Beruflich Dokumente
Kultur Dokumente
GT2012
June 11-15, 2012, Copenhagen, Denmark
GT2012-69536
ABSTRACT Gas oil (flash point > 61 °C) ignited as a free jet at a
Current European Health and Safety Legislation was working pressure consistent with previous hypotheses.
implemented to limit the chance of a serious explosion However, when the jet impinged on a solid surface then the
occurring in the workplace by highlighting potentially resulting spray could be ignited at considerably lower delivery
explosive atmospheres and ensuring that ignition sources are pressures. Although the impingement process is complex, the
not present in these areas. Though hazardous area data will be discussed in light of contemporary models that
classification for gaseous and dust explosion hazards are well predict initial jet/spray characteristics along with post-
established, the same cannot be said for mists especially for impingement characteristics. This paper presents a first step
high flash point liquids. However, a recent literature review of towards consolidating previous studies and improving future
a range of (some fatal) incidents has shown that mist safety guidelines concerned with the risk posed by the
explosions are more common and the consequences more flammability of accidental releases of pressurised high
severe than previously anticipated. This work is, for example, flashpoint fuels.
applicable to the safe use of fuels and lubricants utilised in the
gas turbine power generation and propulsion industries. NOMENCLATURE
Previous studies of jet breakup regimes and idealised Symbol
flammability studies have indicated that low pressure releases Bst Spalding Transfer Number
(<10 bar) of low volatility fuels may still give rise to cp Specific Heat Capacity
combustion hazards. Impingement of accidental releases onto D Droplet Diameter
surfaces has been shown to exacerbate the potential hazard, or d Nozzle Diameter
broaden the range of hazardous release conditions. However, L Length of Nozzle
although a theoretical case can be made for generating La Laplace Number
flammable environments under moderate release conditions, Re Reynolds Number
very little evidence has been provided to bridge the gap Tst Flame Temperature
between ‘idealised’ studies and full-scale incidents. The aim of We Weber Number
this first programme of work is to start the process of bridging μ Viscosity
this gap, leading to well founded safety guidance. ρ Density
The test programme was conducted in a custom built σ Surface Tension
spray chamber located in the Gas Turbine Research Centre φ Equivalence Ratio
(GTRC) of Cardiff University. The fuel was released at a pre- Subscript
defined range of pressures of industrial relevance at a Air
atmospheric temperature. Igniters were positioned at three d Droplet
downstream locations and the continuous electrical discharge f Fuel
had an energy no greater than 4 mJ. Tests were conducted for l Liquid
‘free sprays’ where the spray was directed along the length of min Minimum
the chamber, and for impinging sprays where the spray was o Orifice
aligned to impinge normal to a flat un-heated surface. st Stoichiometric
RESULTS
Due to the confidential nature of the research programme
the delivery pressures utilised for this study cannot be detailed
at this time consequently the results presented are non-
dimensionalised against a Safe Working Pressure (SWP). The
SWP is defined as the pressure at which there were no
recorded ignition events during the free spray experiments. Figure 3: Video still of ignition of free spray at 1.0 SWP
For the free spray experiments the release pressure was set Figure 4 presents two video stills showing pre and
at an initially high value and repeat experiments conducted at post-ignition of the impinging spray at 1.0 SWP. It is
decreasing delivery pressures until no ignition was observed interesting to note that the spray was ignited at Igniter 1
during the three repeated experiments. Figure 2 presents two corresponding to the closest downstream location (60mm)
video stills showing pre and post-ignition of the free spray at which is in contrast to the 1500 mm distance observed for
2.5 SWP. ignition of the free spray.
DISCUSSION
The first observation is that ignition occurred for releases
where the temperature of the fuel is below the flashpoint,
highlighting the fact that flashpoint cannot be used as a
measure for risk assessment relating to mist explosions.
