Fire Banner

Thursday, July 22, 2010

Combustion Institute British Spring Meeting 2010 - Combustion Phenomena in Fire Science

One-Day Meeting on Combustion Phenomena in Fire Science

Combustion Institute British Section: Spring Meeting 2010

Minutes by Stephen Welch, July 2010
BRE Centre for Fire Safety Engineering, The University of Edinburgh

Seven invited speakers from home and abroad represented research interests spanning experimental studies and modelling of fire phenomena.  The meeting opened with a good context setting talk on “Enclosure fires modelling: where are we and where are we going?”, by Prof Bart Merci of Ghent University.  The capabilities of models are progressively advancing but equally if not more important is the knowledge of the user.  Options to increase knowledge of FSE were discussed and the value of Masters programmes in Fire Safety Engineering emphasised. The complexity of fire phenomena and the strong dependence of fire development on details of the input must be recognised. The need for systematic validation exercises working up from simple cases, and recognising measurement uncertainties, remains vital.  Discussion focussed on the problem of models validated for benchscale scenarios being misused in other large-scale applications.

The current limits of our knowledge became rapidly apparent in the next talk on the Buncefield incident by Dougal Drysdale, emeritus professor at Edinburgh University.  Lavishly illustrated by impressive images of tank fires and explosion aftermath, Dougal highlighted a number of thought-provoking aspects of the incident: the possibility that the vapour had been ignited by the pumps turned on to disperse it, with initially puzzling damage features elucidated as pertaining to the reverse flow in the rarefaction wave, all pointing at the pumphouse, and the environmental impact of the use of remaining stocks of old foam concentrate, which had been banned from further use.  The severity and exact nature of the explosion has still not been satisfactorily resolved via modelling studies that were completed with Cartesian hedgerows. Despite opinions expressed at the time that this incident was unique and could never happen again Dougal noted that there have recently been two more of a similar nature.  Hedgerows may indeed have had a role in providing turbulence generation mechanisms and perhaps we need to consider their removal!

Dove-tailing nicely with Dougal’s conclusions, Dr Savio Vianna of Cambridge University picked up the theme of dealing with complex geometries in accidental explosion modelling.  Peak pressures have been well predicted in a range of applications using a Modified Porosity Distributed Resistance (MPDR) model for approximating the flow effects due to complex obstructions (thus potentially of value for Dougal’s hedgerows?!).  Work is ongoing on addressing further aspects of the combustion modelling and the impact of suppression phenomena via deluge and micromist systems.

With another slick progression Prof Kai Luo of Southampton University then took us deeper into the challenges of modelling fire suppression.  Liquid phase effects tend to invalidate most of our existing modelling tools for diffusion flames and attempting to include them we are immediately confronted by severe computational challenges.  Nevertheless, by adopting an Eulerian-Lagrangian Approach with an LES/DNS framework valuable insights into mechanisms have now been achieved – highlighting the need to supply sufficiently small drops which are able to effectively reach the reaction zones and the fact that the cooling effects are dominant over dilution and direct kinetic impacts.  Thus fine mists with large evaporation enthalpies will tend to be most effective but optimum droplet size is dependent on the nature of the fire flows.

Having exhausted the topic of suppression we returned to fluid dynamics and the particular problems of entrainment of air into thermal spill plumes, studied in great detail by Dr Roger Harrison in his work at the University of Canterbury.  These are very relevant practical problems for design of large public spaces but hitherto the spill plume formulae have been constrained by insufficient empirical knowledge, and the application of advanced numerical models, i.e. CFD, limited by other uncertainties.  It was found that spill plume behaviour and entrainment are dependent on the characteristics of the layer flow below the spill edge.  Roger’s work has also resulted in a range of new and improved simplified design formulae for a variety of spill plume scenarios and new guidance on the use of CFD models for these applications.

Coming back to fundamental fire phenomena, Prof John Griffiths of Leeds University addressed the topic of lagging fires, a common problem in industrial environments when potentially flammable fluids leak from pipe work into the surrounding insulation material. Such fires may have devasting consequences, and are neglected at our peril!  The participating phenomena are highly complex, but John’s experimental, numerical and theoretical investigations have revealed the role of different processes related to the nature of the combustion (gas or liquid phase) and the dependence on the fluid properties in interaction with the heating environment, i.e. the energetic effects of vaporisation and the possibility of fluid and vapour movement and recondensation within the porous media. Thus fuel volatility, overlooked in previous studies with mainly involatile liquids, is a key parameter.

The day concluded with a wide-ranging talk on Forest Fire Research by Prof Domingos Viegas of the University of Coimbra.  We were informed of the fundamental experimental research on fire spread dependencies which have clarified basic sensitivities to effects of wind and slope.  At full-scale level the role of convection is vital.  The mechanisms involved in spot fires have been individually examined and fire tornados have been studied in the lab and at full-scale.  The concept of eruptive fire behaviour was described, and the extreme dangers arising from sudden transitions in fire development illustrated by a number of sobering case studies.  The talk concluded with lessons learned from the Australian fires in Victoria in 2009, which claimed 173 lives and destroyed 4000 km2 within 10 hours. 

All of the talks (downloadable*) raised our awareness of the potentially serious consequences of fire in various arenas and the challenge to the fire community in furthering our understanding and knowledge of the fundamental underpinning fire phenomena.  We add to this our responsibility to educate and inform and clearly we have our work cut out and much to do.  In concluding the meeting the awards committee recognised some of the outstanding work already being done in these areas in conferring the best poster awards to  Dr A Snegirev of Saint-Petersburg State Polytechnic University, for his work on modelling spray fires, and to Jamie Stern-Gottfried et al. of Edinburgh University/Arup, for his studies of non-homogeneous fire environments.

Minutes by Stephen Welch, July 2010
BRE Centre for Fire Safety Engineering, The University of Edinburgh

PROGRAMME of Invited Speakers:


Registration, coffee/tea and poster setup


Ghent University
Enclosure fires modelling

University of Edinburgh
2005 Buncefield oil depot explosions

University of Cambridge
Accidental explosions modelling

Lunch and Poster Session

University of Southampton
Fire suppression modelling

University of Canterbury
Fire plume experiments

Coffee/tea and Poster Session

University of Leeds
Lagging fires: Experimental, numerical and theoretical investigations.

University of Coimbra
Forest fires research

Best Poster Awards


Meeting Co-sponsored by IOP Combustion Physics Group

No comments: