at ER = 1.2
147
Addition of hydrogen increases the ignition probability values for all gases under
consideration. The next section exclusively discusses the impact of increasing
hydrogen concentration on various gas characteristics related to ignition energy and
ignition probability.
4.6.4 Impact of hydrogen
Hydrogen minimum ignition energy is about 1/15th that of methane, and the lowering
of the ignition energy for methane-hydrogen mixtures is quite evident if an increasing
concentration of hydrogen is injected into methane. The experiments have enabled
these variations in the ignition energy to be quantified. The first observation was the
lowering of the lowest ignition energy as the hydrogen content was increased. A plot
of lowest ignition energy against total flammables concentration is presented in
Figure 4.34. The shift of the U shape curve towards the right for each of the gas
mixtures under consideration, shows that the most easily ignitable mixtures have
increasing concentrations of hydrogen (total flammables gas). The bottom of the U
curve is brought down with increasing addition of hydrogen, suggesting the lowering
of the MIE values for the flammable gases under consideration. The width of the U
curve, i.e. the rage of concentrations over which ignitions were achieved, increases
with increasing concentration of hydrogen.
100
---0-- Wlethane
- .0 .- 75CH4 + 25H2
- -t;,,- - 50CH4 + 50H2
1
..J 0.1
0.01 +----~---~---~---~---~---~--__4
o 10 20 30 40 50 60 70
Total flammable gas concentration (o/~
Figure 4.34 : Lowest ignition energy at various flammable concentrations
148
The variation in ignition energy at 1%, 50% and 100% ignition probability was studied
for increasing addition of hydrogen. The ignition energies for different methane -
hydrogen mixtures were compared at ER of 0.67, 0.93 and 1.2. The result of the
comparison is presented along with experimental observations of ignition energy in
Table 4.27. Ignition probability increases for increasing ignition energy for all the
gases under consideration. Increasing the concentration of hydrogen reduces the
ignition energy, for all probability values under consideration.
Table 4.27 : Ignition energy at various ignition probabilities
Ignition energy at various Observed

Equivalence probabilities (mJ) Lowest
Gases ignition
ratio energy
IP =1% IP =50% IP =100%- (mJ"i"
Methane 1.2243 4.2138 41.4881 1.8163
75% CH. + 25% H2 0.3499 1.0253 9.4456 0.5840
50% CH. + 50% H2 ER =0.67 0.2691 0.6824 5.8359 0.1590
25% CH. + 75% H2 0.1944 0.6179 5.8979 0.1049
Hydrogen 0.0303 0.1379 1.4799 0.0365
Methane 0.3933 0.7947 5.7998 0.2242
75% CH. + 25% H2 0.1591 0.5852 5.8985 0.1049
50% CH. + 50% H2 ER =0.93 0.1696 0.5949 5.8983 0.0894
25% CH. + 75% H2 0.1224 0.5536 5.9292 0.0663
Hydrogen 0.0303 0.1379 1.4799 0.0244
Methane 0.5934 3.2881 36.8865 0.8349
75% CH. + 25% H2 1.0365 2.3668 18.9533 0.5160
50% CH. + 50% H2 ER =1.2 0.1877 0.6117 5.8979 0.1044
25% CH. + 75% H2 0.2252 0.6463 5.8970 0.1115
Hydrogen 0.0303 0.1379 1.4799 0.0244
(* Ignition ener9Y values are calculated for probability of 99.99%)
149
Ignition probability was quantified for the increased addition of hydrogen. The
variation in the experimentally observed lowest ignition energy against hydrogen
concentration (for different ER) is presented in Table 4.27 and Figure 4.35. The
lowest ignition energy for the lean concentration (ER = 0.67) is reduced by 1.24 mJ
(from 1.82 mJ to 0.58 mJ) when 25% of the methane is replaced by hydrogen.
Further increases in the proportion of hydrogen does not bring further appreciable
lowering of the ignition energy. At a rich concentration (ER = 1.2) ignition energy
changes are quantitatively smaller in magnitude as compared to a lean concentration
(ER = 0.67). The ignition energy change is smallest for ER=0.93 (near stoichiometric
mixture) across the entire range of increasing concentration of hydrogen (ref. Table
4.27 and Figure 4.35).
2
~ ER"'O.67
1.5 -- ------------------------------------------ ___ ER=1.2
~
!
•g 1 -------- -----------------------------------------------
"g
c
;;
~ 0.5
o L"::"'"
o
.." "'=.:""".
=:='
25
====~'·'==-·"
50
.
. "-~:::="··-,~,,-
. ,-d
75 10
Hydrogen percentage In methane ("10)
Figure 4.35 : ·Lowest ignition energy varation with increasing hydrogen percentage
The variation in the ignition energy and the associated ignition probability for the
flammable gases with addition of increasing concentrations of hydrogen, is likely to
change the probability of fire and explosion accidents arising as a result of gas
releases. Overviews of these changes along with the main findings of the study are
discussed in the next chapter.
150
Chapter 5
Conclusions of the Experimental study

In this thesis minimum ignition energy (MIE), lowest ignition energy at various
concentrations and ignition probability are presented and discussed for methane-air,
hydrogen-air and methane-hydrogen-air mixtures. The addition of hydrogen to the
gas characteristics (ignition energy and ignition probability) is also discussed with an
observed and predicted ignition probability for each methane-hydrogen mixture.
Hydrogen, if injected into the natural gas infrastructure (natural gas considered as
methane only) is likely to result in many changes to the gas industry and also to the
end user. As a part of the safety work package of the Naturalhy project, the objective
was to determine experimentally the ignition energy for methane-hydrogen-air
mixtures, along with ignition energies for methane-air and hydrogen-air mixtures at
ambient temperature and pressure. Ignition probabilities associated with the ignition
energy values were also determined from the observed results.
MIE's for most of the flammable hydrocarbon gases are very small (0.3 mJ and less).
Measurement of the MIE is done through various techniques; Capacitive spark
discharge; Laser; and energy delivered by passing electric current through a wire
placed in a flammable gas-air mixture (Martin Hattwing, 2004 and Lewis 1987).
Komai et. aI., 1994 has carried out experiments of igniting flammable gases
(methane, hydrogen etc.) in air using frictional sparks from light alloys. The aim of the
experiment was to" examine the incendivity of the sparks and to determine the safety
associated with handling flammable and explosive gases.
Capacitive spark discharge utilised during this work has been the preferred method
for determining MIE for gases. The stored capacitor energy prior to the electrical
discharge (spark) is considered as the energy delivered to the gas-air mixture. MIE is
evaluated from the quantified spark discharge energy and success of ignition. The
method is also standardised by British, European (BS EN 13821 - 2002) and
American standards (ASTM E 582 -1988).
151
5.1 Main findings of the study
5.1.1 Minimum Ignition Energy
The experimental test runs has provided information on the MIE of the flammable
gases under consideration. The observed MIE for methane (0.22 mJ) is marginally
less than the historical data (0.29 mJ) reported by Lewis and von Elbe, 1987 and
Bjerketvedt et. aI., 1997 (Gas explosion handbook). Concentration of methane
corresponding to the MIE is reported as 8.5% by Lewis and von Elbe, 1987 and 9%
by Bjerketvedt et. al., 1997. This work suggest MIE (0.22 mJ) at a concentration of
9% methane in air. A methane concentration of 8% resulted in a lowest ignition
energy of 0.24 mJ, which is less than the reported MIE values. This lowest ignition
energy value at 8% concentration cannot be described as the MIE because of the
definition of MIE. The value of MIE obtained for hydroge:n of 0.024 mJ is in
agreement with the German Hydrogen Association (D.W.v) value of 0.029 mJ, and
is slightly greater than the Lewis and von Elbe value of 0.019 mJ.
MIE values for other gases (various methane-hydrogen mixtures) were also
determined and found to be between those of methane and hydrogen MIE results.
The most easily ignitable mixture (the concentration at which the MIE is determined)
has an increasing concentration of total flammables, with the addition of hydrogen.
Increasing concentration of hydrogen has the effect of lowering the MIE values of the
gases. The correlation obtained from the test results as presented in Figure 4.23 is :
MIE = 0.2433 * exp (- 0.221* CH') ... (5.1)
Where CH' is the percentage of hydrogen in the methane-hydrogen mixture.
The variation in MIE for gases plotted against the reported values of laminar burning
velocity for hydrogen, methane (Lewis and von Elbe, 1987) and predicted values
(based on correlation proposed by Huanga, et. al. 2006) for methane-hydrogen
mixtures are presented in Figure 4.26. The relationship obtained between the MIE
and the laminar burning velocity (Su) is :
152
MIE = 0.0693 * SU-1.l028 ... (5.2)
Where Su Laminar burning velocity (m/s)
Equations 5.1 and 5.2 provide means of predicting the MIE for either methane,
hydrogen, or methane-hydrogen mixtures.
5.1.2 Ignition Energy
The lowest observed ignition energy for methane and hydrogen within the flammable
range agree reasonably well with the reported data, except at the leanest mean"
concentration of methane. The lowest ignition energy for methane-hydrogen-air
mixtures decreases with increasing concentration of hydrogen in the mixture. These
values cannot be compared to the literature, because of the absence of any previous
work on methane-hydrogen mixtures. The lowest ignition energies have a 'U' shape
dependence on concentration. The characteristic U shape curve was observed for all
the gases under consideration (see Figure 4.27). Ignition energy increases
asymptotically at the lean and rich limits of flammability. The nature of the curve for
all the gases indicated that, it may be possible to collapse the data into a Single
curve. Accordingly the Equivalence Ratio (ER) and the Ignition Energy (lE) was
normalised as presented in Equations 4.18 to 4.20. The Normalised Equivalence
Ratio (NER) and Normalised Ignition Energy were correlated as presented in
Equations 5.3 and 5.4.
For lean gas mixtures (ER < ERM1E )
lE = MIE*exp(3.2342 * NER * NER) and NER = ER - ER"fE ... (5.3)

ER"IE - ERLFL
For rich gas mixtures (ER> ERM1E )
IE=MIE*exp(IO.1l9*NER*NER) and NER=' ER-ER"IE ... (5.4)

ERuFL - ER")E
Where MIE Minimum Ignition Energy for flammable gas (mJ)
ER")E Equivalence Ratio of the gas at the MIE value
153
ERMIE was correlated to the molecular weight (MW) of the gas as presented in
Equation 5.5 (for methane, hydrogen and their mixtures) and Equation 5.6 (for
methane, hydrogen, methane-hydrogen mixture, ethane and propane).
ERM1E = 0.0014 • MW • MW - 0.0307 • MW + 1.081 . .. (5.5)
ERM1E =0.0004' MW' MW -0.012' MW +1.0338 ... (5.6)
The Ignition energy predicted through Equations 5.3 and 5.4 was compared to the
experimentally observed lowest ignition energy. The correlated and the observed
values agree reasonably well with the reported data (for all the gases) as presented
in Figure 4.28. Thus, the experimental data collated in this thesis, has enabled
relationships to be developed that allows the lowest ignition energy of methane-air,
hydrogen-air, methane-hydrogen-air mixture, ethane-air and propane-air mixtures to
be calculated over their flammable ranges.
The ignition probability for methane increases with increasing (spark) energy. The·
gradient of ignition probability against energy is greatest when the fuel-air ratio of the
mixture is close to stoiciometric. The gradient is least when the mixture is lean or
rich.
Ignition probability for hydrogen also increases with increasing energy.

Concentrations of 20% to 30% (hydrogen percentage in air) have a very rapid
increase in ignition probability with energy. The rate of increase (in ignition
probability) with energy for 20%, 25% and 30% can not be differentiated from each
other. Lean and rich mixtures have sloy;er rates of increase in ignition probability as
compared to a near stoichiometric mixture. Hydrogen ignition probability was
observed to be always higher than methane, when compared at similar energy
levels.
Ignition probability for all the gases increases exponentially with energy ultimately
reaching a value of unity described by Equation 4.8.
154
IP = 1- exp[- b(E - E min )] ... (5.7)
The coefficients of the ignition probability equation 'b' and 'Emin' (Energy at zero
ignition probability) determines the nature of the ignition probability curve.
Coefficients band Emin each were correlated to the lowest ig nition energy as follows :
b = 0.3209 * lE(-O·1235) ... (5.8)
Emin = 0.95 * lE ... (5.9)
The Ignition probability (IP) equation after substitution of parameters 'b' and 'Emin' is
IP = 1- expl- 0.3209 * lE-O· 7235 * (E':' 0.95 * lE)] ... (5.1 0)
IP as observed for methane-air, hydrogen-air and methane-hydrogen-air mixtures

was ultimately correlated to equivalence ratio (ER) and source energy (E). A
generalised correlation to calculate 'lP' from the experimental result is :
For lean mixtures (ER < ERM1E )
lE = MIE * exp (3.2342 * NER * NER)
- 0.3209 * {MlE *exp (3.2342 * NER * NER)}-O.7235]

IP=I-exp . ... (5.11)
[ * (E - 0.95 * {MIE* exp (3.2342 * NER * NER)})
ER-ER
Where NER = . MIE
ERM'E - ERLFL
ERMIE = 0.0014* MW*MW -0.0307* MW +1.081

MIE = 0.0693 * Su -1.3028
155
For Rich mixtures (ER > ER MIE )
lE =MIE*exp (10.119* NER* NER)
IP = I-ex [- 0.3209 * {MlE * exp (10.119* NER * NER)}-<J7235]

... (5.12)
P * (E - 0.95 * {MIE * exp (10.119* NER * NER)})

ERuFL - ERMIE
MIE and ERMIE has the same meaning as referred for lean mixtures
The IP relationships help in predicting the ignition probability for methane-air,

hydrogen-air and methane-hydrogen-air mixture from a knowledge of equivalence
ratio (ER) and source energy (E). It may be possible to extend the relationships to
include other flammable gas such as ethane-air and propane-air mixtures by
substituting ERMIE value (Equation 5.6) in the Equations 5.11 and 5.12. However,
experimental data on the ignition probability of these gases is not available to confirm
whether or not this approach is reasonable.
Increasing concentration of hydrogen in methane-hydrogen mixtures decreases MIE

and Ignition energy for the gas mixture. An increase in ignition probability is observed
due to increasing the hydrogen content of the gas mixture. It can be expected that
such changes will result in an increase in incidents of fire ~!nd/or explosion. Chapter
6, discusses this issue, using deterministic values of leakages (for methane and
hydrogen) and ignition source energies. A quantified change in the gas incident
probability from existing knowledge of gas release frequency is presented in
Chapter 6.
156
Chapter 6
Changes in gas incident probability

Introduction of hydrogen into the natural gas infrastructure will change the gas
properties, which were discussed in earlier chapters. The experimental results
indicate that there is greater risk of ignition of hydrogen-air mixtures, compared to
methane-air mixtures. The ignition risks are compared for methane and hydrogen
based on the experimental observations of minimum ignition energy (MIE), lowest
ignition energy at various concentrations and ignition probability.
6.1 Comparing methane and hydrogen releases
Comparison of the probability of ignition of gas releases was done using methane
and hydrogen. Properties of other gases (various mixtures of hydrogen and methane)
were considered to be somewhere between these two extremes. Comparison of the
results for hydrogen and methane show that, hydrogen has an increased flammability
range, a lower ignition energy and a higher ignition probability. Further comparison
based on the energy density shows that hydrogen energy (120 MJ/kg) is more than
twice that of methane (50 MJ/kg). Thus, hydrogen has a somewhat higher
flammability hazard and likely to deliver more destructive energy for equal quantum
(1 kg) of flammable gas oxidised/consumed. Swain et. aI., 2005 expressed the view
that hydrogen disperses so quickly that it is unlikely to create a flammable mixture of
any large volume. Swain's experiments were carried out in open space. In confined
spaces, such as homes, garages, etc., hydrogen will not disperse so readily and the
buoyancy of hydrogen will have less effect, if it releases inside domestic properties.
This is the major concern of this work and the effect of adding hydrogen to natural
gas on the frequency of fires and explosions. Hence, this analysis is based on
confined spaces of domestic gas releases and its fire and explosion potential.
Gas ingress within the domestic property is considered as the starting point of this
analysis. The gas release can have an origin within or outside the house. The
analysis is based on experimental studies carried out for natural gas releases
157
(HSE, 1998). Hydrogen gas releases are considered to occur under similar
conditions, but with a density correction factor to calculate the hydrogen flow rate
from the corresponding methane release rate.
Three methane gas releases; minor, significant and major 0.36 m 3/hr, 1.44 m3/hr and
7 m3/hr, are considered in the analysis, in line with the HSE eRR report 168, 1998.
The minor release corresponds to the gas flow rate from a single gas burner on a
domestic gas cooker. The significant release condition is, when all four burners on a
cooker are ON. The major release condition is corresponding to the maximum likely
leak from a single open end carcass/internal pipe, within the house downstream of
the meter. Typical values of the background ventilation rate of 0.6 air changes per
hour and a room volume of 25 m3 is considered as per the HSE report. Conditions of
perfect mixing are assumed within the room to calculate the steady state gas
concentration. The rate of build up of the gas concentration to the steady state level
as a functio~ of time is presented in Figure 6.1. The actual values of the maximum
steady state concentration calculated under these typical conditions of room volume
and ventilation rates are presented in Table 6.1. The ignition probability for the
energy level of 2 and 5 mJ (energy delivered by various ignition sources present
within the house) are also presented in Table 6.1. Ignition sources have energies far
greater than 2 - 5 mJ, but it's the energy delivered to the gas that is important for
successful ignition, hence the lower values considered in the analysis. Experiments
in which gas concentrations build up within a particular room, also lead to gas-air
mixture accumulation in an adjacent room, even if an interconnecting doorway is
closed (one of the major findings of the HSE, 1998 report). This practical observation
restricts us from reducing the ignition probability value, when. gas concentration
surpasses the stoichiometric concentration. Accordingly under practical situations the
ignition probability can not be considered as negligible (zero) at gas concentration
exceeding UFL values. A conservative approach of risk assessment is adopted here.
If the room concentration exceeds the stoichiometric concentration, ignition
probability is considered as the maximum value observed at the stoichiometric
concentration, for the gases under consideration (either methane or hydrogen in this
present case).
158
Table 6.1 : Ignition probability for methane releases within confined areas
Percentage ignition probability at energy

Methane release m3/hr Conc.(%)
2 mJ 5 mJ
Minor 0.36 2.34 0 0
Significant 1.44 8.8 49 100
Major 7.0 31.82 49 100
Hydrogen is 1/8th times less dense than methane, accordingly the volume leakage
rate for hydrogen is expected to be 2.8 times that of methane. If, when hydrogen is
used to replace some of the natural gas, the pressure is increased in order to supply
an equivalent amount of energy, then this factor would be even greater. This,
coupled with lowering the ignition energy for hydrogen results in increased ignition
probability for all categories of releases; minor, significant and major cases under
consideration. The minor release of hydrogen is likely to form a flammable cloud
(with concentration of 6.4% hydrogen in air, inside the room), which was not the case
with a minor release of methane (concentration of 2.3% methane in air was achieved
under identical conditions, inside the room). The results of the deterministic analysis
are presented in Table 6.2.
60
rrajar leak H2 - _ . ""jar leak CH4
significant leak H2 - • - . - signtificant leak CH4

mnor leak 1-12 .. .. .. .. mnor leak Qi4
LFLH2 - - LFLCH4
---'- ------
.- -----_._._---
~--
-" _. ---- -' - . _.----_.---------
........
._.....J::~=-·=--a -: ""_ ~a-:a ~a·.
~..- ~ a-a
O~~~~~~~~~~-------------~----------------~
"a-a -aaa .- .- ... - .... -- .... _. -- ..................
-a-a •• .... - .. .... .. ..
o Time (hr) 2 3
Figure 6.1 : Gas concentration built up for methane, hydrogen release in room
159
Table 6.2 : Ignition probability for hydrogen releases within confined areas
Percentage ignition probability at energy