The interesting observation from Table 1, is that
impingement of the spray results in an ignitable mixture at
delivery pressures half that of pressures previously considered
safe. There are still practical applications where fuel is
transported at this pressure, and consequently still a risk. The
thermo-fluid dynamics associated with this observation will be
analysed and discussed highlighting reasons for the differences Figure 5: Mass-under-size and number-under-size plots for
in ignitability. various SWP.
The size distribution in a spray has a significant The minimum ignition energy (MIE) required to
influence on its combustion characteristics. For example, ignite a droplet has been studied and quantified previously
ACKNOWLEDGMENTS
The authors would like to thank RWE for funding the
study and the efforts of Mr Steven Morris and Mr Terry
Figure 10: Mass-under-size and number-under-size with MIE
Treherne at the GTRC in helping conduct the experiments.
for 0.5 SWP.
Although outside of the scope of this study it worth
REFERENCES
considering the following. Solid surfaces in close proximity to
Bai, C.X., Rusche, H., Gosman, A.D., Modeling of
a potential release source may also, depending upon their size,
Gasoline Spray Impingement. Atomization and Sprays, 2002.
shape and number, result in confinement of the atomised spray
12: p. 1-27.
and acceleration of the resulting flame, therefore increasing
Ballal, D. R. & Lefebvre, A. H. (1978) Ignition and
not only the probability of ignition but also the consequences
flame quenching of quiescent fuel mists. Proc. R. Soc. Lond.,
of that event. The practical implication of this is that the
A364, p277-294.
environment surrounding such systems may have to be given
Bane, S.P.M., Ziegler, J.L, Boettcher, P.A, Coronel,
more detailed consideration in the risk assessment than has
S.A and Shepherd, J.E. (2011) Experimental investigation of
previously been thought necessary.
spark ignition energy in kerosene, hexane, and hydrogen.
Journal of Loss Prevention in the Process Industries. Article in
CONCLUSIONS
press.
This paper presents a study of hazards associated with
Bowen, P.J. and Shirvill, L.C. (1994). Combustion
high-flash point fuels and how the severity of the hazards can
hazards posed by the pressurized atomization of high-
be increased with spray impingement. The main conclusions
flashpoint liquids. Journal of Loss Prevention in Process
are:
Industries. V7 n3, pp 233-241.
• Unobstructed, ‘low-pressure’ liquid fuel releases
Bowen, P.J. (2011) Combustion Hazards Posed by
through plain orifices have been shown to ignite.
Hybrid Fuel Systems, Keynote Address at 5th European
• Sprays were ignited at temperatures significantly
Combustion Meeting, Scientific Programme, pg. 2, Cardiff
(circa 40K) below the flashpoint of the fuel.
28th June – 1st July 2011.
• A safe working pressure was determined
BSI (2009) Explosive atmospheres - Part 10-1:
experimentally for free spray conditions of a high-
Classification of areas - Explosive gas atmospheres. BS EN
flash point fuel.
60079-10-1:2009, British Standards Institution, London, UK.
• Sprays that impinged on a normal surface ignited at Eckhoff, R.K (2006) Differences and similarities of
0.5 the safe working pressure determined for the free gas and dust explosions: A critical evaluation of the European
jet releases highlighting a significant outstanding ‘ATEX’ directives in relation to dusts. Journal of Loss
issue concerning ‘area classification’ Prevention in the Process Industries, V19, pp 553-560.
• Current models in the literature have been used to EI (2005) Area classification code for installations
indicate the likelihood of flammable sprays, and handling flammable fluids. Part 15 of the IP model code of safe
emphasise the importance of both droplet size and practice in the petroleum industry, Energy Institute, London,
concentration on the MIE required. UK.
• Published models from the literature showed that at Kay, P.J., Bowen, P.J., Witlox, H.W.M. (2010) Sub-
the conditions tested the proportion of the spray cooled and flashing liquid jets and droplet dispersion II.
considered flammable at the SWP was shown to be Scaled experiments and derivation of droplet size correlations.
significantly less compared to a spray at 2.5 SWP,
consistent with our obseverations.