A numerical study on the effects of H2 addition in non-premixed turbulent combustion of C3H8–H2–N2 mixture using a steady flamelet approach

This study has been implemented in two sections. At first, the turbulent jet flame of DLR-B is simulated by combining the k–ε turbulence model and a steady flamelet approach. The DLR-B flame under consideration has been experimentally investigated by Meier et al. who obtained velocity and scalar statistics. The fuel jet composition is 33.2% H₂, 22.1% CH₄ and 44.7% N₂ by volume. The jet exit velocity is 63.2 m/s resulting in a Reynolds number of 22,800. Our focus in the first part is to validate the developed numerical code. Comparison with experiments showed good agreement for temperature and species distribution. At the second part, we exchanged methane with propane in the fuel composition whilst maintaining all other operating conditions unchanged. We investigated the effect of hydrogen concentration on C₃H₈–H₂–N₂ mixtures so that propane mole fraction extent is fixed. The hydrogen volume concentration rose from 33.2% up to 73.2%. The achieved consequences revealed that hydrogen addition produces elongated flame with increased levels of radiative heat flux and CO pollutant emission. The latter behavior might be due to quenching of CO oxidation process in the light of excessive cold air downstream of reaction zone.     

Effect of shock structure on stabilization and blow-off of hydrogen jet flames

Under-expanded hydrogen jet has characteristic shock structure immediately downstream of the nozzle exit. The shock structure depends on the ratio p_EX/p_A, i.e. the ratio of nozzle exit to ambient pressure, and the distributions of velocity and concentration in an under-expanded hydrogen jet depend on characteristics of the shock structure. Therefore, the shock structure should affect the blow-off behaviour of under-expanded hydrogen jet flame. Since this issue has not been investigated in detail, this study aims to close this knowledge gap. The effect of changes in shock structure on lift-off length and blow-off conditions for non-premixed turbulent hydrogen free jet flame has been experimentally investigated. The shock structure was varied by using three types of nozzles: convergent, straight and divergent nozzles. Inlet diameters of nozzles change from 0.31 to 1.04 mm and outlet diameters from 0.34 to 1.7 mm. The static pressure and the ratio of cross-section area at the nozzle inlet to that at the outlet were varying parameters in this study. Hydrogen was horizontally spouted through a nozzle to atmosphere. The maximum static pressure in a nozzle was 13.2 MPa. The experiments revealed that when the hydrogen jet had sequential shock cell structures, which occurred in the range of p_EX/p_A smaller than 2.45, a higher mass flow rate of hydrogen was needed for the stabilization of a jet flame than that for p_EX/p_A larger than 2.45 and that when closed to the ideal expansion (p_EX/p_A = 1), the mass flow rate for stable flame became maximum. In addition, it was observed that the lift-off length of stable flames followed with sequential shock cell structures were almost the same when the minimum cross-section area of used nozzles was constant. However, when hydrogen jet had a shock structure with single Mach disk, the lift-off lengths and the minimum mass flow rate required for the stable jet flame were decreasing with the decrease of the cross-sectional area ratio of the nozzle exit to inlet.     

Stability characteristics of non-premixed turbulent jet flames of hydrogen and syngas blends with coaxial air

The stability characteristics of attached hydrogen (H₂) and syngas (H₂/CO) turbulent jet flames with coaxial air were studied experimentally. The flame stability was investigated by varying the fuel and air stream velocities. Effects of the coaxial nozzle diameter, fuel nozzle lip thickness and syngas fuel composition are addressed in detail. The detachment stability limit of the syngas single jet flame was found to decrease with increasing amount of carbon monoxide in the fuel. For jet flames with coaxial air, the critical coaxial air velocity leading to flame detachment first increases with increasing fuel jet velocity and subsequently decreases. This non-monotonic trend appears for all syngas composition herein investigated (50/50 → 100/0% H₂/CO). OH^∗ chemiluminescence imaging was performed to qualitatively identify the mechanisms responsible for the flame detachment. For all fuel compositions, local extinction close to the burner rim is observed at lower fuel velocities (ascending stability limit), while local flame extinction downstream of the burner rim is observed at higher fuel velocities (descending stability limit). Extrema of the non-monotonic trends appear to be identical when the nozzle fuel velocity is normalized by the critical fuel velocity obtained for the single jet cases.     

Flame stabilization in a lifted non-premixed turbulent hydrogen jet with coaxial air

To understand hydrogen jet liftoff height, the stabilization mechanism of turbulent lifted jet flames under non-premixed conditions was studied. The objectives were to determine flame stability mechanisms, to analyze flame structure, and to characterize the lifted jet at the flame stabilization point. Hydrogen flow velocity varied from 100 to 300 m/s. Coaxial air velocity was regulated from 12 to 20 m/s. Simultaneous velocity field and reaction zone measurements used, PIV/OH PLIF techniques with Nd:YAG lasers and CCD/ICCD cameras. Liftoff height decreased with increased fuel velocity. The flame stabilized in a lower velocity region next to the faster fuel jet due to the mixing effects of the coaxial air flow. The non-premixed turbulent lifted hydrogen jet flames had two types of flame structure for both thin and thick flame base. Lifted flame stabilization was related to local principal strain rate and turbulent intensity, assuming that combustion occurs where local flow velocity and turbulent flame propagation velocity are balanced.     