Hydrogen release m3/hr Conc.(%)
2 mJ 5 mJ
Minor 1.02 6.4 3 8
Significant 4.07 21.4 100 100
Major 19.80 56.9 100 100
Comparison of Table 6.1 and Table 6.2 for Methane and Hydrogen, (using the
deterministic releases of flammable gases suggest some understanding of the
increase in the ignition probability. For minor releases, the rise in ignition probability
was estimated to be 3%, for an ignition source delivering 2 mJ energy to flammable
gas-air mixture. If the source energy is increased to 5 mJ, then the probability
increases to 8%. Significant and major releases result in an increase in ignition
probability from around 50 % to 100 %, (i.e. almost twice the number of cases with
hydrogen will result in fire and/or explosion as compared to the base case) for a 2 mJ
energy. With higher energies (> 2 mJ) the ignition is always certain (ignition
probability is unity) for significant and major releases, hence further increase in
ignition probability because of fuel change is ruled out. This analysis predicts
quantitative increase in the ignition probability when domestic fuel (flammable gas) is
changed from methane to hydrogen.
In real world situations successful ignition of released gas is dependent on many

factors. Thus ignition probability is dependent on numerous parameters. In the
present analysis, ignition probability is considered dependent on probability of
flammable vapour cloud formation and then ignition of this vapour cloud using
surrounding ignition sources. The probability of formation of a flammable cloud is
calculated from the leakage and ventilation rates, which were deterministic in the
above analysis. Probabilistic consideration of leakage and ventilation rates within the
residential area will help to arrive at the probability of flammable cloud formation.
Experiments with common ignition sources within a flammable gas air mixture, will
determine the ratio of source energy and the delivered energy. These studies will
reduce the uncertainties associated with the determination of ignition probability. The
same methodology can then be applied to estimate the quantitative increase in the
ignition probability with the change of domestic fuel from methane to hydrogen.
160
Chapter 7
Issues for future work

The experimental facility used to carry out the ignition energy measurements for
methane-air, hydrogen-air and methane-hydrogen-air gas mixtures was designed
and built at the Department of Chemical Engineering, Loughborough University. The
rig is equipped to handle other flammable gases e.g. propane, butane simply by
replacing the feed gas. Accordingly, ignition energy, ignition probability and minimum
ignition energy for other flammable gases and mixture of gases can be determined
using the existing experimental facility. A few of the limitations of the existing
experimental arrangement are discussed in details in Section 7.1. Issues for future
work are discussed in Section 7.2.
7.1 Limitations of the experimental setup
The operating procedure for the experiments was semi-automatic where the
preparation of the gas mixture (valve operation) was done manually and the ignition
process, with subsequent data recording, completed automatically.
The mixtures for the exeriments involving mixtures of methane and hydrogen were
prepared inside the vessel by the method of partial pressures, by first injecting one of
the gases (either methane or hydrogen) and then adding the second gas as required
for the targeted concentration. The recorded pressure of the first and second gases,
with a consistent pressure difference (between the two gases), gives the flammable
gas mixture concentration to be prepared. The lowest ignitable (practically
achievable), gas mixture of 25 % hydrogen in methane and 75 % hydrogen in
methane were prepared with the lean limit of 8% flammables in air. The lean limit for
these gases can be far below this 8% concentration. The existing method of
preparing the gas mixture from individual gases, prevented, experiments to be
performed below 8% concentration (for 75%CH. + 25%H 2 and 25%CH. + 75%H2).
This technique also meant that the ratio of methane to hydrogen was likely to be
slightly different for each test. Therefore, it is suggested that, rather than mixing the
flammable gases .directly in the combustion chamber, premixed gases should be
used.
161
The experimental setup was designed and fabricated in order to produce sparks of
very low energies to investigate the ignition energy and the ignition probability for
flammable gas air mixture at atmospheric pressure. The energy delivered by the
capacitive spark is considered as the ignition energy during the test runs. The High
Voltage (HV) transformer inside the Extra High Tension (EHT) pulse generator
determines the maximum voltage obtained during the spark process. The electrical
circuit for the spark process operates by triggering a thyristor. A high-voltage pulse
with amplitude of maximum 15 kV (in the present case) is generated by discharging a
primary capacitor of 1 ~F, initially charged at about 300 V, through a HV transformer.
The energy stored on the primary capacitor, which indicates a theoretical upper
energy limit of the spark energy. Losses in the HV transformer and the charging
resistor cause the actual upper energy limit to be significantly lower. The maximum
consistent delivered spark energy achieved during the ignition process was 10 mJ
only. It was observed that the voltage delivered by the HV transformer was
insufficient to charge capacitors with higher capacitance (390 pF). This larger value
capacitor was unable to deliver a spark after triggering the thyristor. The charging
time for the large capacitor was greater compared to the other capacitors. This was
also one of the reasons that no spark was observed inside the vessel when working
with the larger capacitors.
The experimental flammability limits for the gases are heavily dependent on the
ignition energy of the sparking process. Ignition energies have a characteristic 'U'
shape dependence on concentration and this was observed for all the gases under
consideration. Ignition energy increases away from the stoichiometric mixture at an
ever increasing rate and increases, asymptotically, at the limiting values of .
flammability. The experimentally observed flammability limits were narrower than the
reported data. Increasing the spark energy (by due modification of the electrical
circuit) will help broaden the range of experimental gas compositions that can be
studied as the lean and rich limits are approached. Extending the range of
experiments, particular in the vicinity of the lean flammability limit, is important since
little information is available near to the flammability limits:
7.1.1 Modifications proposed to the experimental setup
Human intervention was required during, vacuum creation, gases introduction (valves
operation) and combustion products expulsion. The manual operation of in-situ
162
preparation of methane-hydrogen mixtures, severely limits the lowest (lean) limit of
the gas to be prepared and experimentally tested inside the vessel. Procurement of
premixed gas cylinders with 25% H2 in CH., 50% H2 in CH., and 75% H2 in CH.,
(similar compositions to the one studied in this work) for further experimental runs,
will help reduce the errors in manual preparation of these gases. Moreover the ratio
of the methane and hydrogen gases will be consistent over the entire flammability
range, when gases are drawn from same premixed cylinder.
The pure gases (methane and hydrogen) or the premixed gases would still require
manual operation of feed valves, since various flammable gases will be introduced to
the evacuated vessel, followed by air injection to achieve various flammable
concentrations. Manual preparation often required expulsion of gases, when more
than the desired quantity of flammable gas was introduced to the combustion
chamber. This resulted in loss of flammable gas. Hence high precision metering
pumps can be installed on each of the feed gas lines (including air), to get the exact
predetermined composition of the flammable concentration inside the combustion
chamber.
The stored energy in the primary capacitor was the theoretical upper limit of the
spark energy delivered by the rig. If the primary capaCitor is not large enough, the
discharge capacitor never reaches the breakdown voltage during the high-voltage
pulse charge duration. The maximum achievable upper energy limit from the circuit
was only 10 mJ. A higher value primary capacitor is proposed to obtain higher energy
from the circuit. The primary capaCitor cannot be increased disproportionately. The
circuit can accommodate a certain increase in capacitance, which needs to be
determined only after installation of the increased value capacitance. To increase the
spark energy further, a higher turns ratio HV transformer is required to achieve higher
voltage on the secondary side of the transformer. The higher voltage will be fed
through a charging resistor, causing a voltage build-up on the discharge capacitor
placed downstream of the charging resistor. A spark occurs when the discharge
capacitor finally gets charged to the breakdown voltage of the electrode gap. This
helps in generating a spark with higher value capacitors, delivering higher energy.
Accordingly modifications in the EHT pulse generator unit is proposed to achieve
higher spark energy.
163
The ignition process is visually observed and the transient effects of the ignition
process confirmed through the overpressure wave recordings as generated during
the process. The observation windows provided on the experimental setup can be
utilised to measure various other parameters, if a high speed camera (camera
exceeding the flame speed) is installed outside the windows. The camera can be
interlocked with the spark triggering process to capture photographs of the burning
process. Images of the development of the flame kernel, and the subsequent flame,
including video records can be obtained. Various parameters, like flame speed,
burning velocity, overpressure wave, temperature gradients etc. can be correlated
with each other for the flammable gases under consideration (lIbas et. al. 2006 and
Fairweather et. al. 2009). Initial disturbances created by the electrical discharge,
flame initiation, flame size, flame development, flame growth can also be measured.
The birth of the ignition process can be studied through the flame photographs. More
information of the ignition process can be obtained if a high speed camera is
synchronised with the experimental setup. Accordingly a detailed insight into the
burning process and the ignition process can be obtained.
7.2 Future work
Modifications suggested to the experimental rig, will allow additional parameters to

be monitored (discussed in Section 7.1.1). This will enable a better understanding of
the ignition process to be obtained. The combustion chamber and its associated
ancillary units improve the certainty with which various gas compositions can be
prepared and increase the accuracy of the measurement of ignition energy. The
cylindrical combustion chamber is capable of withstanding the overpressure waves
generated during the combustion process for all flammable gas-air mixtures prepared
at ambient conditions. Such a vessel can be used to study other combustion process
parameters. The parameters which can be measured are numerous, on the subject
matter of ignition, flame development, flame propagation, burning process and
combustion.
Ignition probability for gas release cases are often considered from historical data or
from expert judgement to arrive at an understanding of the risk of fire and/or
explosion. Results from the present experiments can be considered a beginning
phase to detenmining the proportion of gas release incidents that can proceed to fire
164
and/or explosion. In order to further analyse real world situations to arrive at ignition
probability for gas releases, all ignition sources need to be categorised with their
ignition sources energies. As already discussed, the energy delivered during the
sparking process and the energy actually received by the gas is different. Moreover
the enclosure of the ignition sources prevents the flammable gas-air mixture actually
penetrating to the spark location. The degree of confinement or openness of the
sources can be studied for common ignition sources observed in our daily life.
Ignition sources can be placed inside the cylindrical combustion chamber and then
tested with various flammable gas-air mixture compositions for their ignition potential.
The certainty with which these ignition sources can result in the ignition of the
flammable gas can be measured experimentally. Accordingly, various common
sources can be tested and assigned ignition potential. Experiments with different gas
composition for methane and also with hydrogen will help determine the impact of
each gas and its nature on the ignition properties.
Distribution of various ignition sources within the house/confined area and the ignition
potential will help predict the ignition probability of the released gas. The uncertainty
in the ignition probability can thus be reduced (if not eliminated) to arrive at the actual
value of the ignition probability under realistic gas release scenarios in future.
165
References
Acton, M.R., Gosse, A.J. and McCollum, D. (2002). Pipeline risk assessment new
development for natural gas & hydrocarbon pipelines. Advantica, Loughborough, UK.
Advantica (2005 - 2008). Technical discussions with Advantica staff, Loughborough.
Advantica report R 8224, (2005). An overview of the PIPESAFE risk assessment

package for natural gas transmission pipelines.
Astbury, G.R. and Hawksworth, S.J., (2007). 'Spontaneous ignition of hydrogen

leaks: A review of postulated mechanisms'. International Journal of Hydrogen
Energy, 32 (13), pp. 2178 - 2185.
Beckervordersandforth, Christian, Ruhrgas AG, Germany (2004). Gas Demand and

Supply in Europe, Third International Workshop on Oil and Gas Depletion, Berlin,
Germany, 25-26 May, 2004.
Bjerketvedt, D., Bakke, J.R. and Van Wingerden, K. (1997). Gexcon - Gas explosion
handbook, Amsterdam : Elsevier, pp. 1 - 150.
Blanc, M. V., P. G. Guest, G. Von Elbe, and B. Lewis (1947). Ignition of explosive
gas mixtures by electric sparks: I. minimum ignition energies and quenching
distances of mixtures of methane, oxygen and inert gases. The Journal of Chemical
Physics, 15 (11), pp. 798 - 802.
Blything, KW. and Reeves, A.B. (1988). An initial prediction of the BLEVE frequency
of a 100 te. Butane storage vessel, SRDl HSE/ R488/1988.
Britton, L.G. (1992). 'Using material data in static hazard assessment'. Plant /
Operations progress, 11 (2), pp. 56 - 70.
166
Browning, RL. (1969). 'Estimating loss probabilities'. Chemical engineering, 76, Dec.
15, pp. 135.
Calcote, H. F. Gregory, C. A. Barnett, C. M. and Gilmer, R B (1952). 'Spark

Ignition- Effect of Molecular Structure', Industrial and Engineering Chemistry, 44 (11),
pp. 2656 - 2662.
Cashdollar K.L., Zlochowera lA, Greena G.M., Thomasa RA and Hertzbergb M.

(2000). 'Flammability of methane, propane, and hydrogen gases', Journal of Loss
Prevention in the Process Industries 13 (3-5), pp. 327 - 340.
Copper, M., Phillips, A. and Robirison, C. (2004). A study of the causes of ignition of
natural gas release from transmission pipelines Advantica report R 5516.
Cox, AW., Lees, F.P. and Ang, M.L. (1990). Classification of Hazardous Locations. A
report on the Inter - Institutional Group on the Classification of Hazardous Locations,
London: IChemE, UK.
Cronin, P. and Smith B.J., (2000). The potential for Ignition of a Natural gas I Air
mixture by the operation of an Automatic security light, Advantica, R 3974.
Crowl, D.A. & Louvar, J.F (1990). Chemical process safety fundamentals with
applications. ·New Jersey: Prentice Hall.
EGIG, (Gas pipeline incidents - 2008). Report of the European Gas pipeline Incident
data group by EGIG Doe. No. EGIG 08.R0002, (7'h EGIG report, 2005).
Esseghir, M. and Polymeropoulos, C.E. (1988). 'Spark ignition of near the lean limit
CH 4 - air mixtures at low pressures', Combustion and Flame, 73, pp. 99 - 105.
167
Fairweather, M. Ormsby, M.P. Sheppard, C.G.w. and Woolley, R. 'Turbulent burning
rates of methane and methane-hydrogen mixtures', Combustion and Flame, 156 (4),
pp. 780-790.
Hattwig Martin and Steen Henrikus (eds.) (2004). Handbook of explosion prevention
and protection Germany: Wiley VCH Verlag GmbH & Co. KG aA, Weinheim.
HSE (2006) Health and Safety Executive report on gas pipelines unit,
Annual report - 2005/06.
HSE (2002). Offshore hydrocarbon releases statistics and analysis, Report - HSR
2002002.
Hill R.W. and Pool G. (1998). An experimental study of the build-up and dispersal of
natural gas in a house. HSE CRR 169/1998.
HSE (1981). Canvey: A Second Report. Health and Safety Executive, London, 1981.
HSE (1969). Ignition of gases and vapours, review of ignition of flammable gases and
vapours by friction and impact, (The UK Safety in Mines Research Establishment)
Huang, C.C., Shy, S.S., Liua, C.C. and Van Y.Y. (2007). 'A transition on minimum
ignition energy for lean turbulent methane combustion in flamelet and distributed
regimes'. Proceedings of the Combustion Institute, 31 (1), pp. 1401-1409.
Huanga, Z., Zhanga, Y., Zenga, K., Liua, B., Wanga, Q., and Jianga, D., (2006).
'Measurements of laminar burning velocities for natural gas-hydrogen-air mixtures'.
Combustion and Flame 146, (1-2), pp. 302 - 311.
IIbas, M., Crayford, A.P. Yilmaz, I. Bowen, P.J. Syred N. (2006). 'Laminar burning
velocity of hydrogen-air and hydrogen-methane-air mixtures : An experimental
study'. International Journal of Hydrogen Energy, 31 (12), pp.1768-1779.
168
Jeffreys, D.T., Moussa, N.A., Caron, R.M. and AIIan, D.S. (1982). Ignition sources of
LNG vapour clouds', Gas Research Institute GRI - 80 f 0108.
Jones, D.J. (1986). An analysis of reportable Incidents for natural gas Transmission
and gathering lines - 1970 through June 1984, American gas association NG - 18,
Report no. 158.
Johnson D.M. and Wright S.J., (1989). The potential for ignition of Natural gas - Air
mixtures by the operation of a Telephone receiver, Advantica, MRS 15051.
Kim, H.J., Chung, S.H., and Sohn, C.H. (2004). 'Numerical Calculation of Minimum
Ignition Energy for hydrogen and Methane fuels'. Korean Society of Mechanical
Engineers (KSME) Intemational Joumal, 18 (5), pp. 838-846.
Kletz, TA (1977). 'Unconfined vapour cloud explosions'. AIChE eleventh Loss

Prevention Symposium, 11, pp. 50.
Komai, T. Uchida, S. and Umezu, M. (1994). 'Ignition of methane - air mixture by

frictional sparks from light alloys'. Safety science, 17 (2), pp. 91-102.
Kondo, S., Takahashi, A. and Tokuhashi, K. (2003) 'Calculation of minimum ignition

energy of premixed gases'. Journal of Hazardous Materials, 103 (1-2), pp. 11-23.
Kono, M., Kumagai,S. & Sakai, T. (1977). 'The optimum condition for ignition of
gases by composite sparks' .. International Sympo. on Combustion, 16(1), pp. 757- 766
Kono, M., Niu, K. Tsukamoto, T. and Ujiie, Y. (1989). 'Mechanism of flame kernel
formation produced by short duration sparks'. International Symposium on
Combustion, 22 (1), pp. 1643 - 1649
Langan, M.J. and Darby, S. (1998). A review of potential ignition sources during
emergency work. Advantica, R 1545.
169
Lees, F.P. (1980 and 1996) loss prevention in Process Industries., 1 & 2 edition.
London: Butterworth & Co. Ltd.
rd
Lewis, B. and Von Elbe (1987). Combustion, Flames and Explosions of gases, 3
edition. London: Academic press.
Maly, R. and Vogel, M. (1979). 'Initiation and propagation of flame fronts in lean CH4-
air mixtures by the three modes of the ignition spark'. Symposium (Intemational) on
Combustion, 17 (1), pp. 821 - 831.
Mannan, Sam (2005). Lee's Loss prevention in Process Industries, 3"' edition. US :
Elsevier Butterworth- Heinemann.
Moorhouse, J., Williams, A. and Maddison, T.E. (1974). 'An Investigation of the
Minimum Ignition energies of some C1 to C7 hydrocarbons'. Combustion and flames,
23 (2), pp. 203-213.
NACA report, (1957). 'Ignition and flammability of hydrocarbon fuels', National

Advisory Committee for Aeronautics (NACA) report.
Florisson 0., Alliat I., Lowesmith B.J. and Hankinson G., (2006) The value of the
existing Natural gas system for Hydrogen, the sustainable future energy carrier', 23"'
World gas Conference, Amsterdam.
Ono, R., Nifuku, M., Fujiwara, S., Horiguchi, S. and Oda T. (2007). 'Minimum ignition
energy of hydrogen-air mixture: Effects of humidity and spark duration'. Journal of
Electrostatics, 65, (2), pp. 87 - 93
PIPESAFE Manual, Advantica, (2005). Training on PIPESAFE - hazard and risk

assessment package, Advantica, 2005.
Pratt T.H. (1997). Electrostatic Ignitions of Fires and Explosions, US : Wiley - AIChE
170
Randeberg E. (2006). Electric spark ignition of sensitive dust clouds in the sub 1 mJ
range, PhD thesis, University of Bergen, Norway.
Robinson, C. and Ronnie, S. (2005). A model for probability of Ignition of gas

releases from Transmission pipelines, Advantica R 8249.
Sayers J.F. Tewari G.P. and Wilson J.R. (1970). Gas Council Research
Communication GC 171.
Shepherd, J.E., Christopher, K. J. and Lee J.J. (2000). Spark Ignition Energy
Measurements in Jet A, California Institute of Technology, Pasadena, CA 91125, US
(prepared for National Transportation Safety Board, US).
Sofronis, P., Robertson I., and Johnson D. (2005). Hydrogen Embitllement of pipeline
steels: Causes and remediation. University of Illinois, USA.
Spencer, H., Rew, P.J. and Daycock, J.H. (1997)

Ignition probability of flammable gases, HSE, CRR146 I 1997.
Spencer, H., Daycock, J.H. (1997) and Rew, P.J. (1998)

A model for the ignition probability of flammable gases, HSE, CRR 203/1998.
Daycock, J.H. and Rew, P.J. (2004)

