Numerical study of vented hydrogen explosions in a small scale obstructed chamber

There is a growing need to understand and estimate the explosion hazards associated with hydrogen storage and utilisation. This paper presents a comprehensive numerical study on the explosion characteristics of a lean hydrogen-air mixture in a small-scale obstructed vented chamber. The large eddy simulation (LES) technique is employed to study the highly unsteady turbulence-driven explosion when the flame propagates past successive obstructions. A dynamic flame surface density (DFSD) model is applied to the filtered chemical source term in the LES to account for the progressive wrinkling of the deflagrating flame. The driving mechanism of pressure rise and the underlying physics of flame-obstacle interactions are illustrated using the detailed LES results. The paper considers 11 individual flow experimental configurations of various obstacle number, size and location. They are further classified into six groups to investigate the influence of the level of blockage and the separation distance between adjacent obstructions. Critical safety-related parameters including the maximum overpressure and its incidence time are analysed. A comparison with propane is also made to highlight the substantial overpressure and flame acceleration of hydrogen deflagrations. Satisfactory agreements have been obtained between the LES and the experimental data, and this confirms the capability of the developed computational models in capturing essential explosion features and information for the study of vented hydrogen explosions.     

Combustion characteristics of H2/N2 and H2/CO syngas nonpremixed flames

Turbulent nonpremixed H₂/N₂ and H₂/CO syngas flames were simulated using 3D large eddy simulations coupled with a laminar flamelet combustion model. Four different syngas fuel mixtures varying from H₂-rich to CO-rich including N₂ have been modelled. The computations solved the Large Eddy Simulation governing equations on a structured non-uniform Cartesian grid using the finite volume method, where the Smagorinsky eddy viscosity model with the localised dynamic procedure is used to model the sub-grid scale turbulence. Non-premixed combustion has been incorporated using the steady laminar flamelet model. Both instantaneous and time-averaged quantities are analysed and data were also compared to experimental data for one of the four H₂-rich flames. Results show significant differences in both unsteady and steady flame temperature and major combustion products depending on the ratio of H₂/N₂ and H₂/CO in syngas fuel mixture.     

A Review of Research and an Experimental Study on the Pulsation of Buoyant Diffusion Flames and Pool Fires

This paper reviews the past research, experimental techniques and scaling relationships used in the studies of oscillatory buoyant diffusion flames and reports an experimental investigation conducted to determine the pulsating characteristics of such flames. The experimental data were obtained by using three techniques, namely, pressure fluctuation measurements, thermal imaging and high-speed video photography. Present findings are compared with data sets reported in the literature and correlations for pulsation frequency suggested by previous studies are independently verified. Analysis of the experimental data on frequency of pulsations in different burners shows that for a fixed-diameter flame the pulsation frequency is almost independent of fuel flow rate. The equation f=1.68D^−0.5 gives the best approximation for the relationship between pulsating frequency and diameter over a wide range of data. An alternative way of expressing the relationship between the key variables is St=0.52*(1/Fr)^0.505. This proves to be a better way of expressing the relationship since it can include the effect of the fuel flow rate. Slight modifications to this expression allows prediction of flame oscillations under elevated/reduced gravity and isothermal buoyant plumes. This relationship and the observations of the present study confirm the hydrodynamic nature of flame puffing: interplay of buoyancy and fluid motion. © 1996 by John Wiley & Sons, Ltd.     

Dynamic flame surface density modelling of flame deflagration in vented explosion

The propagation of transient, turbulent premixed flames in a vented explosion chamber in the presence of a series of obstacles is numerically investigated by a dynamic formulation for the Flame Surface Density (FSD) with the Large Eddy Simulations (LES) technique. The chemistry is modelled by a one-step overall reaction, which simulates the reaction of a stoichiometric propane-air mixture. The FSD modelling in the reaction rate model is numerically employed with two different sub-grid scale (SGS) models. The first one is based on an empirical correlation of the SGS velocity fluctuations and the second one is based on similarity ideas involved in solving the wrinkled flame front considered as a fractal surface. The numerical predictions are analyzed and compared against an algebraic, simple FSD model together with experimental data. The calculations show that the dynamic FSD models provide superior results as compared with the algebraic FSD model. The comparisons demonstrate the importance of the contributions from the unresolved FSD and provide good agreement with experimental data for the flame structure, overpressure, and burning velocities.     

