Entity

Time filter

Source Type

Saint Denis, France

Chenthil K.K.,Fluidyn Consultancy P Ltd | Anil K.K.R.,Fluidyn Consultancy P Ltd | Tripathi A.,Fluidyn France
Institution of Chemical Engineers Symposium Series | Year: 2015

In general, a fire accident involves various processes such as evaporation of fuel from a pool, formation of jets, laminar and turbulent diffusion, diffusion controlled and kinetics controlled reactions, smoke and soot formation and thermal radiation. The increased computational power enables use of more accurate CFD based models of these processes for large scale fire simulations. This work presents some results of the development and validation of a 3D CFD based model for dispersion of fuel vapours and associated fire hazards. In general, fires occur in a wide range of flow regimes. Unifying the simulation of this wide range of flow under a single combustion model is a challenge. In the present work, a unified model to deal with flames in different regimes such as slow pool fires and fires due to low speed laminar and high speed turbulent jets is implemented by modifying the eddy dissipation concept (EDC) model to include the effect of local laminar or turbulent burning velocities properly. Also, the local flame surface area, which determines local reaction rate, is estimated in a way such that the effect of discretization (mesh size) on the average reaction rate is minimal in order to obtain results which are reasonably grid independent. Spatial distribution of different quantities, such as temperature, mixture fraction, velocity and species concentrations are computed and compared with the experimental measurements available in relevant literature. It is found that the developed model using with an identical set of parameters was able to capture the qualitative behaviour of all the flames with a reasonably good quantitative comparison. © 2015 Amec Foster Wheeler. Source


Lankadasu A.,Fluidyn Consultancy P Ltd | Tripathi A.,Fluidyn France | Saysset S.,GDF SUEZ | Yackow A.,GDF SUEZ | De La Roussiere B.,Entrepose Contracting
Institution of Chemical Engineers Symposium Series | Year: 2015

In September 2012, DNV released the experimental datasets of two research projects conducted to investigate the potential hazards of accidental ruptures of pipelines transporting trans-critical CO2. The transportation is often done via dedicated pipelines at high pressure (.8 MPa), which has high risks of leaking or sudden ruptures. The present paper presents the results of the simulation of three cases among the ones that present the most extreme conditions of the DATASETS. The simulations are conducted in two separate stages: The first stage of the model takes into account the transition flow of supercritical CO2 under the action of large pressure changes through small flow passages such as orifices. It considers steep changes in transport and thermodynamic properties using appropriate equations of state for the wide range of pressures involved and the second stage of the model calculates the large scale dispersion through the ambient air. This model includes consideration of the formation of dry ice due to sudden expansion and subsequent sublimation of dry ice due to entrainment of relatively hot ambient air. It also takes into account the effects of surrounding wind field using an atmospheric boundary layer model. The results have shown that calculated mass flow rate is within the devices accuracy (<11%). At the same time, the concentration and temperatures of CO2 at several distances from the source are well represented by the simulation. © 2015 Amec Foster Wheeler. Source


Kumar P.,DeVry University | Feiz A.-A.,DeVry University | Ngae P.,DeVry University | Barbosa E.,DeVry University | And 4 more authors.
HARMO 2014 - 16th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes, Proceedings | Year: 2014

Dispersion of air pollutants in an urban environment is a challenging problem due to the complexity in interaction of plume and flow field perturbed by the obstacles. The increasing threat of chemical, biological and radiological (CBR) attacks in urban areas has also resulted a significant interest in research on fast identification and detection of these toxic agents. In this study, a computational fluid dynamics (CFD) model is utilized to evaluate the Mock Urban Setting Test (MUST) field tracer experiment that provide a simplified urban-like area. A statistical analysis was performed to compare the concentrations from CFD model simulations with the experimental measurements. Further, the model has been coupled in adjoint mode with a recently proposed inversion technique, based on renormalization theory, for identifying a point source release in an urban like environment of MUST field experiment. The study highlights the detection feasibility of unknown releases in an urban-like environment with a use of more sophisticated model. © Crown Copyright 2014 Dstl. Source


Le Guellec M.,Fluidyn France | Armand P.,CEA DAM Ile-de-France | Vaton G.,Fluidyn France | Souprayen C.,Fluidyn France
HARMO 2011 - Proceedings of the 14th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes | Year: 2011

Major public surface rail station has in the past already been concerned and can be again the place of unexpected event like contaminant release during an attack. In such situation, emergency response efforts need to be optimized. The critical questions are: Where is the emission point? How much material was released? Accurate estimation of the source term is essential to manage the emergency planning and mitigate consequences inside the rail station and in Urban vicinity. One solution is to integrate the observed data at sensors with a predictive model to provide probabilistic estimates of the unknown source term parameters. In such semi-confined volumes with many openings and obstacles, flow streamlines and turbulence fields shows complex patterns in and between halls, ticketing rooms, platforms...etc. High-resolution CFD simulations with external meteorological/street scale flow forcing (and possibly internal ventilation forcing) are required for the prediction of the toxic cloud motion. The inverse problem is then solved by a sampling of predictive simulations guided by statistical comparisons with measured data. This approach uses Bayesian Inference with stochastic sampling based on Markov chain (MCMC). The results of the flow analysis may trigger the location of permanent or mobile detection systems. The event reconstruction when performed indicates the probability distribution functions of the source being at a particular location with a release rate. This paper presents firstly briefly the numerical methodology selected for the source event reconstruction. Then, an application case on the railway station Gare de Lyon in Paris is detailed with a description of the complex flows and some toxic release modeling which have been used for the source term reconstruction testing. Source


Ho Tin Noe I.,SIAAP | Siino F.,SIAAP | Bara C.,SIAAP | Urvoy Y.,SETUDE Ingenieur Conseil | And 4 more authors.
Chemical Engineering Transactions | Year: 2010

The SIAAP1 has implemented a new specific tool called SYPROS based on analysing atmospheric dispersion for forecasting and monitoring the odour impact of wastewater treatment plants on the surrounding urban areas. Meteorological data and Total Reduced Sulphur concentrations monitored on-site are used in order to provide estimations of odour emissions generated by wastewater treatment facilities. The atmospheric dispersion of these odours is then assessed using a specific three-dimensional numerical model which provides information on the real-time odour impact of wastewater treatment plants. The weather forecast, the set-up of the facilities in use and data on wastewater quality are all used to assess the odour emissions generated. This method of evaluation is based on the results of a three-year on-site data analysis project. The odour impact of the wastewater treatment plants for the next two days is computed using forecasted odour emissions and weather forecast data with the help of another specific three-dimensional numerical model which is available for both internal purposes and public use. Copyright ©2010, AIDIC Servizi S.r.l. Source

Discover hidden collaborations