Development of a method for the determination of on-site ignition probabilities, HSE,
RR 226 I 2004.
Stern Jonathan (2002). 'Security of I European Natural gas supplies', The royal
institute of International affairs, UK.
Swaffield F., (1999). An experimental study of the Ignition of Natural gas - Air
mixtures by electrical equipment, Advantica, R 3229.
Swain, M.R., Filoso, PA and Swain, M.N. (2005). 'Ignition of lean hydrogen-air
mixtures'. International Journal of Hydrogen Energy 30 (13-14), pp. 1447 -1455.
171
Toriyama, Y and Saito, S. (1942). 'On Ignition of Inflammable gas'. Journal of the
Institute of Electrical Engineers, Japan 62, pp. 427.
Townsend, N. A. and Fearnehough G.D. (1986). Control of risk from UK gas

transmission pipelines, British Gas plc. Research and Development Division, UK.
Transportation research board, US (2004). Special report 281, 'Transmission

pipelines and Land use: A risk - informed approach',
(URL : http://onlinepubs.trb.orglonlinepubs/srlsr281.pdf).
UKOPA (2005). United Kingdom Onshore Pipeline Operators Association (UKOPA)

pipeline fault database by Advantica, Report no. R 8099.
Willcocks John and Yong Bai (2000). 'Risk based inspection and integrity
management of pipeline system'. The Proceedings of the Tenth (2000) International
Offshore and Polar Engineering conference
LINKS
Department of Trade and Industry, UK, URL : www.dti.gov.uk/ [Dec. 2007]
Environment agency, UK, URL : www.environment-agency.gov.uk, [Dec. 2007]
German hydrogen association, URL : www.dwv.info.de [May. 2007]
National Gird, UK, URL : www.nationalgrid.co.uk, [Jan 2008]
Naturalhy, URL : www.naturalhY.net. [Dec. 2007]
Office of pipeline safety, USA, URL : http://ops.dot.gov/stats/stats.htm, [Jan. 2008]
Office of the Fire Marshal, Ontario - www.ofm.gov.on.ca [Oct. 2005]
Schatz research centre, URL : www.humboldt.edu/-serc/h2safety.html. [Aug. 2007]
Transco manual - piping gas for you, URL : www.transco.uk.com [Oct. 2005]
STANDARDS
EN 13673 - 1 (2003). Determination of the maximum explosion pressure and the
maximum rate of pressure rise of gases and vapours-part 1: max. explosion
pressure.
BS EN 13821 (2002). Potentially explosive atmospheres - Explosion prevention and
protection - Determination of minimum ignition energy of dust/air mixtures.
ASTM E 582 (1988). Standard Test Method for Minimum Ignition Energy and
Quenching Distance in Gaseous Mixtures.
172
Appendix A
Research work presentation and meetings attended
S.No. Name of event Venue Date
I Conferences I Technical meetings
NATURALHY Project - Safety Work Loughborc;>ugh 26

1
Package meeting University, UK March 2008
Shell Global
NATURALHY Project meeting- 16
2 Solutions,
Technical discussions on ignition energy October, 2007
Cheshire, UK
Miramar
11- 13
The 2nd International Conference on Palace,
3 September,
Hydrogen Safety (HYSAFE) San Sebastian,
2007
SPAIN
Holiday Inn,
NATURALHY project - fourth meeting of 28-29
4 Stratford on
the full consortium of the project June, 2007
Avon, UK
National Grid
NATURALHY Project - Safety Work 27
5 Office,
Package meeting June, 2007
Warwick, UK
Institution of
Defence-in-depth: Prevention of
Structural 28
6 Escalation of Fires and Explosions,
Engineers, February, 2007
FABIG technical meeting.
London·
27-30
Indian Chemical Engineering Congress GNFC Ltd.,
7 December,
Conference (CHEMCON) 2006 Gujarat, INDIA
2006
Safety Implications of a Hydrogen Institution of
Economy: Risk Assessment and Design Structural 15
8
against Fires and Explosions, Engineers, February, 2006
FABIG technical meetings London
11 Large scale experiments

g Large scale experiments for build up of 05
Spadeadam,
gas in confined Ivented enclosures and November
Cumbria, UK
gas explosions at test site 2005
III Training
10- 11
Introduction to PIPESAFE and ORDER - Advantica,
10 November
risk assessment packages Loughborough
2005
Hazard Identification, Consequence
Loughborough 18 - 22,
11 analYSis and Risk assessment (Module -
University, UK July 2005
CGP 073)
173
Appendix B
Properties of Hydrogen and Methane gas
Gas property Methane Hydrogen
Gas Density at NTP (kg/m 3 ) 0.72 kg/m 3 0.09 kg/m 3
Energy density (Higher heating value, 50 MJ I kg 120 MJ I kg

Lower heating value as MJ per kg) (55.53, 50.02) (141.86,119.93)
Flammability limits (% vol) 5 % - 15 % 4 % -75 %
Minimum Ignition Energy (mJ) 0.29 0.02 mJ
Diffusivity (cm 2/sec) 0.16 0.61
Autoignition temperature (K) 813 853
Water solubility in 100 mL 3.5 mg @ 170 C 0.16 mg @20°C
Melting point (K) 91 14
Boiling point (K) 109 20
Flash point (K) 85 20
Flame temperature (K) 2148 2318
Max. Laminar burning velocity in air (m/s) 0.43 3.46
Detonation limits (% Vol.) 6.3-13.5 11/18-59
Flame quenching distance (gap) 2.16 mm 0.6 mrn
Flames visibility visible invisible
Molecular weight 16 2
The above data is gleaned from various sources (books, standards, research papers
and internet search) and hence references are not provided.
174
Appendix C
Calculations for the cylindrical combustion chamber design
The cylindrical combustion chamber (Figure C.1) was fabricated from a 6 inch
(schedule 40) stainless steel (SS) seamless pipeline section with welded flanges on
either side of the pipe ends. The ends were closed using stainless steel flanges
specially fabricated to house the polycarbonate plate and then bolted to the cylinder
ends through covering flanges.
The actual cylindrical combustion chamber measurements are:

Internal diameter (ID) = 155 mm
Outer diameter (OD) = 169 mm
Wall thickness = 7 mm
Chamber Length = 232 mm
Flange (welded to pipe) outside diameter = 279 mm
Flanges (welded to pipe) thickness =24mm
Flanges (for closures) ID = 155 mm
Flanges (for closures) OD = 279 mm
Flanges (for closures) thickness =24mm
Polycarbonate housing diameter = 197 mm
o ring gasket nominal bore = 175 mm
o ring gasket diameter =5mm
Total thickness of polycarbonate plate (for window) = 25 mm

Polycarbonate plate inside the shell = 10 mm
Polycarbonate plate inside flanges = 15 mm
Bolt pitch circle diameter = 242 mm
Diameter of hole for bolts = 10 mm
Diameter of each bolts = 9.5 mm
Nos. of bolts = 16
T
Cl> 279
24
L..k:::::J---- 232 - - -
(all dimensions are in mm)
Figure C.1 : Details of the windows and the flange assembly for the cylindrical
combustion chamber
175
C.1 Sample calculation for Overpressure expected during the explosion
The maximum overpressure for confined gas explosion within experimental

cylindrical combustion chamber was based on the assumption, that during the
combustion of a stoichiometric mixture of fuel gas with air, the adiabatic flame
temperature is reached for the burning gases. In practice the actual flame
temperature achieved will be less than the adiabatic flame temperature, but the
cylindrical combustion chamber was designed for the worst case scenario (EN 13673
: 2003). Accordingly two different cases were considered to determine in maximum
adiabatic flame temperature during each process.
Case I : Stoichiometric mixtures of methane - air

Case '11 : Stoichiometric mixtures of hydrogen- air
Technical data used
% fuel gas Adiabatic Ambient Ambient (initial)

M.
Fuel at stoic. flame temp (initial) temp. pressure
wt. ratio Tf (K) Ti (K) Pambient (MPa)
Methane 16 9.5 2148 288 0.1
Hydrogen 2 29.5 2318 288 0.1
During the combustion process define
Nu = number of moles or all reactants

Nb = number of moles or all products formed
Case I: Combustion reaction for methane
CH. + 2 O2 + 7.52 N2 ----> CO2 + 2 H2 0 + 7.52 N2
NR = 1+ 2 + 7.52 = 10.52 NP =1+ 2 + 7.52 =10.52

NP 1 NR = 10.52/10.52 = 1
Expansion factor 'EF' = (Tf 1Ti) • (NPI NR)
= (2148/288)· (1)
= 7.45
Maximum Pressure generated during explosion
Pmax = Pambient * EF
176
Pmax = 0.1 * 7.45 MPa.
Case 11 : Combustion reaction for hydrogen
H2 + 0.5 O2 + 1.88 N2 ----> H20 + 1.88 N2

NR = 1+0.5 + 1.88 = 3.38 NP = 1+1.88 = 2.88
NP 1 NR = 2.88 1 3.38 = 0.85
Expansion factor 'EF' = (Tf 1Ti) * (NPI NR)
= (2318/288 ) * (0.85)
= 6.84
Maximum Pressure generated during explosion
Pmax= Pambient * EF
Pmax = 0.1 * 6.84 MPa.
Flammable mixture Max. Pressure (MPa)

Case I : Stoichiometric mixtures of methane - air 0.745
Case 11 : Stoichiometric mixtures of hydrogen- air 0.684
For cylindrical combustion chamber design purposes the maximum peak

overpressure generated during the combustion process in Case I or Case 11 was
assumed to be P(combust;onj = 0.8 MPa.
C.2 Sample calculation for the shell thickness
Cylindrical combustion chamber parameters
Shell thickness (ths) = 7 mm (0.007 m)

Shell diameter (d) =155mm (0.155 m)
2
Allowable working stress for SS, considered ( f ) = 150 MPa (1500 kg/cm )
Overpressure expected P(combus,;onj = 0.8 MPa
Based on the hoop stress (Ph) equation, the cylindrical combustion chamber can
withstand a maximum pressure of
= 2*f* ths/d
= 2 * 150 * 0.007 1 0.155
= 13.5 MPa
177
Safety factor for the vessel shell = Ph I P(combustion)
= 17
The safety factor of 17 was based on the fact that the pressure inside the cylindrical
combustion chamber was considered static, whereas the pressure exerted during the
explosion and subsequent overpressure wave impact was dynamic. Even if the
dynamic pressure (overpressure) was twice the static pressure, the cylindrical
combustion chamber shell thickness would be strong enough to carry out the
experiments (explosions inside the cylindrical combustion chamber)
C.3 Sample calculation for polycarbonate plate window thickness
Window parameters
Window diameter (d w ) = 155 mm (0.155 m)

Window thickness ( thw) = 15 mm (0.015 m)
Polycarbonate Tensile strength (f) =
50 MPa
Overpressure (p) = 0.8 MPa
thw = d w* .,J(3116) * (pi f) (ref. Brownell L.E. and Young E.H.

Process equipment design, 1977)
Or
Limiting pressure' p , for a window thickness' thw' will be
p=(16/3)* f*(thwldj
= (16/3) * 50*(151155)2
= 2.50 MPa
Safety factor for polycarbonate window
= 3.13
The safety factor of 3.13 for the windows was based on static pressure inside the
cylindrical combustion chamber, similar to the analysis carried out for the shell
thickness. The same logic of static and dynamic pressure was applied here, for the
calculation of window thickness. It was inferred that in the case of dynamic pressure
also, the windows would be capable of withstanding the overpressure wave
generated through the explosion.
178
C.4 Sample calculation during vacuum formation
Cylindrical combustion chamber parameters

Vessel Length (I) = 232 mm
Outside diameter (do) = 169 mm .
Vessel wall thickness ( th s ) = 7 mm
Vessel Length I Outside diameter (11 do )= 232/169 = 1.37
Outside diameter I thickness = 169/7 = 24

Factor B = P (dolt) = 1400 (Ref. Brownell L.E. and Young E.H. Process
equipment design, 1977, pp. 147, )
where
p (allowable pressure)= 1400 I (24)
= 58 psi
Since the atmosphere can exert a pressure of only 15 psi. (outside pressure = 14.7
psi) in case of a full vacuum (pressure inside vessel = zero psi), where as the
cylindrical combustion chamber can withstand a negative pressure of 58 psi.
Comment : The cylindrical combustion chamber is able to withstand stresses

generated due to explosive pressure and during the complete vacuum
formation inside the cylindrical combustion chamber.
179
C.S Flange specification
As per ASTM A 181 (BS 4504 : 1989) flange specifications for the 6 inch Stainless
steel pipe are as follows:
Outside diameter = 13.5 in. (350 mm)

Min. thickness =1.125in. (28.6 mm)
Bolt pitch circle diameter = 11.75 in. (298.45 mm)
Diameter of hole for bolts = 0.875 in. (22.2 mm)
Diameter of bolts = 0.75 in. (19 mm)
Nos. of bolts =8
Actual flange specifications used
Outside diameter = 279 mm

Flange thickness = 24mm
Bolt pitch circle diameter = 242 mm
Diameter of hole for bolts = 10 mm
Diameter of bolts = 9.5 mm
Nos. of bolts = 16
The standard (BS I ASTM) flange specifications are based on industrial use of the 6
inch pipe, utilised for very high pressures. The maximum pressure achieved in our
experimental cylindrical combustion chamber is disproportionately less than the
industrial pressures. Hence present flange specifications utilised in fabrication of the
experimental cylindrical combustion chamber are adequate as confirmed through the
calculations presented above.
180
Appendix 0
Risk Assessment Record for the Assessment Number
Department
Assessor Hemant Mathurkar Date 08.05.2007
Chemical Engineering
nition measurement
Risk Identified Assessors Risk Assessors Actions to

Persons at Risk
(see Hazard prompt list over page) Controls in Place Rating Further Reduce Risk
(Groups or nos.)
Methane: is a colourless persons CVi,<;Ud'" area
odourless gas. It is flammable and may form appropriate protective equipments,
working in lab A detection and extraction
flammable and explosive mixtures with air. Eliminate ignition sources. The rig is
Asphyxia may result if the oxygen Area (S - 038) system at ambient pressure and its volume
concentration is reduced to below 18% by is 4 litres, so only minor leaks from
displacement. vessel are envisaged in case of
gas I persons gases area use I

odourless and flammable. It forms flammable appropriate protective equipments,
working in lab detection and extraction
and explosive mixtures with air over a wide Eliminate ignition sources. The rig is
range of concentrations. Hydrogen burns with Area (S - 038) system. at ambient pressure and its volume
an almost inVisible blue flame Gas is non- is 4 litres, so only minor leaks from
toxic, and the primary health hazard is vessel are envisaged in case of
Supervisors Comment
I medium
Assessors Signature Supervisors v,~, "nu, ~
Reassessment Date Date: 10.05.2007
181
Risk Assessment Record Assessment Number
Location 2 nd. Pilot Laboratory
Department
Room NoJArea: S- 038 Assessor Hemant Mathurkar Date 08.05.2007
Chemical Engineering
I c, ""\,Ii m""C:1 III rig,
Risk Identified Assessors Risk Assessors Actions to

Persons at Risk
(see Hazard prompt list over page) Controls in Place Rating Further Reduce Risk
(Groups or nos.)
IH.M.L. see 1)
Overpressure of experimental rig, before or All persons The line pressure and i ~ "o~"oto area, use all a~~' v""a,o
following Ignition. vessel pressure is protective equipments. Adopting the
working in lab
continuously monitored and safe procedure of disconnecting the
Area (S - 038) controlled. Excessive hose from the ignition vessel and
pressure can be vented then venting the excess pressure to
through three way valve. atmosphere.
Rig is over-designed with
safety factor =17.
Windows designed with
=
safety factor 3.
Gas leak in line: Methane J Hydrogen : All persons Fumes J flammable gases ~~ediulT' Close all vy... ,vo' valves and
Gases are colourless, odourless and detection and extraction regulators. Minor leaks will be
working in lab
flammable. It forms flammable and explosive system. diluted to below the LFL limits. For
mixtures with air over a wide range of area major leaks Contact emergency
concentrations. Gases are non-toxic, and Check for leaks using leak telephone number (888).
health hazard are asphyxiation by detection fluid
~,. : of
,~ '" : Supervisors Comment
• . risk rating is medium
~I "h.M." ""oh
Assessors Signature Supervisors Signature
-0-
.T .0.
Reassessment Date Date: 10.05.2007
lllV~[SlLy
182
Machinery & Work Equipment - Category 1 Hazardous chemicals (CHIP P.1 0) Hazards associated with Work Organisation
Mechanical Hazards Explosive (Contractors/Service) - Category 5
Electric Shock Oxidising
VehiclesfTransport Extremely flammable Hazards associated with Work Environment -
Hand Tools High flammable Category 6
Flammable Temperature
Hazards associated with Place of Work - Very toxic Heating
Category 2 Toxic Ventilation
Slips, trips and falls on a level Harmful Lighting
Falls from a height Corrosive
Falling objects/materials Irritant Other types of hazard - Category 7
Striking objects Sensitising Violence
Contact with hot/cold materials/surfaces Carcinogenic Stress
Storage and stacking Entry into confined space/lack of oxygen Drugs
Space and confined work area Substance abuse
Hazards associated with activity, methods of
Hazards associated with materials, substances work - Category 4
& physical agents - Category 3 Manual handling
Dust, fume and gases Upper Limb Disorders/Repetitive Strain
Biological hazards/infection Injury
NoiseNibration Visual fatigue
Compressed gases Posture
1Kl~§~ lKl~n~Iil~§ ~1NI~~~v il'lI@~~!YIlMlv ~@W~

CONSEQUENCE: LIKELIHOOD: RISK RATING =
3 Major (e.g. death or major injury as per 3 High (where certain or near certain harm will CONSEQUENCE x LIKELIHOOD
RIDDOR or irreversible health damage) occur
2 Serious (e.g. injuries causing absence of 2 Medium (where harm will frequently occur) High 6 - 9
more 1 Low (where harm will seldom occur) Medium 2-4
than three days or significant health effects- Low 1
reversible)
1 Slight (other injuries requiring first aid and
minor ill health effects)
183
Appendix D (continued)
Department Equipment Description:- Ignition energy measurement rig Assessor Date
Chemical Engineering Area: - Room No. S- 038 (2'" Pilot Laboratory) Hemant Mathurkar 08.05.2007
Risk Identified Risk Rating Action-Control Required to

OPERATION' Persons at Risk
(see Hazard prompt list Controls in Place (H. M. L see over
ACTIVITY (Groups of numbers) further reduce risk
over page) page)
Direct contact to any of the Electrical hazard Single person Ignition can only be MEDIUM Operations strictly as per the
high voltage device (direct contact) carried out when key operating procedure. Avoid
is inserted wet conditions on the floor
and near the devices.
Changing the capacitor' Electrical hazard Single person Interlock control MEDIUM Earth capacitors before
resistors in the ECU (direct contact) system changing' removing and
resistor' capacitor
Mechanical impact by a Electrical hazard Single person ECU is closed LOW No shelf above the ECU.
conductor material/ (only if it is simultaneous Nothing is kept on the ECU.
dropped object on ECU with the triggering pulse,
OR manual triggering
Managers Comments/Approval:
Assessors Comment: No further risks identified
Date
Reassessment Comments:
09.05.2007
Date Managers Signature Date
Assessors Signature 10.05.2007
10.05.2007
184
Machinery Hazard Prom'-'-p...;;,t...;..L_is....;t_ _ _ _ _ _-:Risk Assessment Method & Estimation Ratings
r:P~R=-=O=-'B,;.:A::..:B=-=I,:.;L~ITY~O"""'F=-----,
MECHANICAL HAZARDS HAZARDOUS SUBSTANCES
OCCURRENCE of that harm
Crushing Toxic fluids RISK is a severity and
Shearing Toxiclgas/misUfumes/dust function frequency and duration of
Cuttings/Severing Flammable liquids related to of of the exposure
Entanglement Flamm.gas/misUfumes/dust the possible
Drawing-infTrapping Explosive sUbstances considere harm for the probability of occurrence
Impact Biological substances d hazard considered of hazardous event
Stabbing/Puncture Other hazardous substances hazard
Friction/Abrasion possibility to avoid or
WORK ACTIVITY HAZARDS
High Pressure Fluid Injection limit the hanm
SlipsfTrips/Falls Highly repetitive actions
Falling/Moving object Stressful posture RISK RATINGS (HIGH, MEDIUM, LOW)
Other mechanical hazards Lifting/Handling
Mental overload/Stress SEVERITY •
ELECTRICAL HAZARDS 3 • MAJOR (eg. death or major injury as per
Visual fatigue
Direct contact Poor workplace design RIDDOR or irreversible health damage)
Indirect contact Other workplace hazards 2 SERIOUS (eg. injuries causing absence of more
Electrostatic phenomena than three days or significant health effects·
WORK ENVIRONMENT reversible).
Short circuiUOverload
Source of ignition Localised hot surfaces 1 MINOR (eg. first aid treatments and other lost time
Other electrical hazards Localised cold surfaces injuries)
Significant noise PROBABILITY
RADIATION HAZARDS 3· HIGH (where certain or near curtain harm will
Significant vibration
occur).
Lasers Poor lighting
2· MEDIUM (where harm will frequently occur).
Electro-magnetic effects HoUcold ambient temperature
Ionising/Non-ion. radiation Other work environment 1· LOW (where harm will seldom occur)
RISK RATING = SEVERITY X PROBABILITY
Other radiation hazards Hazards HIGH 6-9 MEDIUM 2·4 LOW 1
185
REFERENCE NUMBER: 869 Form:COSHH Page no.
CONTROL OF SUBSTANCES HAZARDOUS TO HEALTH
RISK ASSESSMENT FOR THE EXPERIMENTAL RIG
Before filling in this form, please read the Notes, as indicated.
This form MUST BE COMPLETED prior to the commencement of any work involving
risks to health from a hazardous substance, so that a suitable and sufficient assessment of
health risks is made.
Notel: Persons completing this form should make themselves aware of the Health and Safety
Commission Approved Codes of Practice "Control of Substances Hazardous to Health" and
"Control of Carcinogenic Substances", the HSE booklet "COSHH Assessments".
PART A RISK EVALUATION
At Department Chemical Engineering
A2 Title of Work Activity Research activity (Ignition energy measurement for gas mixture)
Note:2 Choose a title or give a serial number so as to facilitate departmental filing and/or risk
assessments.
A3 Locations(s) of Work 12 nd. Pilot laboratory (S - 038)

A4 Hazardous Substance(s) Classification
(Note 3) Tick I or more boxes
Very Toxic
Toxic .
Harmful Highly flammable

Corrosive
Irritant
SubstanceCsl with MEL or OES
Dust
Carcinogen(or sus1"'ct carcinogen)
~icro-organism
Specify particularly dangerous or hazardous substance(s) (Note5)
Specify particularly dangerous or hazardous substance(s) (Note5)
1 Methane (gas), Hydrogen (gas)
Total number of hazardous substances involved in the work activity 2
Note 3: The COSHH Regulations do NOT apply where either the Control of Asbestos at Work or
theControl of Lead at Work Regulations apply or where the risk to health is solely from
radiation noise, pressure, explosive or flammable properties, heat or cold, nor to medicines
administered to patients.
A substance should be regarded as hazardous to health if it is hazardous in the form in which it

occurs in the work activity, including by-products and waste residues.
186
(a) Any substance which is listed as "very toxic","toxic","corrosive","harmful", or"irritant"

in Part I A of the Approved List for the Classification, Packaging and Labelling of
Dangerous Substance Regulations, 1984, (2nd edition onwards) is a substance hazardous
to health.
(b) Any substance which has a MEL(Maximum Exposure Limit) or OES(Occupational
Exposure Standard) given in the HSE Guidance Note EHAO (current year date) is a
substance hazardous to health.
(c) Micro-organisms which create a hazard to the health of any person, where the hazard
arises out of or in connection with a work activity. Hazard classification of pathogens is
given in the booklet "Categorisation of Pathogens etc", Advisory Committee on
Dangerous Pathogens. See also the relevant' Advisory Committee on Genetic
ManipulationlHealth and Safety Executive Notes.
(d) A dust ofany kind is a substance hazardous to health when present in a
"substantial"concentration. See the Approved Code of Practice, para 2(1) and HSE
Guidance Note EHAO.
(e) Any other substance is hazardous if it creates a risk to health comparable to any of the
above .
Note 4: Refer to the HSC Approved Code of Practice,"Control of Carcinogenic Substances".