Measurements and LES calculations of turbulent premixed flame propagation past repeated obstacles

Measurements and Large Eddy Simulations (LES) have been carried out for a turbulent premixed flame propagating past solid obstacles in a laboratory scale combustion chamber. The mixture used is a stoichiometric propane/air mixture, ignited from rest. A wide range of flow configurations are studied. The configurations vary in terms of the number and position of the built-in solid obstructions. The main aim of the present study is two folded. First, to validate a newly developed dynamic flame surface density (DFSD) model over a wide range of flow conditions. Second, to provide repeatable measurements of the flow and combustion in a well-controlled combustion chamber. A total of four groups are derived for qualitative and quantitative comparisons between predicted results and experimental measurements. The concept of groups offers better understanding of the flame–flow interactions and the impact of number and position of the solid baffle plates with respect to the ignition source. Results are presented and discussed for the flame structure, position, speed and accelerations at different times after ignitions. The pressure–time histories are also presented together with the regimes of combustion for all flow configurations during the course of flame propagation.     

Burning syngas in a high swirl burner: Effects of fuel composition

Flame characteristics of swirling non-premixed H₂/CO syngas fuel mixtures have been simulated using large eddy simulation and detailed chemistry. The selected combustor configuration is the TECFLAM burner which has been used for extensive experimental investigations for natural gas combustion. The large eddy simulation (LES) solves the governing equations on a structured Cartesian grid using a finite volume method, with turbulence and combustion modelling based on the localised dynamic Smagorinsky model and the steady laminar flamelet model respectively. The predictions for H₂-rich and CO-rich flames show considerable differences between them for velocity and scalar fields and this demonstrates the effects of fuel variability on the flame characteristics in swirling environment. In general, the higher diffusivity of hydrogen in H₂-rich fuel is largely responsible for forming a much thicker flame with a larger vortex breakdown bubble (VBB) in a swirling flame compare to the H₂-lean but CO-rich syngas flames.     

Modelling of instabilities in turbulent swirling flames

A large eddy simulation-based data analysis procedure is used to explore the instabilities in turbulent non-premixed swirling flames. The selected flames known as SM flames are based on the Sydney swirl burner experimental database. The governing equations for continuity, momentum and mixture fraction are solved on a structured Cartesian grid and the Smagorinsky eddy viscosity model with dynamic procedure is used as the sub-grid scale turbulence model. The thermo-chemical variables are described using the steady laminar flamelet model. The results show that the LES successfully predicts the upstream first recirculation zone generated by the bluff body and the downstream second recirculation zone induced by swirl. Overall, LES comparisons with measurements are in good agreement. Generated power spectra and snapshots demonstrate oscillations of the centre jet and the recirculation zone. Snapshots of flame SM1 showed irregular precession of the centre jet and the power spectrum at a downstream axial location situated between the two recirculation zones showed distinct precession frequency. Mode II instability defined as cyclic expansion and collapse of the recirculation zone is also identified for the flame SM2. The coupling of swirl, chemical reactions and heat release exhibits Mode II instability. The presented simulations demonstrate the efficiency and applicability of the LES technique to swirl flames.     

The impact of variable demand upon the performance of a combined cycle gas turbine (CCGT) power plant

This paper presents experimentally measured data showing the impact of variable demand on a modern 800 MW CCGT plant. The results contrasting the performance of the plant when operating under optimum conditions with those measured when modulating the output to match dispatch instructions is presented and compared. These contrasts include the impact of step changes, continual modulation and both hot and cold starts of the plant. The results indicate the changes in fuel used per MWh, CO₂ emitted per MWh and the NOx emissions under different operating modes. From the subsequent analysis significant increases were recorded in both fuel used and CO₂ emitted when the plant departs from optimum operating conditions. When the plant is requested to cease generating due to over capacity on the system, major increases in the emissions of NOx, when required to restart generation together with large increases in the fuel used and CO₂ emitted per MWh, can be observed.     

SPATIAL DISCRETIZATION ERRORS IN THE HEAT FLUX INTEGRAL OF THE DISCRETE TRANSFER METHOD

The discrete transfer method (DTM) is a widely used algorithm for the computation of radiative heat transfer in enclosures. The truncation error in the heat flux integral associated with the method is the difference between the actual heat flux and its DTM approximation. Estimates were presented by Versteeg et al. [1, 2] of the error associated with the discretization of the hemisphere around irradiated points in enclosures filled with transparent and nonscattering participating media. In this article we quantify the errors due to the spatial discretisation of the enclosure surface and medium conditions. We have studied radiation problems in enclosures with nonhomogeneous absorbing/emitting media with nonuniform surface intensity based on Hsu and Farmer's benchmark case E1 [3]. Our error estimates are found to be in excellent agreement with the actual DTM errors and have been used successfully to predict the convergence rates of the DTM as the control-volume mesh is refined.