Note 5: A Part B2 "Scheme of Work" must be completed for this type of work activity.
AS Grounds for Concluding Exposure is not a Risk to Health
Quantities or rate of use of substance(s) are too small to constitute any risk to health under
foreseeable circumstances or use, even if control measures broke down. (Yes / No)
Note 6: If there are reasonable grounds for reaching the conclusion that risks are insignificant, "finish
this assessment now by signing page 6.
A6 Route by which the Substances are Hazardous to Health.

(Tick I or more boxes)
Inhalation N/A
Ingestion N/A
Skin Absorption N/A
Direct Contact,Skin or eyes N/A
Injection (via sharps) N/A
A7 What could be the Effect of Exposure to the above Hazardous Substances? (Tick boxes)
Single Acute Exposure : Serious .f

Not Serious
Not Known
Repeated low Exposures : Serious

Not Serious
Not Known .f
Adverse Effect Could Be : longTenn
Short Tenn
Not Known .f
Effects could be hannful to the Human Reproductive System: Yes
No .f
Not Known
The Micro-organism could infect an individual Yes

No .f
Not Known
An infected per-mn could infect others Yes

No .f
Not Known
187
AS Engineering Control Measures:

(Tick I or more boxes)
Note 7: The fume cupboard must be used in accordance with the University's local guidance or code
of practice
The work will require some other local exhaust ventilation (Yes I No) Yes
If Yes lease s ecif below:
The work will be carried out in a glove box (Yes I No) ~

or other sealed system (Yes I No ~
If Other please specify below:
I Contaminant within --------------
The work will be carried out in a laboratory at the required biological containment N/A
level (Yes /No)
and in a biological safety cabinet, c1ass(insert number where relevant) N/A
ACDP Category N/A
Genetic Manipulation Category N/A
A9 Personal Protective Equipment Requirements

The following personal protective equipment may be necessary for a part or all of the work.
Eye Protection (Yes I No) Yes

Face Protection (Yes I No) No
Hand Protection (Yes I No) No
Foot Protection (Yes I No) No
Respiratory Protection (Yes I No) No
Other (e.g. protective clothing) (Yes I No) Yes
Labcoat
188
PARTB DESCRIPTION OF THE WORKING PRACTICE
Note 8: Part B2 of this form must always be completed for work by postgraduate research students
and for 3rd and 4th year undergraduates carrying out similar research work.
Bl Instructions for the Work Activity

(Tick I box only)
D
The work activity consists of well documented routine procedures carried out
frequently in a controlled environment and requiring only simple and easily
understandable verbal instructions
Note 9: Where an assessment of risk is simple and obvious and where the work activity is straight-
forward and clear verbal instructions can be given easily, a written scheme of work (Part B2)
is unnecessary. Complete the other sections of Part B.
The work activity consists of procedures requiring a specific scheme of work. ",.JL_--I
LI
Note 5: A Part B2 "Scheme of Work" must be completed for this type of work activity.
B2 Scheme of Work (Continue on a separate sheet, if necessary)
Note 10: The scheme of work is a statement of how the work activity is going to be carried out safely. It
should specify the ways in which the hazardous substances are to be used or handled, and
should give sufficient details to identify the precautions necessary to control the risks that
arise from working with the hazardous substances.
Please refer to operating procedure
189
B3 Training for the Work Activity
Specific training will be required :(Yes I No) IYes

Note 11: Any special training required to ensure that persons involved in the work activity can operate
safely should be described here. This is particularly important so that persons can understand
and comply effectively with the scheme of work (B2), where this had been formulated.
Training in use and handling of pressurised cylinders.

Training in handling high voltage equipments and interlock system
B4 Supervision
Note 12: The level of supervision must always be appropriate to the competence of the individuals
involved in the work activity.
~
The supervisor will approve straightforward routine work in progress: (Yes I No) es
The supervisor will specifically approve the scheme of work, B2 (Yes I No) Yes
The supervisor will provide supervision personally to control the work (Yes I No) No
B5 Monitoring
Note 13: For the majority of work, atmospheric monitoring should not be necessary for protecting
health, providing sufficient thought has gone into ensuring the adequacy of control measures
in relation to risks, and the control measures are properly used and maintained. For further
information on monitoring and health surveillance see the Approved code of Practice,
paragraphs 66 to 92, inclusive.
Monitoring for airborne contaminant will be required: (Yes I No) ~

Biological monitoring of workers will be required: (Yes I No) ~
B6 Contingency Planning
Note 14: Contingency planning is required to limit the extent of the risk arising from an uncontrolled
release of a hazardous substance and for regaining control as quickly as possible.
Written emergency instructions will be provided for workers and others who
might be affected, on site: (Yes I No)
Provision of the following may be required in an emergency:
Spill Neutralisation Chemicals: (Yes I No)

If Yes please specify below:
Eye Irritation Point N/A

Body Showe N/A
Other First Aid Provisions N/A
Breathing Apparatus (with trained operator) N/A
External Emergency Services N/A
Poison antidote N/A
If Yes please specify below:
I
190
B7 Disposal of Waste Residues - (Not Applicable)
In-house to District Council Waste Collection, after rendering safe (Yes I No)
In-house to drain, after rendering safe (Yes I No)
In-house to incinerator, after rendering safe if appropriate (Yes I No)
To specialist licensed Waste Disposal contractor (Yes I No) (Please specify): .
Other (e.g. inter-departmental rjes I No) (Please specify below):
I
B8 Implications for other Persons
Academic Staff (Yes I No) Yes

Postgraduate Staff (Yes I No) Yes
Postgraduate Students (Yes I No) Yes
Undergraduate Students (Yes I No) Yes
Technical Staff(Yes I No) Yes
Cleaning Staff (Yes I No) No
Contractor (Yes I No) No
Visitors (Yes I No) Yes
Others (Yes I No) N/A
B9 Accreditation
(i) Signature of Assessor
Signature of Supervisor
- - - - - - - Name -H.-Mathurkar
- - - - Date 08.05.07
_________________ Name -=:G:.:..-,H"a",nki=·n::::s",o::::n__ Date 08.05.07
(ii) Signature of All Persons receiving a copy of this Risk Assessment.
Note 15: A copy of this assessment must be given to each postgraduate research student andlor to each
4th year undergraduate doing like work, and he/she must sign a receipt, at B9(11)
I/We have received a copy of Parts A and B of this Risk Assessment.
Signature Name Date
TonyEyre 08.05.2007
(Continue on a separate sheet, if necessary)
,
(iii) Date of Next Assessment
______________ 2 ___________ 3
Note 16: This assessment should be reviewed immediately if there is any reason to suppose that the
original assessment is no longer valid due to significant changes in the work activity, arising
for example, from the introduction of new hazardous substances, new personnel, changes in
procedures or reported ill-health. Otherwise, the assessment should be reviewed annually.
A·COPY OF THIS ASSESSMENT MUST BE RETAINED BY THE HEAD OF

DEPARTMENT, OR HlSIHER REPRESENTATIVE, FOR AS LONG AS IT IS RELEVANT.
191
Appendix E
Operating procedure
This section details the sequence of actions I activities to be performed (Operating
procedure) during the experimental test runs. The abbreviations used and the valves
referred during the description of the operating procedure are presented in the
Figure 3-1.
Start power supplv
Switch ON the following devices
• The computer (PC) and start the LabVIEW interface (hnmfinaI.VI)

• The Extra High Tension (EHT) pulse generator unit - i.e. Spark generator
• The Oscilloscope TDS 3034 B
• The National Instruments housing box
• The Vacuum pump
Line pressurization
1. Close manifold valves (V1, V2 and V3) and vessel valves (V5 and V6)
2. Open the gas cylinder regulator valves and set flammable gases delivery
pressure of around of around 50 kPa (max 10 psig) and the dry air delivery
pressure of around 100 kPa (max 15 psig)
3. Ensure coupling CV4 is connected to the vessel
Vessel filling through the method of partial pressure

4. Start the vacuum pump, open valve V5 and turn valve V6 to vacuum pump
5. Evacuate the vessel completely till steady state pressure transducer (P1)
reading is obtained
6. Close valves V5 & V6, stop the vacuum pump and record the pressure
transducer reading P1, with a click on "AM to File"
7. Slowly open valve V3 to introduce CH.,(and I or valve V2 for H 2 ) and record
the steady state pressure transducer P1 reading again after closing V3, with a
click on "AM to File"
8. Introduce dry air through Valve V1, so that the steady state vessel pressure
reaches just above atmospheric after closing V1. Record the pressure
transducer (P1) reading with a click on "AM to File"
9. Detach the coupling CV4, to isolate the vessel from the gas cylinders.
Observe the pressure transducer - P2 reading
10. Place the gas filler pipe to stored position and lock it
Capacitive Discharge system

11. Insert key and turn it to "SAFE", check the interlock display on EHT pulse
generator
12. Ensure the gas filler pipe to stored position (locked position)
13. Open the Electronic Component Unit (ECU) and ground "Monitor Point" in
ECU with earth probe. If necessary change resistor and discharge capacitor
14. Close and latch the ECU
15. Check interlock shows - CLOSED, Switch key lock to ACTIVE and Ready for
Ignition
192
Ignition
16. Ignite gas mixture from LabVIEW control panel with a click on "Fire & Record"
button
17. LabVIEW programme records the pressure (transducer - P2) and temperature
(thermocouple T1) rise as function of time in separate files out-'press.LMV
and ouUemp.LMV respectively.
18. Transient values of Voltage and Current will be displayed by oscilloscope.
The peak value of voltage (during spark initiation phase) as displayed by
oscilloscope is recorded manually
19. If the gas mixture is ignited by spark (confirmed through pressure and
temperature rise) then ref. step no. 21, to evacuate the vessel and repeat
experiments with fresh gas air mixture
20. If the gas mixture did not ignite, then increase the spark energy using larger
capaCitor in ECU box as follows:
• Open ECU, Check interlock show "OPEN". This will operate discharge
circuits in the EHT pulse generator
• Ground "Monitor Point" in ECU with earth probe (this will remove any
reSidual charge in the discharge capacitor). Change the capacitor and
resistors as desired
• Repeat step no. 16. A maximum of four such ignitions attempted
before finally evacuating the vessel (step no. 21) and starting with
fresh gas air mixture
Exhaust
21. Open valve V5 .
22. Open valve V6 to atmosphere (to vent excess pressure - if any - to
atmosphere)
23. Repeat procedure from step 4 to 20 with different (or same) composition of
flammable gas - air mixture andlor Capacitor I Resistor combination
The test runs were repeated using various concentrations of CH4 , H, and Air.
Shutdown (end experiments for the day)

For Electrical svstem
• Open ECU, Check interlocks show "OPEN". This will operate discharge
circuits in the EHT pulse generator
• Ground "Monitor Point" in ECU with earth probe (this will remove any residual
charge in the discharge capacitor)
• Switch key lock to "SAFE"
• Switch OFF the EHT pulse generator power
• If leaving the equipment unattended, remove key
For Mechanical system

• Ensure all valves (V1, V2, V3, V5 and V6) are closed and the coupling CV4 is
attached to the vessel
• Turn off cylinder regulators and depressurise all the lines completely by
opening valves V1, V2 and V3 one at a time into the vessel.
• Depressurise vessel by opening valves V5 and V6 to atmosphere, check the
pressure transducer readings (P1 & P2) and again close valves V5 and V6
completely
193
Appendix F
Electrical circuit and Instrumentation
,'"
,
" <0/ "
.
, "
,,
•
~
~
"
, "
1- _ _ ___ ,
c;i}l~ ~:- -~ ------- :

-~U
"
!i
\1
, ,
"
I"
i:
I,
I
,I
"
"
j
.
f I'
~ ;;
~ l'
,
~i
, >,
; l!
'" I'I
\- >
- ~ I!,
"
(., -J
u __
"
I
Figure F,1 : Rectifier circuit for the spark generation
194
~
c
iil
:n
'"
CJ
6"
(') ~
"'"
C. y./rle.Tci
w"
(Q
iil
3
0'
-
~
:::T
<D
r
Ql
rr
<
m
~. ~
::§: T~.
"0
Cl
(Q
iil
3
3
<D
~l
Write"To
[j]
Appendix G
Specifications of the Pressure Transducers
250
P (kPa) = 41.011 * (volts) - 0.6465
R2 = 1
200
-.
~150
Cl)
~
:l
~ 100
~
c..
50
0
0 1 2 3 4 5
Pressure (Volts)
Figure G.1: Calibration curve for the Pressure transducer (P1)
1200
P (kPa) = 206.58 * (volts) - 0.2375
_ _ _ _ R2=1_ _ _ _
1000
m
c.. 800
~
f!
:l
600
en
en
Cl)
c.. 400
~
200
0
0 2 3 4 5
Pressure (Volts)
Figure G.2 : Calibration curve for the Pressure transducer (P2)
196
Appendix H
Table H.1 : Experimental runs with methane - air mixture
CH4conc. H2 cone. Total flammables Air cone. Energy Success of

(%) (%) conc. ('I.) ('I.) (m:;) Ignition
5.60 0 5.60 94.40 3.84 No
5.60 0 5.60 94.40 8.28 No
5.60 0 5.60 94.40 10.73 No
5.60 0 5.60 94.40 13.13 No
5.78 0 5.78 94.22 4.03 No
5.78 0 5.78 94.22 5.69 No
5.78 0 5.78 94.22 10.26 No
5.78 0 5.78 94.22 6.98 No
5.81 0 5.81 94.19 3.92 No
5.81 0 5.81 94.19 3.16 No
5.81 0 5.81 94.19 11.62 No
5.81 0 5.81 94.19 6.28 No
5.82 0 5.82 94.18 3.57 No
5.82 0 5.82 94.18 11.30 No
5.82 0 5.82 94.18 12.25 No
5.82 0 5.82 94.18 6.28 No
6.26 0 6.26 93.74 3.56 No
6.26 0 6.26 93.74 8.24 No
6.26 0 6.26 93.74 7.38 No
6.26 0 6.26 93.74 4.98 No
5.62 0 5.62 94.38 9.07 No
5.62 0 5.62 94.38 8.46 No
5.62 0 5.62 94.38 12.25 No
5.62 0 5.62 94.38 6.74 No
5.78 0 5.78 94.22 3.57 No
5.78 0 5.78 94.22 6.76 No
5.78 0 5.78 94.22 9.69 No
5.78 0 5.78 94.22 6.98 No
5.81 0 5.81 94.19 3.98 No
5.81 0 5.81 94.19 4.69 No
5.81 0 5.81 94.19 13.73 No
5.81 0 5.81 94.19 6.74 No
5.79 0 5.79 94.21 3.56 No
5.79 0 5.79 94.21 11.52 No
5.79 0 5.79 94.21 11.68 No
5.79 0 5.79 94.21 9.59 No
6.01 0 6.01 93.99 4.53 No
6.01 0 6.01 93.99 11.30 No
6.01 0 6.01 93.99 7.84 No
6.01 0 6.01 93.99 7.22 No
6.16 0 6.16 93.84 2.27 No
6.16 0 6.16 93.84 6.75 No
6.16 0 6.16 93.64 5.88 No
6.16 0 6.16 93.84 11.96 Yes
6.21 0 6.21 93.79 2.72 No

6.21 0 6.21 93.79 12.64 No
6.21 0 6.21 93.79 9.79 No
6.21 0 6.21 93.79 9.12 No
6.28 0 6.28 93.72 3.57 No

6.28 0 6.28 93.72 7.46 No
197
CH4 conc. H2 conc. Total flammables Air conc. Energy Success of
(%) (%) conc. (%) (%) (mJ) lanition
6.28 0 6.28 93.72 11.12 No
6.28 0 6.28 93.72 7.47 No
6.03 0 6.03 93.97 1.99 No

6.03 0 6.03 93.97 6.59 No
6.03 0 6.03 93.97 8.28 No
6.03 0 6.03 93.97 13.13 No
6.15 0 6.15 93.85 2.82 No

6.15 0 6.15 93.85 12.88 No
6.15 0 6.15 93.85 10.85 No
6.15 0 6.15 93.85 9.31 No
6.21 0 6.21 93.79 3.56 No

6.21 0 6.21 93.79 7.23 No
6.21 0 6.21 93.79 10.31 No
6.21 0 6.21 93.79 6.74 No
6.21 0 6.21 93.79 2.20 No

6.21 0 6.21 93.79 5.70 No
6.21 0 6.21 93.79 8.62 No
6.21 0 6.21 93.79 10.79 No
6.15 0 6.15 93.85 2.64 No

6.15 0 6.15 93.85 12.42 No
6.15 0 6.15 93.85 10.79 No
6.15 0 6.15 93.85 6.98 No
6.10 0 6.10 93.90 3.30 No

6.10 0 6.10 93.90 7.22 No
6.10 0 6.10 93.90 11.40 No
6.10 0 6.10 93.90 6.98 No
6.24 0 6.24 93.76 2.27 No

6.24 0 6.24 93.76 5.35 No
6.24 0 6.24 93.76 8.65 No
6.24 0 6.24 93.76 11.68 No
6.00 0 6.00 94.00 11.96
(average) . (average·) (average) (average) (Lowest Energ·y)
6.97 0 6.97 93.03 3.89 No

6.88 0 6.88 93.12 5.74 Yes
7.23 0 7.23 92.77 4.39 Yes
7.05 0 7.05 92.95 7.73 Yes
6.81 0 6.81 93.19 4.02 Yes
6.64 0 6.64 93.36 12.19 Yes
6.64 0 6.64 93.36 4.55 Yes
6.64 0 6.64 93.36 8.09 Yes
6.25 0 6.25 93.75 12.64 No
6.72 0 6.72 93.28 4.02 No
6.94 0 6.94 93.06 10.87 Yes
6.81 0 6.81 93.19 4.83 Yes
6.59 0 6.59 93.41 3.89 Yes
6.68 0 6.68 93.32 3.95 Yes
6.42 0 6.42 93.58 4.15 Yes
6.27 0 6.27 93.73 6.00 Ves
6.40 0 6.40 93.60 4.78 No
6.22 0 6.22 93.78 4.83 No
6.27 0 6.27 93.73 4.28 Yes
6.99 0 6.99 93.01 4.28 No
6.85 0 6.85 93.15 4.69 Yes
6.52 0 6.52 93.48 6.36 No
6.66 0 6.66 93.34 4.02 Yes
6.51 0 6.51 93.49 5.28 Yes
6.42 0 6.42 93.58 5.43 No
6.44 0 6.44 93.56 6.87 No
6.45 0 6.45 93.55 9.83 Yes
6.65 0 6.65 93.35 13.34 Yes
6.65 0 6.65 93.35 12.42 Yes
6.51 0 6.51 93.49 4.83 Yes
198
CH4conc. H2 conc. Total flammables Air conc. Ener9]) Success of
(%1 (%t conc.(%) ('10) (mJ Ignition
6.65 0 6.65 93.35 4.99 Yes
6.27 0 6.27 93.73 12.42 No
6.22 0 6.22 93.78 6.69 No
6.30 0 6.30 93.70 7.57 No
6.51 0 6.51 93.49 9.06 No
6.21 0 6.21 93.79 4.55 Yes
6.37 0 6.37 93.63 3.82 No
6.36 0 6.36 93.64 3.89 No
6.71 0 6.71 93.29 7.37 Yes
6.76 0 6.76 93.24 7.91 No
6.11 0 6.11 93.89 4.98 Yes
6.72 0 6.72 93.28 10.87 Yes
6.39 0 6.39 93.61 12.42 No
6.77 0 6.77 93.23 8.11 No
6.19 0 6.19 93.81 6.52 No
6.38 0 6.38 93.62 10.87 Yes
6.29 0 6.29 93.71 12.64 Yes
6.27 0 6.27 93.73 4.83 Yes
6.49 0 6.49 93.51 6.69 No
6.36 0 6.36 93.64 6.67 Yes
6.49 0 6.49 93.51 10.87 Yes
6.31 0 6.31 93.69 6.36 Yes
6.46 0 6.46 93.54 6.52 Yes
6.63 0 6.63 93.37 7.03 No
6.75 0 6.75 93.25 4.83 Yes
6.49 0 6.49 93.51 6.36 No
6.39 0 6.39 93.61 5.27 No
6.24 0 6.24 93.76 4.57 Yes
6.53 0 6.53 93.47 4.28 No
6.29 0 6.29 93.71 4.69 Yes
6.47 0 6.47 93.53 6.36 No
6.34 0 6.34 93.66 6.20 Yes
6.38 0 6.38 93.62 10.04 No
6.31 0 6.31 93.69 10.87 Yes
6.65 0 6.65 93.35 10.41 No
6.24 0 6.24 93.76 4.02 No
6.72 0 6.72 93.28 4.08 Yes
6.49 0 6.49 93.51 4.28 No
6.87 0 6.87 93.13 3.95 Yes
6.79 0 6.79 93.21 5.27 No
6.65 0 6.65 93.35 4.83 No
6.61 0 6.61 93.39 5.57 No
6.64 0 6.64 93.36 4.98 Yes
7.08 0 7.08 92.92 4.83 Yes
6.49 0 6.49 93.51 4.69 No
6.90 0 6.90 93.10 5.43 Yes
6.47 0 6.47 93.53 5.25 Yes
6.74 0 6.74 93.26 4.83 No
6.68 0 6.68 93.32 4.83 No
7.08 0 7.08 92.92 4.28 No
6.72 0 6.72 93.28 1.66 No
7.37 0 7.37 92.63 1.55 No
6.65 0 6.65 93.35 2.80 No
6.67 0 6.67 93.33 1.31 No
6.87 0 6.87 93.13 3.34 No
6.51 0 6.51 93.49 1.11 No
6.93 0 6.93 93.07 4.34 No
6.86 0 6.86 93.14 2.88 No
6.85 0 6.85 93.15 1.81 Yes
6.31 0 6.31 93.69 3.38 No
6.61 0 6.61 93.39 5.95 Yes
6.80 0 6.80 93.20 4.84 No
6.74 0 6.74 93.26 4.13 Yes
6.38 0 6.38 93.62 2.78 No
6.40 0 6.40 93.60 3.05 No
6.91 0 6.91 93.09 1.80 No
7.09 0 7.09 92.91 4.53 Yes
6.44 0 6.44 93.56 2.28 No
6.20 0 6.20 93.80 5.83 No
199
CH4 cOrXi
%
H2 CO(~~i
%
Total flammables
cone. (%)
Air cone.
(%)
Energy
(mJ)
Success of
Ignition
6.23 0 6.23 93.77 5.16 No
6.66 0 6.66 93.34 4.32 No
6.92 0 6.92 93.08 3.56 No
6.72 0 6.72 93.28 4.23 No
6.66 0 6.66 93.34 3.57 No
6.53 0 6.53 93.47 2.27 No
6.80 0 6.80 93.20 2.88 No
6.80 0 6.80 93.20 5.00 Yes
6.70 0 6.70 93.30 2.04 No
6.80 0 6.80 93.20 6.92 No
6.94 0 6.94 93.06 4.53 No
6.38 0 6.38 93.62 4.84 No
6.91 0 6.91 93.09 2.27 Yes
6.88 0 6.88 93.12 2.41 No
6.86 0 6.86 93.14 2.64 No
7.03 0 7.03 92.97 3.21 No
6.35 0 6.35 93.65 2.38 No
6.34 0 6.34 93.66 2.43 No
6.38 0 6.38 93.62 2.20 Yes
6.69 0 6.69 93.31 2.41 No
6.77 0 6.77 93.23 2.57 No
6.85 0 6.85 93.15 1.82 Yes
6.62 0 6.62 93.38 2.13 Yes
6.40 0 6.40 93.60 2.35 No
6.38 0 6.38 93.62 1.80 No
6.61 0 6.61 93.39 2.20 No
6.79 0 6.79 93.21 3.74 Yes
6.69 0 6.69 93.31 3.13 No
6.79 0 6.79 93.21 2.57 No
6.58 0 6.58 93.42 3.29 Yes
6.58 0 6.58 93.42 3.08 No
6.73 0 6.73 93.27 0.40 No
6.89 0 6.89 93.11 0.55 No
6.72 0 6.72 93.28 0.50 No
6.77 0 6.77 93.23 0.78 No
7.01 0 7.01 92.99 0.39 No
6.57 0 6.57 93.43 0.30 No
7.07 0 7.07 92.93 0.94 No
6.96 0 6.96 93.04 1.17 No
6.91 0 6.91 93.09 0.50 No
6.43 0 6.43 93.57 0.87 No
6.85 '0 6.85 93.15 0.32 No
6.14 0 6.14 93.86 0.44 No
6.89 0 6.89 93.11 0.75 No
7.01 0 7.01 92.99 0.52 No
7.01 0 7.01 92.99 0.24 No
7.19 0 7.19 92.81 0.99 No
6.47 0 6.47 93.53 1.17 No
6.31 0 6.31 93.69 1.26 No
6.25 0 6.25 93.75 0.44 No
6.95 0 6.95 93.05 0.55 No
6.70 0 6.70 93.30 0.48 No
7.00 0 7.00 93.00 0.80 No
6.81 0 6.81 93.19 0.41 No
6.69 0 6.69 93.31 0.30 No
6.68 0 6.68 93.32 0.87 No
6.60 0 6.60 93.40 0.82 No
6.84 0 6.84 93.16 0.67 No
6.84 0 6.64 93.16 0.43 No
6.72 0 6.72 93.28 0.45 No
6.64 0 6.84 93.16 0.43 No
7.05 0 7.05 92.95 0.27 No
6.46 0 6.46 93.54 0.61 No
6.93 0 6.93 93.07 0.55 No
6.95 0 6.95 93.05 0.87 No
6.35 0 6.35 93.65 0.53 No
6.43 0 6.43 93.57 0.70 No
6.38 0 6.38 93.62 0.32 No
6.35 0 6.35 93.65 0.35 No
200
CH4 cone. H2 conc. Total flammables Air cone. Energy Success of
(%) (%) conc. ('lot ('to) (mJ) Ignitlon
7.24 0 7.24 92.76 0.78 No
6.74 0 6.74 93.26 0.32 No
6.70 0 6.70 93.30 1.17 No
6.45 0 6.45 93.55 0.23 No
6.42 0 6042 93.58 0040 No
6.72 0 6.72 93.28 0.55 No
6.94 0 6.94 93.06 0.50 No
6.80 0 6.80 93.20 0.78 No
6.92 0 6.92 93.08 0.46 No
6.60 0 6.60 93040 0.94 No
7.02 0 7.02 92.98 0.99 No
6.65 0 6.65 93.35 0.34 No
6.91 0 6.91 93.09 2.32 No
6.91 0 6.91 93.09 4.13 No
6.91 0 6.91 93.09 7.20 No
6.91 0 6.91 93.09 11.26 No
6.63 0 6.63 93.37 2.27 No
6.63 0 6.63 93.37 5.27 No
6.63 0 6.63 93.37 11.19 No
6.63 0 6.63 93.37 11040 No
'6.62 0 6.62 93.38 2.76 No
6.62 0 6.62 93.38 3.93 No
6.62 0 6.62 93.38 9.63 No
6.62 0 6.62 93.38 13.13 Yes
6.77 0 6.77 93.23 0.94 No
6.77 0 6.77 93.23 6.09 No
6.77 0 6.77 93.23 7.91 No
6.77 0 6.77 93.23 10.79 Yes
6.70 0 6.70 93.30 2.37 No
6.70 0 6.70 93.30 6.30 No
6.70 0 6.70 93.30 8.08 Yes
6.84 0 6.84 93.16 0.15 No
6.84 0 6.84 93.16 3.54 No
6.84 0 6.84 93.16 3.39 No
6.84 0 6.84 93.16 10.22 Yes
6.66 0 6.66 93.34 0.52 No
6.66 0 6.66 93.34 2.53 No
6.66 0 6.66 93.34 4.44 Yes
6.67 0 6.67 93.33 6.06 No
6.67 0 6.67 93.33 3.99 No
6.67 0 6.67 93.33 8.28 Yes
6.68 0 6.68 93.32 1.65 No
6.68 0 6.68 93.32 2.34 No
6.68 0 6.68 93.32 3.89 No
6.68 0 6.68 93.32 8.35 Yes
6.85 0 6.85 93.15 0.72 No

6.85 0 6.85 93.15 1.82 Yes
6.80 0 6.80 93.20 0.43 No
6.80 0 6.80 93.20 6.30 No
6.80 0 6.80 93.20 6.04 No
6.80 0 6.80 93.20 8.07 No
6.59 0 6.59 93.41 0.45 No

6.59 0 6.59 93041 8.09 No
6.59 0 6.59 93041 6.21 No
6.59 0 6.59 93.41 7041 No
201
CH4conc. H2conc. Total flammables Air cone. Energy Success of
(%) (%) conc. (%) (%) (mjj Ignition
6.58 0 6.58 93.42 1.43 No
6.58 0 6.58 93.42 2.20 No
6.58 0 6.58 93.42 3.64 No
6.58 0 6.58 93.42 8.54 Yes
6.42 0 6.42 93.58 0.34 No
6.42 0 6.42 93.58 1.92 Yes
6.88 0 6.88 93.12 0.32 No
6.88 0 6.88 93.12 1.94 No
6.88 0 6.88 93.12 3.04 No
6.88 0 6.88 93.12 7.41 No
6.55 0 6.55 93.45 0.30 No
6.55 0 6.55 93.45 2.06 No
6.55 0 6.55 93.45 3.76 No
6.55 0 6.55 93.45 8.31 No
6.51 0 6.51 93.49 0.40 No
6.51 0 6.51 93.49 5.95 No
6.51 0 6.51 93.49 3.76 No
6.51 0 6.51 93.49 7.61 No
6.52 0 6.52 93.48 0.28 No
6.52 0 6.52 93.48 1.98 No
6.52 0 6.52 93.48 4.42 No
6.52 0 6.52 93.48 8.54 No
6.92 0 6.92 93.08 0.72 No

6.92 0 6.92 93.08 2.64 No
6.92 0 6.92 93.08 4.53 Yes
6.66 0 6.66 93.34 2.42 No
6.66 0 6.66 93.34 3.48 No
6.66 0 6.66 93.34 4.96 No
6.66 0 6.66 93.34 8.84 Yes
6.69 0 6.69 93.31 1.82
(averaoe) (averaoe) (averaoe) (averaoe) (Lowest Eneroy)
7.71 0 7.71 92.29 0.17 No
7.71 0 7.71 92.29 1.12 No
7.71 0 7.71 92.29 1.45 No
7.71 0 7.71 92.29 3.74 Yes
7.80 0 7.80 92.20 0.34 No
7.80 0 7.80 92.20 2.70 No
7.80 0 7.80 92.20 2.48 Yes
7.80 0 7.80 92.20 0.00 No
7.72 0 7.72 92.28 0.27 No
7.72 0 7.72 92.28 1.47 No
7.72 0 7.72 92.28 3.88 No
7.72 0 7.72 92.28 6.73 No
7.76 0 7.76 92.24 0.35 No

7.76 0 7.76 92.24 2.81 No
7.76 0 7.76 92.24 2.20 Yes
7.76 0 7.76 92.24 0.00 No
7.75 0 7.75 92.25 0.27 No

7.75 0 7.75 92.25 1.78 No
7.75 0 7.75 92.25 11.74 No
7.75 0 7.75 92.25 5.32 No
7.77 0 7.77 92.23 0.30 No

7.77 0 7.77 92.23 0.32 No
7.77 0 7.77 92.23 1.48 Yes
7.95 0 7.95 92.05 0.30 No

7.95 0 7.95 92.05 0.34 No
202
CH4 conc.
(%l
H2 conc.
(%)
Total flammables
cone. (%)
Air conc.
(%) En(~~ Success of
Ignition
7.95 0 7.95 92.05 1.43 Yes
7.92 0 7.92 92.08 0.34 No

7.92 0 7.92 92.08 0.66 No
7.92 0 7.92 92.08 3.22 Yes
7.89 0 7.89 92.11 0.30 No

7.89 0 7.89 92.11 0.76 No
7.89 0 7.89 92.11 3.41 Yes
7.96 0 7.96 92.04 0.31 Yes
8.08 0 8.08 91.92 0.31 No

8.08 0 8.08 91.92 0.59 Yes
8.12 0 8.12 91.88 0.28 No

8.12 0 8.12 91.88 0.26 No
8.12 0 8.12 91.88 1.48 Yes
8.12 0 8.12 91.88 0.88 No

8.12 0 8.12 91.88 0.99 No
8.12 0 8.12 91.88 2.64 No
8.12 0 8.12 91.88 3.75 Yes
8.06 0 8.06 91.94 0.56 No

8.06 0 8.06 91.94 0.99 No
8.06 0 8.06 91.94 0.38 Yes
7.84 0 7.84 92.16 0.24 No

7.84 0 7.84 92.16 0.46 Yes
8.12 0 8.12 91.88 0.25 No

8.12 0 8.12 91.88 0.43 Yes
8.19 0 8.19 91.81 0.61 No

8.19 0 8.19 91.81 0.99 No
8.19 0 8.19 91.81 0.24 Yes
8.12 0 8.12 91.88 0.44 Yes

8.12 0 8.12 91.88 0.60 No
8.12 0 8.12 91.88 0.93 No
8.12 0 8.12 91.88 0.29 Yes
8.15 0 8.15 91.85 0.96 Yes

7.96 0 7.96 92.04 0.24
(average) (average) (average) (average) (Lowest Ene;gy)
8.63 0 8.63 91.37 0.17 No

8.63 0 8.63 91.37 1.41 Yes
8.66 0 8.66 91.34 0.16 No

8.66 0 8.66 91.34 0.55 Yes
8.64 0 8.64 91.36 0.19 No

8.64 0 8.64 91.36 1.73 No
8.64 0 8.64 91.36 4.54 No
8.64 0 8.64 91.36 5.55 No
8.74 0 8.74 91.26 0.27 No

8.74 0 8.74 91.26 0.55 No
8.74 0 8.74 91.26 2.27 Yes
8.75 0 8.75 91.25 0.93 No

8.75 0 8.75 91.25 0.27 Yes
8.79 0 8.79 91.21 0.42 No

8.79 0 8.79 91.21 0.22 Yes
8.81 0 8.81 91.19 0.32 Yes
203
CH4conc. H2 conc. Total fJammables Air conc. Energy Success of
('Iol (%) conc.(%) (%) (mJ) Ignition
8.99 0 8.99 91.01 0.45 Yes
8.85 0 8.85 91.15 0.30 Yes
9.07 0 9.07 90.93 0.29 Yes
9.11 0 9.11 90.89 0.34 No

9,11 0 9.11 90.89 0.61 Yes
9.14 0 9.14 90.86 0.53 No

9.14 0 9.14 90.86 1.51 Yes
9.14 0 9.14 90.86 0.43 No

9.14 0 9.14 90.86 1.04 Yes
9.02 0 9.02 90.98 0.27 No

9.02 0 9.02 90.98 0.46 Yes
9.03 0 9.03 90.97 0.41 No

9.03 0 9.03 90.97 0.70 Yes
9.07 0 9.07 90.93 0.44 No

9.07 0 9.07 90.93 0.45 Yes
9,18 0 9.18 90.82 0.33 No

9.18 0 9.18 90.82 0.44 Yes
9.21 0 9.21 90.79 0.70 No

9.21 0 9.21 90.79 0.36 Yes
9.13 0 9.13 90.87 0.16 No

9.13 0 9.13 90.87 0.76 Yes
9.11 0 9.11 90.89 0.49 No

9.11 0 9.11 90.89 0.87 No
9.11 0 9.11 90,89 0.82 Yes
8.95 0 8.95 91.05 0.22
(average) (average) (average) (average) (Lowest Energy)
10.38 0 10.38 89.62 0.82 No

10.38 0 10,38 89.62 0.99 Yes
10.65 0 10.65 89.35 0.93 No

10.65 0 10.65 89.35 3.41 Yes
10.57 0 10.57 89.43 0.68 No

10.57 0 10.57 89.43 5.02 Yes
10.57 0 10.57 89.43 0.22 No

10.57 0 10.57 89.43 2.06 No
10.57 0 10.57 89.43 5.16 Yes
10.78 0 10.78 89.22 1.04 No

10.78 0 10.78 89.22 0.83 Yes
10.71 0 10.71 89.29 0.70 No

10,71 0 10.71 89.29 1.28 No
10.71 0 10.71 89.29 3.04 Yes
10.60 0 10.60 89.40 1.17 No

10.60 0 10.60 89.40 2.32 No
10.60 0 10.60 89.40 7.99 No
10.60 0 10.60 89.40 6.20 Yes
10.63 0 10.63 89.37 0.23 No

10.63 0 10.63 89.37 2.42 No
10.63 0 10.63 89.37 3.64 No
10.63 0 10.63 89.37 7.84 No
10.64 0 10.64 89.36 0.60 No
204
CH4 conc. H2 conc. Total flammables Airconc. Energy Success of
(%) (%) conc. (%) (%) (mJ) h:anition
10.64 0 10.64 89.36 1.06 No
10.64 0 10.64 89.36 1.52 Yes
10.62 0 10.62 89.38 0.91 No

10.62 0 10.62 89.38 7.67 No
10.62 0 10.62 89.38 6.36 Yes
10.57 0 10.57 89.43 0.93 No

10.57 0 10.57 89.43 1.15 Yes
10.57 0 10.57 89.43 0.00 No
10.62 0 10.62 89.38 0.85 No

10.62 0 10.62 89.38 5.33 No
10.62 0 10.62 89.38 6.52 Yes
10.75 0 10.75 89.25 0.62 No

10.75 0 10.75 89.25 3.75 No
10.75 0 10.75 89.25 6.52 No
10.75 0 10.75 89.25 ·6.07 No
10.62 0 10.62 89.38 0.42 No

10.62 0 10.62 89.38 2.53 No
10.62 0 10.62 89.38 10.04 No
10.53 0 10.53 89.47 0.44 No

10.53 0 10.53 89.47 0.55 No
10.53 0 10.53 89.47 2.93 Yes
10.58 0 10.58 89.42 0.80 No

10.58 0 10.58 89.42 4.50 No
10.58 0 10.58 89.42 12.64 No
10.58 0 10.58 89.42 10.85 Yes
10.69 0 10.69 89.31 0.76 No

10.69 0 10.69 89.31 8.37 No
10.69 0 10.69 89.31 8.28 Yes
10.63 0 10.63 89.37 0.83 No

10.63 0 10.63 89.37 4.00 No
10.63 0 10.63 89.37 6.46 Yes
10.71 0 10.71 89.29 0.55 No

10.71 0 10.71 89.29 2.20 No
10.71 0 10.71 89.29 9.79 Yes
10.82 0 10.82 89.18 0.50 No

10.82 0 10.82 89.18 2.57 No
10.82 0 10.82 89.18 9.79 No
10.82 0 10.82 89.18 7.98 No
10.63 0 10.63 89.37 0.83
205
Appendix H (continued)
Table H.2 : Experimental runs with hydrogen- air mixture
H, cog~i CH, cO~~i Total flamma.b:~~ Air cO(~~i En~~~ Success of

% % cone. % % lonltion
5.48 0 5.48 94.52 0.73 No
5.48 0 5.48 94.52 0.51 No
5.48 0 5.48 94.52 1.56 No
5.48 0 5.48 94.52 6.36 No
5.76 0 5.76 94.24 0.51 No
5.76 0 5.76 94.24 4.00 No
5.76 0 5.76 94.24 8.76 No
5.76 0 5.76 94.24 10.08 No
5.77 0 5.77 94.23 1.39 No
5.77 0 5.77 94.23 7.91 No
5.77 0 5.77 94.23 6.10 No
5.77 0 5.77 94.23 12.54 No
5.92 0 5.92 94.08 1.18 No
5.92 0 5.92 94.08 2.87 No
5.92 0 5.92 94.08 6.69 No
5.92 0 5.92 94.08 9.79 No
6.00 0 6.00 94.00 0.79 No
6.00 0 6.00 94.00 2.32 No
6.00 0 6.00 94.00 12.42 No
6.00 0 6.00 94.00 8.76 No
6.12 0 6.12 93.88 0.69 No
6.12 0 6.12 93.88 2.37 No
6.12 0 6.12 93.88 12.64 No
6.12 0 6.12 93.88 9.53 No
6.17 0 6.17 93.83 0.58 No
6.17 0 6.17 93.83 3.41 No
6.17 0 6.17 93.83 3.50 No
6.17 0 6.17 93.83 13.76 No
6.16 0 6.16 93.84 1.39 No
6.16 0 6.16 93.64 7.91 No
6.16 0 6.16 93.84 6.10 No
6.16 0 6.16 93.84 11.96 Yes
6.19 0 6.19 93.81 0.37 No

6.19 0 6.19 93.81 2.23 No
6.19 0 6.19 93.81 4.06 No
6.19 0 6.19 93.81 4.00 No
6.22 0 6.22 93.78 0.44 No
6.22 0 6.22 93.78 2.53 No
6.22 0 6.22 93.78 11.52 No
6.22 0 6.22 93.78 9.28 No
6.25 0 6.25 93.75 0.44 No
6.25 0 6.25 93.75 2.53 No
6.25 0 6.25 93.75 11.52 No
6.25 0 6.25 93.75 9.28 No
6.30 0 6.30 93.70 1.06 No
6.30 0 6.30 93.70 0.76 Yes
6.40 0 6.40 93.60 0.18 No

6.40 0 6.40 93.60 1.65 No
6.40 0 6.40 93.60 6.07 No
6.40 0 6.40 93.60 6.10 No
206
Hl cone. CH" cone. Total flammables Air cone. Energy Success of
(%) ('/.t conc.(%) .(%) (mY) Ignition
6.54 0 6.54 93.46 0.50 No
6.54 0 6.54 93.46 2.37 No
6.54 0 6.54 93.46 11.09 No
6.54 0 6.54 93.46 9.28 No
6.78 0 6.78 93.22 1.06 No
6.78 0 6.78 93.22 6.30 No
6.78 0 6.78 93.22 8.99 No
6.78 0 6.78 93.22 10.22 No
6.82 0 6.82 93.18 0.18 No
6.82 0 6.82 93.18 1.65 No
6.82 0 6.82 93.18 6.07 No
6.82 0 6.82 93.18 6.10 No
6.90 0 6.90 93.10 1.67 No
6.90 0 6.90 93.10 4.69 No
6.90 0 6.90 93.10 8.62 No
6.90 0 6.90 93.10 9.04 No
6.92 0 6.92 93.08 0.33 No
6.92 0 6.92 93.08 2.83 No
6.92 0 6.92 93.08 2.36 No
6.92 0 6.92 93.08 10.87 No
6.99 0 6.99 93.01 1.18 No
6.99 0 6.99 93.01 2.87 No
6.99 0 6.99 93.01 5.25 No
6.99 0 6.99 93.01 5.60 Yes
6.98 0 6.98 93.02 1.87 No
6.98 0 6.98 93.02 4.43 No
6.98 0 6.98 93.02 7.20 No
6.98 0 6.98 93.02 10.88 No
6.32 0 6.32 93.67 0.76
9.44 0 9.44 90.56 0.28 No
9.44 0 9.44 90.56 1.31 No
9.44 0 9.44 90.56 3.14 No
9.44 0 9.44 90.56 9.01 No
9.59 0 9.59 90.41 0.26 No
9.59 0 9.59 90.41 0.22 No
9.59 0 9.59 90.41 1.05 Yes
9.61 0 9.61 90.39 0.25 No
9.61 0 9.61 90.39 1.08 Yes
9.74 0 9.74 90.26 0.10 Yes
9.58 0 9.58 90.42 0.37 No
9.58 0 9.58 90.42 0.34 No
9.58 0 9.58 90.42 1.02 Yes
9.64 0 9.64 90.36 0.30 No
9.64 0 9.64 90.36 0.34 No
9.64 0 9.64 90.36 1.48 Yes
9.75 0 9.75 90.25 0.17 No
9.75 0 9.75 90.25 0.70 No
9.75 0 9.75 90.25 2.27 No
9.75 0 9.75 90.25 2.32 Yes
9.94 0 9.94 90.06 0.41 No

9.94 0 9.94 90.06 0.31 No
9.94 0 9.94 90.06 2.11 No
9.94 0 9.94 90.06 0.62 Yes
10.10 0 10.10 89.90 0.38 No
207
----- ---
H2 cone. CH. cone. Total flammables Air cone. Energy Success of

(%) (%) conc. (%) (%) (mJ) Ignition
10.10 0 10.10 89.90 0.72 No
10.10 0 10.10 89.90 1.28 No
10.10 0 10.10 89.90 1.56 Ves
10.27 0 10.27 89.73 1.08 No

10.27 0 10.27 89.73 2.06 No
10.27 0 10.27 89.73 2.34 No
10.27 0 10.27 89.73 4.63 Ves
10.39 0 10.39 89.61 0.57 No

10.39 0 10.39 89.61 0.72 Ves
10.70 0 10.70 89.30 0.30 No

10.70 0 10.70 89.30 1.59 No
10.70 0 10.70 89.30 3.76 No
10.70 0 10.70 89.30 3.91 Ves
11.04 0 11.04 88.96 0.13 No

11.04 0 11.04 88.96 1.27 No
11.04 0 11.04 88.96 1.33 No
11.04 0 11.04 88.96 4.39 No
11.22 0 11.22 88.78 1.18 No

11.22 0 11.22 88.78 0.68 No
11.22 0 11.22 88.78 1.24 No
11.22 0 11.22 88.78 2.81 Ves
11.15 0 11.15 88.85 0.24 No

11.15 0 11.15 88.85 0.27 Ves
11.64 0 11.64 88.36 0.15 No

11.64 0 11.64 88.36 0.13 No
11.64 0 11.64 88.36 0.38 Ves
11.50 0 11.50 88.50 0.29 No

11.50 0 11.50 88.50 0.39 Ves
11.68 0 11.68 88.32 0.16 No

11.68 0 11.68 88.32 0.19 No
11.68 0 11.68 88.32 0.22 No
11.68 0 11.68 88.32 0.15 Ves
12.02 0 12.02 87.98 0.30 No

12.02 0 12.02 87.98 0.38 No
12.02 0 12.02 87.98 0.27 No
12.02 0 12.02 87.98 3.67 Ves
12.73 0 12.73 87.27 0.55 No

12.73 0 12.73 87.27 0.66 Ves
10.65 0 10.65 89.35 0.10
(averaoe) (averaoe) (average) (average) (Lowest Enerov)
19.51 0 19.51 80.49 0.37 Ves
19.67 0 19.67 80.33 0.33 Ves
19.78 0 19.78 80.22 0.15 Ves
20.13 0 20.13 79.87 0.13 Ves
20.31 0 20.31 79.69 0.20 No

20.31 0 20.31 79.69 0.14 Ves
20.46 0 20.46 79.54 0.20 Ves
20.95 0 20.95 79.05 0.16 Ves
20.30 0 20.30 79.70 0.11 No

20.30 0 20.30 79.70 0.07 Ves
208
H2 conc. CH 4 conc. Total flammables Air conc. Energy Success of
{%} {%} conc. {%} (%) (m:i) Ignition
21.10 0 21.10 78.90 0.27 No
21.10 0 21.10 78.90 0.14 Ves
21.70 0 21.70 78.30 0.15 No

21.70 0 21.70 78.30 0.12 Ves
21.87 0 21.87 78.13 0.20 No

21.87 0 21.87 78.13 0.14 Ves
21.18 0 21.18 78.82 0.13 Ves
21.19 0 21.19 78.81 0.04 No

21.19 0 21.19 78.81 0.04 Ves
21.25 0 21.25 78.75 0.87 Ves
21.25 0 21.25 78.75 0.55 Ves
21.28 0 21.28 78.72 0.08 Ves
22.51 0 22.51 77.49 0.36 Ves
21.56 0 21.56 78.44 0.20 No

21.56 0 21.56 78.44 0.14 Ves
21.60 0 21.60 78.40 0.15 Ves
21.72 0 21.72 78.28 0.09 Ves

20.96 0 20.96 79.04 0.04
(average) (average-r (average) (average) (Lowest Energy)
24.09 0 24.09 75.91 0.05 No

24.09 0 24.09 75.91 0.36 Ves
24.38 0 24.38 75.62 0.09 No

24.38 0 24.38 75.62 0.Q3 Ves
24.41 0 24.41 75.59 0.15 Ves
24.53 0 24.53 75.47 0.08 Ves
25.51 0 25.51 74.49 0.10 No

25.51 0 25.51 74.49 0.10 Ves
25.51 0 25.51 74.49 0.05 Ves
25.72 0 25.72 74.28 0.08 Ves
25.74 0 25.74 74.26 0.10 Ves
25.77 0 25.77 74.23 0.12 Ves
25.80 0 25.80 74.20 0.17 Ves
25.65 0 25.65 74.35 0.16 Ves
26.11 0 26.11 73.89 0.03 Ves
26.19 0 26.19 73.81 0.13 No

26.19 0 26.19 73.81 0.26 Ves
26.24 0 26.24 73.76 0.11 Ves
25.60 0 25.60 74.40 0.12 Ves
26.07 0 26.07 73.93 0.11 Ves
25.60 0 25.60 74.40 0.11 Ves
26.22 0 26.22 73.78 0.12 No
209
H2 conc. CH 4 conc. Total flammables Air conc. Energy Success of
(%) (%) conc. (%) (%) (mJ) h:lOition
26.22 0 26.22 73.78 0.12 Yes
25.76 0 25.76 74.24 0.08 Yes
25.87 0 25.87 74.13 0.12 Yes

25.47 0 25.47 74.53 0,03
(average) (average) (average) (average) (Lowest Eneray)
29.86 0 29.86 70.14 0.03 Yes
29.11 0 29.11 70.89 0.03 Yes
29.37 0 29.37 70.63 0.06 Yes
29.54 0 29.54 70.46 0.06 Yes
29.21 0 29.21 70.79 0.06 No

29.21 0 29.21 70.79 0.02 Yes
29.26 0 29.26 70.74 0.04 No

29.26 0 29.26 70.74 0.05 Yes
29.45 0 29.45 70.55 0.05 Yes
29.70 0 29.70 70.30 0.04 No

29.70 0 29.70 70.30 0.03 Yes
30.07 0 30.07 69.93 0.06 Yes
30.05 0 30.05 69.95 0.07 Yes
30.10 0 30.10 69.90 0.06 Yes
31.07 0 31.07 68.93 0.06 Yes
31.03 0 31.03 68.97 0.16 No

31.03 0 31.03 68.97 0.10 Yes
31.96 0 31.96 68.04 0.07 No

31.96 0 31.96 68.04 0.11 Yes
30.62 0 30.62 69.38 0.24 Yes
31.38 0 31.38 68.62 0.08 Yes
30.87 0 30.87 69.13 0.05 No

30.87 0 30.87 69.13 0,07 Yes
31.50 0 31.50 68.50 0.19 Yes
31.34 0 31.34 68.66 0.08 Yes
31.31 0 31.31 68.69 0.07 Yes

30.38 0 30.38 69.62 0.02
43.73 0 43.73 56.27 0.06 No

43.73 0 43.73 56.27 0.43 Yes
44.26 0 44.26 55.74 0.19 Yes
44.70 0 44.70 55.30 0.07 No

44.70 0 44.70 55.30 0.06 Yes
44.18 0 44.18 55.82 0.09 Yes
46.09 0 46.09 53.91 0.15 No

46.09 0 46.09 53.91 0.14 Yes
45.00 0 45.00 55.00 0.12 No
210
H2 conc. CH, conc: Total flammables Air conc. Energy Success of
(%) (%) conc. (%) (%) (mY) Ignition
45.00 0 45.00 55.00 0.15 Yes
44.99 0 44.99 55.01 0.22 Yes
45.73 0 45.73 54.27 0.19 No

45.73 0 45.73 54.27 0.21 Yes
46.34 0 46.34 53.66 0.17 No

46.34 0 46.34 53.66 0.23 Yes
46.49 0 46.49 53.51 0.26 No

46.49 0 46.49 53.51 0.29 Yes
47.00 0 47.00 53.00 0.32 No

47.00 0 47.00 53.00 0.28 Yes
47.36 0 47.36 52.64 0.12 No

47.36 0 47.36 52.64 0.08 Yes
45.41 0 45.41 54.59 0.56 Yes
43.32 0 43.32 56.68 0.35 Yes
45.09 0 45.09 54.91 0.31 Yes
46.51 0 46.51 53.49 0.19 No

46.51 0 46.51 53.49 0.22 Yes
46.78 0 46.78 53.22 0.32 Yes
46.68 0 46.68 53.32 0.19 No

46.68 0 46.68 53.32 0.26 Yes
47.68 0 47.68 52.32 0.21 Yes
47.48 0 47.48 52.52 0.19 No

47.48 0 47.48 52.52 0.37 Yes
45.68 0 45.68 54.32 0.06
59.24 0 59.24 40.76 0.45 No

59.24 0 59.24 40.76 0.44 No
59.24 0 59.24 40.76 1.47 No
59.24 0 59.24 40.76 2.64 Yes
58.02 0 58.02 41.98 0.19 No

58.02 0 58.02 41.98 0.30 No
58.02 0 58.02 41.98 0.59 No
58.02 0 58.02 41.98 0.95 Yes
59.99 0 59.99 40.01 0.21 No

59.99 0 59.99 40.01 0.22 No
59.99 0 59.99 40.01 0.22 Yes
59.71 0 59.71 40.29 0.17 No

59.71 0 59.71 40.29 0.20 No
59.71 0 59.71 40.29 0.75 No
59.71 0 59.71 40.29 1.09 Yes
60.27 0 60.27 39.73 0.27 No

60.27 0 60.27 39.73 0.26 Yes
61.16 0 61.16 38.84 0.18 No

61.16 0 61.16 38.84 0.17 No
61.16 0 61.16 38.64 0.73 Yes
62.30 0 62.30 37.70 0.18 No

62.30 0 62.30 37.70 0.21 No
62.30 0 62.30 37.70 0.25 Yes
211
H2 cone. CH 4 cone. Total flammables Air cone. Energy Success of
{'Iol {'M cone. ('to) {'M (mjj !9!1ition
61.21 0 61.21 38.79 0.25 No
61.21 0 61.21 38.79 0.17 No
61.21 0 61.21 38.79 0.27 No
61.21 0 61.21 38.79 0.28 Yes
61.31 0 61.31 38.69 0.18 No

61.31 0 61.31 38.69 1.07 No
61.31 0 61.31 38.69 2.91 No
61.31 0 61.31 38.69 4.55 Yes
63.00 0 63.00 37.00 0.17 No

63.00 0 63.00 37.00 0.98 No
63.00 0 63.00 37.00 3.13 No
63.00 0 63.00 37.00 4.28 Yes
63.10 0 63.10 36.90 0.65 No

63.10 0 63.10 36.90 1.39 No
63.10 0 63.10 36.90 1.86 Yes
63.24 0 63.24 36.76 0.17 No

63.24 0 63.24 36.76 1.06 No
63.24 0 63.24 36.76 3.34 Yes
63.54 0 63.54 36.46 0.17 No

63.54 0 63.54 36.46 1.05 No
63.54 0 63.54 36.46 2.93 Yes
63.62 0 63.62 36.38 0.16 No

63.62 0 63.62 36.38 1.02 No
63.62 0 63.62 36.38 2.08 No
63.62 0 63.62 36.38 4.96 Yes
63.68 0 63.68 36.32 0.25 No

63.68 0 63.68 36.32 1.07 No
63.68 0 63.68 36.32 3.23 Yes
64.04 0 64.04 35.96 0.29 No

64.04 0 64.04 35.96 1.27 No
64.04 0 64.04 35.96 3.21 Yes
64.75 0 64.75 35.25 0.75 No

64.75 0 64.75 35.25 2.64 No
64.75 0 64.75 35.25 3.05 No
64.75 0 64.75 35.25 3.13 Yes
63.17 0 63.17 36.83 1.08 No

63.17 0 63.17 36.83 4.94 No
63.17 0 63.17 36.83 3.34 Yes
62.56 0 62.56 37.44 0.25 No

62.56 0 62.56 37.44 1.12 No
62.56 0 62.56 37.44 3.34 Yes
63.64 0 63.64 36.36 0.21 No

63.64 0 63.64 36.36 1.16 No
63.64 0 63.64 36.36 3.34 No
63.64 0 63.64 36.36 4.83 No
62.11 0 62.11 37.89 0.22
. (aver"lJ~ (averaae) _(averaae) (averaae) (Lowest Eneray)
212
Table H.3 : Experimental runs with 75 % methane + 25 % Hydrogen with air

CH4conc. H2 cone. Total flammables Air cone. Energy Success of
(%) (%) conc. (%) (%) (mjj lanitlon
6.18 1.99 8.17 91.83 0.19 No
6.18 1.99 8.17 91.83 1.27 Yes
6.30 2.29 8.60 91.40 0.31 No

6.30 2.29 8.60 91.40 4.35 Yes
6.34 2.31 8.65 91.35 0.70 Yes
6.05 1.79 7.84 92.16 0.25 No

6.05 1.79 7.84 92.16 0.98 No
6.05 1.79 7.84 92.16 3.23 No
6.05 1.79 7.84 92.16 9.63 No
5.99 2.02 8.00 92.00 0.30 No

5.99 2.02 8.00 92.00 3.10 Yes
6.28 1.77 8.05 91.95 0.76 No

6.28 1.77 8.05 91.95 3.87 Yes
6.21 2.00 8.22 91.78 0.25 No

6.21 2.00 8.22 91.78 1.20 Yes
6.21 1.77 7.98 92.02 0.19 No

6.21 1.77 7.98 92.02 1.39 Yes
5.81 2.25 8.06 91.94 0.58 Yes
6.12 1.79 7.92 92.08 0.61 Yes
5.91 2.07 7.98 92.02 0.58 Yes
5.61 2.06 7.67 92.33 0.62 Yes
6.02 2.25 8.27 91.73 0.62 Yes
5.72 1.75 7.47 92.53 0.36 No

5.72 1.75 7.47 92.53 1.56 No
5.72 1.75 7.47 92.53 5.05 Yes
6.21 2.35 8.56 91.44 0.61 Yes
5.81 1.85 7.65 92.35 0.65 Yes
5.95 1.75 7.70 92.30 0.42 No

5.95 1.75 7.70 92.30 2.32 Yes
5.97 1.89 7.86 92.14 0.73 Yes
5.56 2.04 7.60 92.40 0.24 No

5.56 2.04 7.60 92.40 1.27 Yes
5.90 2.09 7.99 92.01 0.53 No

5.90 2.09 7.99 92.01 1.02 Yes
6.01 2.01 8.02 91.98 0.27
8.44 2.61 11.05 88.95 0.11 Yes
8.35 2.20 10.55 89.45 0.13 Yes
7.91 2.17 10.07 89.93 0.10 Yes
8.23 2.74 10.97 89,03 0.11 Yes
213
CH4conc. H2 conc. Total flammables Airconc. Energy Success of
(%) (%) conc. (%) (%) (mJ) lanltlon
7.81 2.39 10.20 89.80 0.32 No
7.81 2.39 10.20 89.80 0.49 Yes
8.23 2.96 11.19 88.81 0.12 Yes
7.60 2.98 10.58 89.42 0.11 Yes
8.46 2.58 11.04 88.96 0.10 Yes
7.97 2.59 10.56 89.44 0.16 Yes
8.18 2.61 10.79 89.21 0.11 Yes
7.54 2.87 10.41 89.59 0.12 Yes
8.09 2.25 10.35 89.65 0.28 No

8.09 2.25 10.35 89.65 0.21 No
8.09 2.25 10.35 89.65 1.78 Yes
7.58 2.39 9.97 90.03 0.10 Yes
8.27 2.83 11.10 88.90 0.15 Yes
8.05 2.44 10.49 89.51 0.11 Yes
8.08 2.81 10.89 89.11 0.21 Yes
8.33 2.91 11.24 88.76 0.39 No

8.33 2.91 11.24 88.76 0.42 No
8.33 2.91 11.24 88.76 2.32 Yes
8.05 2.75 10.81 89.19 0.33 No

8.05 2.75 10.81 89.19 0.27 No
8.05 2.75 10.81 89.19 1.24 Yes
7.39 2.68 10.07 89.93 1.06 Yes
8.00 2.98 10.97 89.03 0.16 Yes

8.03 2.64 10.67 89.33 0.10
9.70 3.17 12.87 87.13 0.31 No

9.70 3.17 12.87 87.13 0.32 No
9.70 3.17 12.87 87.13 0.52 Yes
9.77 3.22 12.98 87.02 0.25 No

9.77 3.22 12.98 87.02 0.40 No
9.77 3.22 12.98 87.02 2.32 Yes
10.26 3.24 13.50 86.50 0.42 No

10.26 3.24 13.50 86.50 1.69 No
10.26 3.24 13.50 86.50 4.34 Yes
9.79 3.38 13.17 86.83 0.42 No

9.79 3.38 13.17 86.83 0.49 No
9.79 3.38 13.17 86.83 1.27 Yes
10.50 3.10 13.60 86.40 0.52 No

10.50 3.10 13.60 86.40 3.94 Yes
9.66 3.18 12.64 87.16 0.39 No

9.66 3.18 12.84 87.16 3.22 No
9.66 3.18 12.84 87.16 7.83 Yes
10.32 3.15 13.47 86.53 0.42 No

10.32 3.15 13.47 86.53 2.27 Yes
10.54 3.87 14.42 85.58 0.36 No

10.54 3.87 14.42 85.58 1.35 Yes
214
('to) ('to) conc. ('to) ('to) (mJ) lanition
10.02 3.46 13.48 86.52 0.20 No
10.02 3.46 13.48 86.52 1.16 No
10.02 3.46 13.48 86.52 2.93 Yes
10.08 3.53 13.61 86.39 0.36 No
10.08 3.53 13.61 86.39 1.65 No
10.08 3.53 13.61 86.39 3.88 Yes
9.71 3.20 12.91 87.09 0.26 No
9.71 3.20 12.91 87.09 1.56 No
9.71 3.20 12.91 87.09 3.99 Yes
10.47 3.07 13.55 86.45 0.65 No
10.47 3.07 13.55 86.45 1.97 No
10.47 3.07 13.55 86.45 2.20 Yes
10.03 3.67 13.70 86.30 0.25 No
10.03 3.67 13.70 86.30 2.75 No
10.03 3.67 13.70 86.30 5.61 Yes
10.17 3.04 13.21 86.79 0.88 No
10.17 3.04 13.21 86.79 4.57 No
10.17 3.04 13.21 86.79 5.48 Yes
10.57 3.68 14.25 85.75 0.43 No
10.57 3.68 14.25 85.75 2.02 No
10.57 3.68 14.25 85.75 3.55 No
10.57 3.68 14.25 85.75 9.63 Yes
9.93 3.73 13.66 86.34 0.73 No
9.93 3.73 13.66 86.34 4.35 No
9.93 3.73 13.66 86.34 5.89 Yes
10.35 3.55 13.90 86.10 0.43 No
10.35 3.55 13.90 86.10 3.67 No
10.35 3.55 13.90 86.10 5.75 No
10.35 3.55 13.90 86.10 10.04 No
9.67 3.65 13.32 86.68 0.55 No
9.67 3.65 13.32 86.68 3.60 No
9.67 3.65 13.32 86.68 4.94 Yes
9.78 3.78 13.55 86.45 0.44 No
9.78 3.78 13.55 86.45 3.87 No
9.78 3.78 13.55 86.45 8.99 Yes
10.59 3.19 13.78 86.22 0.61 No

10.59 3.19 13.78 86.22 2.54 No
10.59 3.19 13.78 86.22 5.35 No
10.59 3.19 13.78 86.22 7.55 No
10.1 3.39 13.49 86.51 0.98
(average) (average) (average') (average) (Lowest Energy)
11.90 4.32 16.22 83.78 0.58 No
11.90 4.32 16.22 83.78 2.16 No
11.90 4.32 16.22 83.78 4.71 No
11.90 4.32 16.22 83.78 10.41 Yes
12.70 4.21 16.91 83.09 0.39 No
12.70 4.21 16.91 83.09 2.06 No
12.70 4.21 16.91 83.09 4.98 No
12.70 4.21 16.91 83.09 6.94 Yes
12.33 4.19 16.52 83.48 0.99 No

12.33 4.19 16.52 83.48 4.94 No
12.33 4.19 16.52 83.48 8.15 No
12.33 4.19 16.52 83.48 10.04 No
12.89 3.99 16.88 83.12 0.56 No

12.89 3.99 16.88 83.12 2.64 No
215
CH4 conc. H2 conc. Total fJammables Air conc. Energy Success of
(%) (%) conc. (%) (%) (mJ) lanltion
12.89 3.99 16.88 83.12 7.05 No
12.89 3.99 16.88 83.12 9.23 Yes
12.16 4.65 16.82 83.18 0.34 No

12.16 4.65 16.82 83.18 2.31 No
12.16 4.65 16.82 83.18 5.19 Yes
12.69 4.05 16.74 83.26 0.46 No

12.69 4.05 16.74 83.26 3.29 No
12.69 4.05 16.74 83.26 4.71 No
12.69 4.05 16.74 83.26 7.55 No
12.27 4.67 16.94 83.06 0.72 No
12.27 4.67 16.94 83.06 3.80 No
12.27 4.67 16.94 83.06 7.83 No
12.27 4.67 16.94 83.06 6.65 No
12.88 4.20 17.09 82.91 0.53 No
12.88 4.20 17.09 82.91 2.81 No
12.88 4.20 17.09 82.91 6.18 No
12.88 4.20 17.09 82.91 11.74 Yes
12.37 4.30 16.67 83.33 0.60 No

12.37 4.30 16.67 83.33 1.78 No
12.37 4.30 16.67 83.33 13.58 No
12.37 4.30 16.67 83.33 9.79 No
12.32 4.41 16.72 83.28 0.75 No

12.32 4.41 16.72 83.28 2.98 No
12.32 4.41 16.72 83.28 7.83 No
12.32 4.41 16.72 83.28 11.74 No
12.41 4.60 17.01 82.99 0.48 No

12.41 4.60 17.01 82.99 4.33 No
12.41 4.60 17.01 82.99 9.43 Yes
12.76 3.96 16.73 83.27 0.41 No

12.76 3.96 16.73 83.27 2.32 No
12.76 3.96 16.73 83.27 5.89 No
12.76 3.96 16.73 83.27 6.36 No
12.60 4.12 16.71 83.29 0.44 No

12.60 4.12 16.71 83.29 2.81 No
12.60 4.12 16.71 83.29 6.17 No
12.60 4.12 16.71 83.29 10.04 No
12.47 4.72 17.19 82.81 0.78 No

12.47 4.72 17.19 82.81 4.50 No
12.47 4.72 17.19 82.81 9.16 No
12.47 4.72 17.19 82.81 8.45 No
12.30 4.12 16.42 83.58 0.34 No
12.30 4.12 16.42 83.58 2.37 No
12.30 4.12 16.42 83.58 5.75 No
12.30 4.12 16.42 83.58 6.04 No
12.07 3.96 16.03 83.97 0.20 No
12.07 3.96 16.03 83.97 2.32 No
12.07 3.96 16.03 83.97 5.89 No
12.07 3.96 16.03 83.97 6.36 No
12.00 4.34 16.33 83.67 0.42 No

12.00 4.34 16.33 83.67 4.13 No
12.00 4.34 16.33 83.67 10.87 No
12.28 3.96 16.23 83.77 1.33 No
12.28 3.96 16.23 83.77 4.43 No
12.28 3.96 16.23 83.77 8.82 No
12.28 3.96 16.23 83.77 10.49 No
I
216
•
En(~~~
CH4conc. H2 cone. Total flammables Air cone. Success of
(%) (%) conc. (%) (%) Ignition
12.01 4.11 16.11 83.89 0.33 No
12.01 4.11 16.11 83.89 1.78 No
12.01 4.11 16.11 83.89 5.89 No
12.01 4.11 16.11 83.89 4.62 No
13.02 4.66 17.67 82.33 0.73 No
13.02 4.66 17.67 82.33 3.35 No
13.02 4.66 17.67 82.33 6.17 No
13.02 4.66 17.67 82.33 9.23 No
12.42 4.28 16.70 83.30 5.19
217
Table HA : Experimental runs with 50% methane +50% Hydrogen with air
CH4 cone. H2 cone. Total flammables Air cone. Energy Success of

(%) (%) conc. (%) (%) (mJ) lonlllon
2.80 2.83 5.64 94.36 0.40 No
2.80 2.83 5.64 94.36 4.43 No
2.80 2.83 5.64 94.36 11.74 Yes
3.06 3.17 6.23 93.77 0.44 No

3.06 3.17 6.23 93.77 4.18 No
3.06 3.17 6.23 93.77 12.42 Yes
3.00 3.48 6.48 93.52 0.36 No

3.00 3.48 6.48 93.52 3.67 No
3.00 3.48 6.48 93.52 9.23 Yes
2.81 2.83 5.64 94.36 0.42 No

2.81 2.83 5.64 94.36 2.47 No
2.81 2.83 5.64 94.36 10.41 Yes
3.15 3.44 6.59 93.41 0.94 No

3.15 3.44 6.59 93.41 5.95 No
3.15 3.44 6.59 93.41 8.48 No
3.15 3.44 6.59 93.41 8.54 No
2.71 3.20 5.91 94.09 0.08 No
2.71 3.20 5.91 94.09 2.54 No
2.71 3.20 5.91 94.09 5.09 Yes
2.71 3.20 5.91 94.09 0.00 No
3.40 3.21 6.61 93.39 0.87 No
3.40 3.21 6.61 93.39 3.53 No
3.40 3.21 6.61 93.39 7.83 No
3.40 3.21 6.61 93.39 11.26 No
2.81 3.24 6.05 93.95 0.20 No
2.81 3.24 6.05 93.95 0.82 No
2.81 3.24 6.05 93.95 2.72 Yes
3.37 3.38 6.75 93.25 0.72 No

3.37 3.38 6.75 93.25 6.92 No
3.37 3.38 6.75 93.25 6.57 No
3.37 3.38 6.75 93.25 12.25 No
3.30 2.79 6.09 93.91 0.87 No
3.30 2.79 6.09 93.91 3.80 No
3.30 2.79 6.09 93.91 8.31 No
3.30 2.79 6.09 93.91 10.58 No
2.76 3.11 5.87 94.13 0.72 No
2.76 3.11 5.87 94.13 5.72 No
2.76 3.11 5.87 94.13 7.51 No
2.76 3.11 5.87 94.13 11.40 No
2.76 3.40 6.17 93.83 0.42 No

2.76 3.40 6.17 93.83 1.65 No
2.76 3.40 6.17 93.83 13.34 Yes
3.32 3.31 6.63 93.37 0.50 No

3.32 3.31 6.63 93.37 1.83 No
3.32 3.31 6.63 93.37 8.90 Yes
2.68 2.80 5.48 94.52 0.99 No

2.68 2.80 5.48 94.52 4.07 No
2.68 2.80 5.48 94.52 8.15 No
2.68 2.80 5.48 94.52 8.54 No
218
CH4 conc. H2 conc. Total flammables Air cone. Energy Success of
(~J ('tol conc. ('tol ('tol (m:il lanition
3.18 3.16 6.35 93.65 0.45 No
3.18 3.16 6.35 93.65 1.60 No
3.18 3.16 6.35 93.65 12.64 No
3.18 3.16 6.35 93.65 9.41 No
3.07 3.39 6.46 93.54 0.33 No

3.07 3.39 6.46 93.54 2.71 No
3.07 3.39 6.46 93.54 9.43 Yes
3.30 2.70 5.99 94.01 0.34 No

3.30 2.70 5.99 94.01 3.30 No
3.30 2.70 5.99 94.01 8.46 Yes
3.30 3.29 6.59 93.41 0.50 No

3.30 3.29 6.59 93.41 1.65 No
3.30 3.29 6.59 93.41 11.52 No
3.30 3.29 6.59 93.41 13.13 No
2.83 3.02 5.85 94.15 0.41 No

2.83 3.02 5.85 94.15 2.53 No
2.83 3.02 5.85 94.15 10.87 Yes
2.64 2.82 5.46 94.54 0.28 No

2.64 2.82 5.46 94.54 1.83 No
2.64 2.82 5.46 94.54 6.79 Yes
3.01 3.13 6.14 93.86 2.72
. (aver~ge) (averaael (averaael Javeragel (Lowest Enerayl
5.38 4.85 10.23 89.77 0.22 No

5.38 4.85 10.23 89.77 1.69 No
5.38 4.85 10.23 89.77 2.88 Yes
4.90 4.68 9.58 90:42 0.31 Yes
5.17 5.24 10.41 89.59 0.25 Yes
4.76 5.05 9.82 90.18 0.18 Yes
4.94 5.34 10.28 89.72 0.55 No

4.94 5.34 10.28 89.72 1.92 No
4.94 5.34 10.28 89.72 2.34 Yes
5.29 5.38 10.67 89.33 0.19 Yes
5.10 5.36 10.46 89.54 0.17 Yes
4.62 5.10 9.72 90.28 0.43 Yes
4.22 5.31 9.53 90.47 0.49 Yes
4.89 5.07 9.96 90.04 0.70 No
4.89 5.07 9.96 90.04 3.16 Yes
4.72 5.13 9.85 90.15 0.35 No

4.72 5.13 9.85 90.15 1.27 Yes
4.80 5.19 9.99 90.01 0.22 Yes
4.73 4.73 9.46 90.54 0.34 Yes
5.31 5.11 10.41 89.59 1.01 Yes
4.87 4.79 9.67 90.33 1.01 No

4.87 4.79 9.67 90.33 1.39 Yes
4.89 5.09 9.98 90.02 0.17 Yes
5.36 5.24 10.60 89.40 0.16 Yes
219
(%) (%) conc. (%) (%) (mJ) lanition
5.54 4.51 10.05 89.95 0.72 No
5.54 4.51 10.05 89.95 2.81 Yes
5.19 4.71 9.90 90.10 0.24 Yes
5.12 4.84 9.97 90.03 0.17 Yes

4.99 5.04 10.03 89.97 0.16
6.68 6.92 13.60 86.40 0.09 Yes
7.00 6.94 13.93 86.07 0.12 Yes
6.83 6.47 13.30 86.70 0.09 Yes
7.07 6.63 13.70 86.30 0.12 Yes
6.70 6.88 13.58 86.42 0.41 No

6.70 6.88 13.58 86.42 1.04 Yes
7.31 6.94 14.25 85.75 0.10 Yes
6.74 6.85 13.58 86.42 0.17 Yes
6.66 6.67 13.33 86.67 0.11 Yes
6.39 6.62 13.01 86.99 0.10 Yes
7.03 6.77 13.79 86.21 0.39 No

7.03 6.77 13.79 86.21 0.32 Yes
6.49 6.35 12.84 87.16 0.11 Yes
6.79 6.36 13.15 86.85 0.12 Yes
6.68 6.45 13.13 86.87 0.11 Yes
6.84 6.68 13.52 86.48 0.27 No

6.84 6.68 13.52 86.48 0.28 Yes
6.72 6.63 13.35 86.65 0.12 Yes
6.62 6.57 13.19 86.81 0.10 Yes
6.65 6.49 13.15 86.85 0.15 Yes
7.09 6.84 13.93 86.07 0.10 Yes
6.79 6.70 13.49 86.51 0.40 No

6.79 6.70 13.49 86.51 0.55 No
6.79 6.70 13.49 86.51 2.16 Yes
6.84 6.85 13.69 86.31 0.35 No

6.84 6.85 13.69 86.31 0.66 No
6.84 6.85 13.69 86.31 2.31 Yes
6.80 6.68 13.48 86.52 0.09
7.88 8.21 16.09 83.91 0.14 No

7.88 8.21 16.09 83.91 0.31 No
7.88 8.21 16.09 83.91 2.36 Yes
8.12 8.Q1 16.12 83.88 0.13 Yes
8.17 8.24 16.41 83.59 0.10 Yes
8.06 8.17 16.22 83.78 0.13 Yes
7.84 . 8.30 16.14 83.86 0.41 No

7.84 8.30 16.14 83.86 1.04 Yes
220
(%) (%) conc. ('Iol (%) {mJ>- Ignition
8.24 7.95 16.19 83.81 0.11 Yes
8.41 7.74 16.15 83.85 0.10 Yes
8.53 8.37 16.91 83.09 0.12 Yes
7.88 7.92 15.80 84.20 0.11 Yes
7.73 7.84 15.57 84.43 0.42 No

7.73 7.84 15.57 84.43 0.31 Yes
8.37 7.88 16.25 83.75 0.11 Yes
8.07 8.04 16.12 83.88 0.11 Yes
8.06 8.24 16.30 83.70 0.11 Yes
8.19 8.08 16.27 83.73 0.25 No

8.19 8.08 16.27 83.73 0.26 Yes
8.29 8.16 16.44 83.56 0.13 Yes
7.83 8.35 16.18 83.82 0.10 Yes
8.39 8.12 16.51 83.49 0.14 Yes
8.02 8.14 16.16 83.84 0.10 Yes
7.82 7.70 15.52 84.48 0.22 No

7.82 7.70 15.52 84.48 0.62 No
7.82 7.70 15.52 84.48 2.98 Yes
7.83 7.93 15.76 84.24 0.36 No

7.83 7.93 15.76 84.24 0.62 No
7.83 7.93 15.76 84.24 2.81 Yes
8.09 8.07 16.16 83.84 0.10
10.36 9.63 19.99 80.01 0.27 No

10.36 9.63 19.99 80.Q1 4.53 No
10.36 9.63 19.99 80.Q1 3.77 No
10.36 9.63 19.99 80.Q1 7.73 No
10.81 10.10 20.91 79.09 1.06 No

10.81 10.10 20.91 79.09 3.87 No
10.81 10.10 20.91 79.09 4.11 Yes
9.49 10.01 19.50 80.50 2.04 No

9.49 10.01 19.50 80.50 2.37 No
9.49 10.01 19.50 80.50 4.83 No
9.49 10.01 19.50 80.50 11.82 No
10.26 10.53 20.79 79.21 0.99 No

10.26 10.53 20.79 79.21 0.41 No
10.26 10.53 20.79 79.21 1.24 No
10.26 10.53 20.79 79.21 6.65 No
10.40 10.04 20.44 79.56 0.72 No

10.40 10.04 20.44 79.56 6.18 No
10.40 10.04 20.44 79.56 6.70 No
10.40 10.04 20.44 79.56 9.53 No
9.51 10.39 19.89 80.11 0.36 No

9.51 10.39 19.89 80.11 2.42 No
9.51 10.39 19.89 80.11 4.73 No
9.51 10.39 19.89 80.11 5.12 No
10.11 9.98 20.09 79.91 0.88 No

10.11 9.98 20.09 79.91 1.87 No
10.11 9.98 20.09 79.91 4.23 No
221
CH4 conc. H2 conc.
En(~~
Total flammables Air conc . Success of
. (%1 (0t.). conc. (%) (%) Ignition
10.11 9.98 20.09 79.91 7.16 No
9.71 9.67 19.38 80.62 0.32 No
9.71 9.67 19.38 80.62 2.82 No
9.71 9.67 19.38 80.62 5.28 No
9.71 9.67 19.38 80.62 11.62 No
9.96 10.05 20.01 79.99 0.34 No
9.96 10.05 20.01 79.99 3.30 No
9.96 10.05 20.01 79.99 8.65 Yes
10.02 10.57 20.60 79.40 0.19 No

10.02 10.57 20.60 79.40 6.18 No
10.02 10.57 20.60 79.40 3.05 No
10.02 10.57 20.60 79.40 6.54 No
10.10 10.26 20.36 79.64 0.69 No
10.10 10.26 20.36 79.64 2.01 No
10.10 10.26 20.36 79.64 6.87 No
10.10 10.26 20.36 79.64 7.16 Yes
10.28 9.66 19.94 80.06 0.32 No

10.28 9.66 19.94 80.06 1.43 No
10.28 9.66 19.94 80.06 6.52 No
10.28 9.66 19.94 80.06 11.96 No
10.48 10.04 20.51 79.49 0.41 No
10.48 10.04 20.51 79.49 1.92 No
10.48 10.04 20.51 79.49 3.77 No
10.48 10.04 20.51 79.49 5.42 No
9.46 9.88 19.34 80.66 0.29 No
9.46 9.88 19.34 80.66 1.87 No
9.46 9.88 19.34 80.66 4.58 No
9.46 9.88 19.34 80.66 10.05 No
9.94 9.73 19.67 80.33 0.99 No
9.94 9.73 19.67 80.33 3.80 No
9.94 9.73 19.67 80.33 9.88 No
9.94 9.73 19.67 80.33 11.40 Yes
9.45 10.15 19.60 80.40 0.45 No

9.45 10.15 19.60 80.40 1.60 No
9.45 10.15 19.60 80.40 12.64 No
9.45 10.15 19.60 80.40 9.41 No
10.04 9.53 19.57 80.43 0.44 No
10.04 9.53 19.57 80.43 3.75 No
10.04 9.53 19.57 80.43 8.09 No
10.43 9.66 20.09 79.91 0.34 No
10.43 9.66 20.09 79.91 2.72 No
10.43 9.66 20.09 79.91 11.30 Yes
10.38 9.59 19.97 80.03 0.15 No

10.38 9.59 19.97 80.03 1.92 No
10.38 9.59 19.97 80.03 3.61 No
10.38 9.59 19.97 80.03 8.54 No
10.76 10.57 21.32 78.68 0.76 No
10.76 10.57 21.32 78.68 3.35 No
10.76 10.57 21.32 78.68 2.74 No
10.76 10.57 21.32 78.68 10.66 No
12.00 12.10 24.10 75.90 0.99 No

12.00 12.10 24.10 75.90 2.52 No
12.00 12.10 24.10 75.90 3.57 No
12.00 12.10 24.10 75.90 6.04 No
12.00 12.10 24.10 75.90 7.56 No
222
(%) (%) conc. (%) (%) (mJ) Ignition
15.07 16.42 31.49 68.51 0.34 No
15.07 16.42 31.49 68.51 2.27 No
15.07 16.42 31.49 68.51 3.41 No
15.07 16.42 31.49 68.51 8.81 No
18.95 19.63 38.59 61.41 0.51 No

18.95 19.63 38.59 61.41 2.11 No
18.95 19.63 38.59 61.41 6.36 No
18.95 19.63 38.59 61.41 6.52 No
10.10 10.00 20.10 79.90 4.11
223
Table H.5 : Experimental runs with 25% methane + 75% Hydrogen with air
Hz cone. CH" cone. Total flammables Air cone. Energy Success of

(%) (%) conc. (%) (%) (mJ) lonitlon
5.97 2.40 8.37 91.63 0.72 No
5.97 2.40 8.37 91.63 6.92 No
5.97 2.40 8.37 91.63 9.68 No
5.97 2.40 8.37 91.63 8.76 No
6.01 2.63 8.64 91.36 0.56 No
6.01 2.63 8.64 91.36 6.30 No
6.01 2.63 8.64 91.36 9.16 No
6.01 2.63 8.64 91.36 10.31 No
6.07 2.29 8.36 91.64 0.94 No
6.07 2.29 8.36 91.64 4.14 Yes
6.46 1.96 8.42 91.58 0.36 No

6.46 1.96 8.42 91.58 2.98 No
6.46 1.96 8.42 91.58 5.75 No
6.46 1.96 8.42 91.58 9.53 Yes
6.22 2.67 8.89 91.11 0.25 No

6.22 2.67 8.89 91.11 2.75 No
6.22 2.67 8.89 91.11 5.75 No
6.22 2.67 8.89 91.11 7.38 Yes
6.17 2.29 8.45 91.55 0.61 No

6.17 2.29 8.45 91.55 5.38 No
6.17 2.29 8.45 91.55 8.65 Yes
5.87 1.92 7.79 92.21 0.65 No

5.87 1.92 7.79 92.21 2.06 No
5.87 1.92 7.79 92.21 9.68 Yes
6.09 2.20 8.30 91.70 0.87 No

6.09 2.20 8.30 91.70 3.80 No
6.09 2.20 8.30 91.70 12.73 No
6.09 2.20 8.30 91.70 9.28 No
5.75 2.00 7.75 92.25 0.42 No

5.75 2.00 7.75 92.25 2.47 No
5.75 2.00 7.75 92.25 9.06 Yes
6.31 1.88 8.20 91.80 0.73 No

6.31 1.88 8.20 91.80 5.95 No
6.31 1.88 8.20 91.80 12.73 No
6.31 1.88 8.20 91.80 10.58 No
6.22 2.39 8.61 91.39 0.42 No
6.22 2.39 8.61 91.39 1.65 No
6.22 2.39 8.61 91.39 10.41 Yes
6.45 2.57 9.02 90.98 0.40 No

6.45 2.57 9.02 90.98 3.48 No
6.45 2.57 9.02 90.98 11.74 Yes
6.42 2.40 8.81 91.19 0.76 No

6.42 2.40 8.81 91.19 3.48 No
6.42 2.40 8.81 91.19 10.97 No
6.42 2.40 8.81 91.19 11.96 No
6.26 2.16 8.41 91.59 0.20 No
6.26 2.16 8.41 91.59 0.82 No
6.26 2.16 8.41 91.59 2.72 No
6.26 2.16 8.41 91.59 6.44 Yes
224
Hz cone. CH. cone. Total flammables
Air cO(~~i Energy Success of
- (%) (%) conc. (%) % (mj) Ignition
6.23 2.27 8.50 91.50 0.87 No
6.23 2.27 8.50 91.50 3.53 No
6.23 2.27 8.50 91.50 12.73 No
6.23 2.27 8.50 91.50 13.43 No
6.04 2.18 8.23 91.77 0.44 No

6.04 2.18 8.23 91.77 4.07 No
6.04 2.18 8.23 91.77 12.42 Yes
5.82 2.32 8.14 91.86 0.36 No

5.82 2.32 8.14 91.86 3.67 No
5.82 2.32 8.14 91.86 12.19 No
5.90 2.51 8.41 91.59 0.94 No

5.90 2.51 8.41 91.59 6.42 No
5.90 2.51 8.41 91.59 13.56 No
5.90 2.51 8.41 91.59 11.37 Yes
6.30 2.23 8.52 91.48 0.08 No

6.30 2.23 8.52 91.48 2.06 Yes
6.61 2.16 8.77 91.23 0.42 No

6.61 2.16 8.77 91.23 3.84 No
6.61 2.16 8.77 91.23 4.71 Yes
6.16 2.27 8.43 91.57 2.06
10.24 3.09 13.32 86.68 0.12 Yes
10.13 3.32 13.45 86.55 0.16 Yes
9.86 3.59 13.44 86.56 0.10 Yes
9.93 3.41 13.34 86.66 0.11 Yes
10.18 3.84 14.02 85.98 0.32 Yes
9.77 3.31 13.08 86.92 0.12 Yes
9.98 3.60 13.58 86.42 0.11 Yes
10.14 3.62 13.77 86.23 0.10 No

10.14 3.62 13.77 86.23 1.09 Yes
10.07 3.03 13.09 86.91 0.16 Yes
10.02 3.54 13.56 86.44 0.11 Yes
10.15 3.80 13.94 86.06 0.39 Yes
10.57 3.60 14.16 85.84 0.50 No

10.57 3.60 14.16 85.84 1.13 No
10.57 3.60 14.16 85.84 2.48 Yes
9.78 3.71 13.49 86.51 0.32 No

9.78 3.71 13.49 86.51 1.73 Yes
9.72 3.62 13.34 86.66 0.12 Yes
10.20 3.87 14.07 85.93 0.16 Yes
10.38 3.29 13.67 86.33 0.42 No

10.38 3.29 13.67 86.33 0.56 No
10.38 3.29 13.67 86.33 2.58 Yes
9.83 3.03 12.87 87.13 0.11 Yes
10.89 3.28 14.18 85.82 0.76 No

10.89 3.28 14.18 85.82 0.88 Yes
225
Hz conc. CH. conc. Total flammables Air conc. Energy Success of
('I.) l'!.t conc .. {%) .1%) (mJ) Ignition
10.41 3.49 13.91 86.09 0.12 Yes
10.50 3.90 14.40 85.60 0.48 No

10.50 3.90 14.40 85.60 0.75 Yes
10.14 3.50 13.64 86.36 0.10
13.61 5.03 18.64 81.36 0.11 Yes
13.40 4.73 18.13 81.87 0.10 Yes
13.60 4.52 18.12 81.88 0.08 Yes
13.27 4.71 17.98 82.02 0.08 Yes
12.87 4.37 17.23 82.77 0.18 Yes
12.91 4.33 17.24 82.76 0.11 Yes
13.44 4.22 17.66 82.34 0.10 Yes
13.53 4.60 18.13 81.87 0.08 Yes
13.36 4.64 18.00 82.00 0.08 Yes
12.82 4.29 17.11 82.89 0.11 Yes
13.33 4.76 18.08 81.92 0.17 Yes
13.14 4.30 17.45 82.55 0.51 No

13.14 4.30 17.45 82.55 0.56 No
13.14 4.30 17.45 82.55 1.73 Yes
13.47 4.65 18.12 81.88 0.19 Yes
13.25 4.50 17.76 82.24 0.12 Yes
13.42 4.45 17.88 82.12 0.09 Yes
12.80 4.27 17.07 82.93 0.19 No

12.80 4.27 17.07 82.93 0.24 No
12.80 4.27 17.07 82.93 0.67 Yes
13.69 4.74 18.43 81.57 0.07 Yes
13.51 5.08 18.59 81.41 0.52 No

13.51 5.08 18.59 81.41 0.76 Yes
13.67 4.67 18.34 81.66 0.11 Yes
13.75 4.29 18.04 81.96 0.45 No

13.75 4.29 18.04 81.96 0.46 No
13.75 4.29 18.04 81.96 1.47 Yes
13.34 4.56 17.90 82.10 0.07
16.00 5.84 21.85 78.15 0.18 Yes
16.33 5.90 22.23 77.77 0.17 Yes
15.48 5.78 21.26 78.74 0.14 Yes
15.32 5.08 20.40 79.60 0.42 No

15.32 5.08 20.40 79.60 0.46 No
15.32 5.08 20.40 79.60 2.64 Yes
16.10 5.75 21.86 78.14 0.35 Yes
15.62 5.53 21.15 78.85 0.14 Yes
226
Hz conc. CH"conc. Total flammables Air conc. Energy Success of
.{'1!l (%) cone ... (%) (%) (mi) Ignition
16.19 5.03 21.22 78.78 0.12 Yes
15.58 5.72 21.30 78.70 0.12 Yes
16.24 5.59 21.83 78.17 0.17 Yes
15.44 5.28 20.72 79.28 0.11 Yes
15.81 6.04 21.86 78.14 0.19 Yes
16.22 5.02 21.24 78.76 0.49 No

16.22 5.02 21.24 78.76 0.61 No
16.22 5.02 21.24 78.76 2.42 Yes
15.58 5.88 21.46 78.54 0.31 No

15.58 5.88 21.46 78.54 0.40 No
15.58 5.88 21.46 78.54 2.06 Yes
15.87 5.82 21.69 78.31 0.14 Yes
15.73 5.65 21.38 78.62 0.17 Yes
15.48 5.19 20.67 79.33 0.21 Yes
15.60 5.50 21.09 78.91 0.12 Yes
16.14 5.53 21.67 78.33 0.70 No

16.14 5.53 21.67 78.33 0.58 Yes
15.53 5.86 21.40 78.60 0.38 No

15.53 5.86 21.40 78.60 0.33 No
15.53 5.86 21.40 78.60 0.73 Yes
16.42 5.86 22.28 77.72 0.11 Yes

15.84 5.59 21.43 78.57 0.11
20.64 6.51 27.15 72.85 0.22 No

20.64 6.51 27.15 72.85 0.31 No
20.64 6.51 27.15 72.85 1.24 Yes
20.37 7.23 27.59 72.41 0.17 No

20.37 7.23 27.59 72.41 0.42 No
20.37 7.23 27.59 72.41 1.39 No
20.37 7.23 27.59 72.41 5.89 Yes
20.61 7.15 27.76 72.24 0.22 No

20.61 7.15 27.76 72.24 2.37 No
20.61 7.15 27.76 72.24 5.48 No
20.61 7.15 27.76 72.24 6.20 Yes
19.62 6.59 26.21 73.79 0.28 No

19.62 6.59 26.21 73.79 2.37 No
19.62 6.59 26.21 73.79 5.89 No
19.62 6.59 26.21 73.79 7.20 Yes
20.61 7.02 27.63 72.37 0.36 No

20.61 7.02 27.63 72.37 0.42 No
20.61 7.02 27.63 72.37 1.39 No
20.61 7.02 27.63 72.37 3.52 Yes
20.19 6.41 26.60 73.40 0.33 No

20.19 6.41 26.60 73.40 0.42 No
20.19 6.41 26.60 73.40 1.12 No
20.19 6.41 26.60 73.40 8.93 Yes
20.00 6.96 26.96 73.04 0.42 No

20.00 6.96 26.96 73.04 0.42 No
20.00 6.96 26.96 73.04 1.39 No
20.00 6.96 26.96 73.04 5.60 Yes
227
H2 con~i CH4 conc. Total flammables Air conc. Energy Success of
(% ('10) conc. (%) (%) (mJ) Ignition
20.31 7.10 27.41 72.59 0.30 No
20.31 7.10 27.41 72.59 0.42 No
20.31 7.10 27.41 72.59 1.39 Yes
20.29 7.33 27.63 72.37 0.31 No
20.29 7.33 27.63 72.37 0.42 No
20.29 7.33 27.63 72.37 1.39 No
20.29 7.33 27.63 72.37 11.74 Yes
20.66 6.89 27.55 72.45 0.41 No
20.66 6.89 27.55 72.45 2.81 No
20.66 6.89 27.55 72.45 6.42 No
20.66 6.89 27.55 72.45 10.04 No
21.12 6.64 27.76 72.24 0.30 No
21.12 6.64 27.76 72.24 2.81 No
21.12 6.64 27.76 72.24 6.67 No
21.12 6.64 27.76 72.24 7.55 No
20.20 6.79 27.00 73.00 0.30 No
20.20 6.79 27.00 73.00 1.05 No
20.20 6.79 27.00 73.00 8.79 No
20.20 6.79 27.00 73.00 10.04 Yes
19.58 6.43 26.01 73.99 0.32 No
19.58 6.43 26.01 73.99 0.33 No
19.58 6.43 26.01 73.99 1.27 No
19.58 6.43 26.01 73.99 6.46 Yes
20.18 6.74 26.92 73.08 0.24 No
20.18 6.74 26.92 73.08 2.54 No
20.18 6.74 26.92 73.08 3.34 No
20.18 6.74 26.92 73.08 4.69 No
20.20 6.65 26.85 73.15 0.23 No
20.20 6.65 26.85 73.15 0.42 No
20.20 6.65 26.85 73.15 1.39 No
20.20 6.65 26.85 73.15 11.16 Yes
20.45 7.27 27.72 72.28 0.22 No
20.45 7.27 27.72 72.28 0.42 No
20.45 7.27 27.72 72.28 1.39 No
20.45 7.27 27.72 72.28 3.34 Yes
20.62 6.57 27.19 72.81 0.31 No
20.62 6.57 27.19 72.81 2.54 No
20.62 6.57 27.19 72.81 3.34 No
20.62 6.57 27.19 72.81 11.09 No
20.42 6.65 27.07 72.93 0.76 No

20.42 6.65 27.07 72.93 2.75 No
20.42 6.65 27.07 72.93 9.51 Yes
20.42 6.65 27.07 72.93 0.00 No
20.57 6.83 27.40 72.60 0.33 No
20.57 6.83 27.40 72.60 1.39 No
20.57 6.83 27.40 72.60 3.34 No
20.57 6.83 27.40 72.60 8.71 No
21.08 7.01 28.10 71.90 0.48 No
21.08 7.01 28.10 71.90 0.98 No
21.08 7.01 28.10 71.90 4.58 No
21.08 7.01 28.10 71.90 5.27 No
20.39 6.36 27.22 72.78 1.24
(average) (averaoe) (averaoe) (averaoe) (Lowest Enerov)
228

Minimum Ignition Energy and Ignition Probability For Methane, Hydrogen and Their Mixtures

Uploaded by

Document Informationclick to expand document informationjournalf

Document Informationclick to expand document information

Copyright:

Available Formats

Minimum Ignition Energy and Ignition Probability For Methane, Hydrogen and Their Mixtures

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Minimum Ignition Energy and Ignition Probability For Methane, Hydrogen and Their Mixtures

Uploaded by

Copyright:

Available Formats

Loughborough University

Minimum ignition energy

This item was submitted to Loughborough University's Institutional Repository

• A Doctoral Thesis. Submitted in partial fullment of the requirements for

Metadata Record: https://dspace.lboro.ac.uk/2134/19907

Publisher: c Hemant Mathurkar

Please cite the published version.

... to my daughter ...

A Doctoral Thesis Submitted in partial fulfillment

Department of Chemical Engineering

© Hemant Mathurkar, 2009

I express my gratitude to Professor Geoff Hankinson, Chair of Safety Engineering,

I am grateful to me parents because of whom, I am on this earth and trying to

I would like to thank my organisation, NEERI and specially my Head of Department

I am also grateful to the people of the town of Loughborough and Loughborough

I extend my heartfelt thanks and warmest gratitude to one and all.

1.1 Background ................................................................................................ 1

1.2 Motivation .................................................................................................. 3

1.3 The Naturalhy projec!... .............................................................................. 5

1.4 Likelihood of failure .................................................................................. 11

1.4.1 The UKOPA pipeline fault database .................................................. 12

1.4.2 The EGIG report on gas pipeline incidents ....................................... 14

1.4.3 The US Department of Transport (DOT) data ................................... 15

1.4.4 Low pressure pipeline incidents ....................................................... 18

1.5 Impact of hydrogen .................................................................................. 19

1.5.1 Impact on pipeline failures ................................................................ 19

1.5.2 . Impact on ignition probability ............................................................ 21

1.6 The Thesis structure ................................................................................ 23

2 Literature Review ............................................................................................. 25

2.1 Ignition theory .......................................................................................... 25

2.2 Historical data on ignition of gas releases ................................................ 27

2.2.1 High pressure releases .................................................................... 27

2.2.2 Low pressure releases ..................................................................... 29

2.3 Sources of Ignition ................................................................................... 31

2.3.2 Ignition sources for transmission pipeline release ............................ 34

2.3.3 Ignition sources for Low pressure release ........................................ 40

2.3.4 Analysis of Ignition sources .............................................................. 45

2.4 Ignition model for gas releases ................................................................ 47

2.4.1 Cox, Lees and Ang model ................................................................ 49

2.4.2 PIPESAFE - Ignition probability model for transmission pipeline ...... 50

2.4.3 WS Atkins model .............................................................................. 51

2.4.4 Ignition probability estimate for low pressure releases ...................... 53

2.4.5 Model limitation ................................................................................ 56

2.5 The Minimum Ignition Energy of gases .................................................... 57

2.5.1 MIE determination ............................................................................ 58

2.5.2 MIE and quenching distance ......................... 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

2.5.3 Historical MIE data ........................................................................... 63

2.5.4 Limitations of MIE apparatus ............................................................ 65

2.6 Ignition energy and ignition probability for gases ...................................... 65

3 Design and Operation of the Experimental facility ............................................ 67

3.1 Introduction .............................................................................................. 67

3.2 Conceptual design ................................................................................... 67

3.3 Gas supply, delivery and evacuation system ........................... ,................ 68

3.4 Cylindrical combustion chamber .............................................................. 71

3.5 The Capacitive spark discharge circuit.. ................................................... 73

3.6 Instrumentation ........................................................................................ 76

3.6.1 Pressure Transducers ...................................................................... 76

• A Doctoral Thesis. Submitted in partial fullment of the requirements for