Analysis of transient supersonic hydrogen release,dispersion and combustion

A hydrogen leak from a facility, which uses highly compressed hydrogen gas (714 bar, 800 K) during operation was studied. The investigated scenario involves supersonic hydrogen release from a 10 cm² leak of the pressurized reservoir, turbulent hydrogen dispersion in the facility room, followed by an accidental ignition and burn-out of the resulting H₂-air cloud. The objective is to investigate the maximum possible flame velocity and overpressure in the facility room in case of a worst-case ignition. The pressure loads are needed for the structural analysis of the building wall response. The first two phases, namely unsteady supersonic release and subsequent turbulent hydrogen dispersion are simulated with GASFLOW-MPI. This is a well validated parallel, all-speed CFD code which solves the compressible Navier-Stokes equations and can model a broad range of flow Mach numbers. Details of the shock structures are resolved for the under-expanded supersonic jet and the sonic-subsonic transition in the release. The turbulent dispersion phase is simulated by LES. The evolution of the highly transient burnable H₂-air mixture in the room in terms of burnable mass, volume, and average H₂-concentration is evaluated with special sub-routines. For five different points in time the maximum turbulent flame speed and resulting overpressures are computed, using four published turbulent burning velocity correlations. The largest turbulent flame speed and overpressure is predicted for an early ignition event resulting in 35–71 m/s, and 0.13–0.27 bar, respectively.     

Swirl effects on harmonically excited,premixed flame kinematics

This paper describes the response of a swirling premixed flame with constant burning velocity to non-axisymmetric harmonic excitation. This work extends prior studies of axisymmetric forcing, which have shown that wrinkles are excited on the flame that propagate downstream along the mean flame surface at a speed given by U_o cos ψ, where U_o is the mean flow velocity and ψ is the flame angle. The swirl component in the flow field introduces an azimuthal transport mechanism for disturbances on the flame. As such, the flame response at any given position is a superposition of flame wrinkles excited at earlier times, upstream axial locations, and different azimuthal positions. These swirl transport effects do not arise in problems where axisymmetric flames are subjected to axisymmetric excitation, but enter quite prominently in the presence of non-axisymmetries, such as when the flame is subjected to transverse excitation. The solution characteristics are strongly dependent upon the ratio of angular rotation rate to excitation frequency, denoted by σ = Ω/ω, which describes the fraction of azimuthal rotation a disturbance makes in one acoustic period. When σ ? 1 and σ ? 1, the axial wavelength of flame wrinkles scales with the convective wavelength, λ_c, but becomes much longer for σ ～ O(1). The spatial variation in phase of flame wrinkling is also strongly dependent upon σ. Regardless of swirl number, flame wrinkles propagate in helical spirals along the solution characteristics at a phase speed equal to the local tangential velocity. The axial phase characteristics of flame wrinkling at a fixed azimuthal location, as would be measured by laser sheet imaging, are much more complex. For σ < 1, the wrinkles exhibit the familiar negative roll-off character for the phase with axial downstream distance, indicative of an axially convecting disturbance. The slope of this phase roll-off decreases with increasing σ, however, and becomes zero at σ = 1 for a compact flame. For σ > 1, the wrinkles actually have a positive roll-off character for the phase with axial downstream distance, indicating a flame wrinkle with a negative trace velocity, but whose actual propagation velocity is positive. Finally, these results show that while the flame response to transverse acoustic excitation is quite strong locally, its spatially integrated effect is much smaller for acoustically compact flames. This suggests that the dominant mechanism through which the flame responds globally to transverse excitation is the induced vortical and longitudinal acoustic fluctuations.     

Effects of density ratio and differential diffusion on flame accelerative propagation of H2/O2/N2 mixtures

The effects of density ratio and differential diffusion on premixed flame propagation of H₂/O₂/N₂ mixtures are investigated by constant volume combustion chamber. The density ratio and differential diffusion are controlled independently by adjusting the O₂/N₂ ratio and equivalence ratio. Results show that the density ratio has no effect on turbulent burning velocity while the differential diffusion has a promotion effect on turbulent burning velocity. The onsets of laminar flame acceleration are promoted by both density ratio and differential ratio. The turbulent flames perform a continuous acceleration propagation and the dependence between flame propagation speed and flame radius can be characterized as (dR/dt)/(σ·S_L) ～ R^0.33～0.37, which is lower than the 1/2 power law. The acceleration parameters of laminar flames and turbulent flames (u^’/S_L = 1) are around 0.17 and 0.36 respectively, and both of them are not affected by density ratio and differential diffusion. The empirical formula m = 0.19·(u^’/S_L)^0.4+0.17 is concluded to quantitatively describe the accelerative characteristics of laminar and turbulent flames. The current study indicates that the acceleration of laminar flames is mainly induced by flame intrinsic instability, and the latter can affect the acceleration onset but not affect the fractal excess. The acceleration of turbulent flames is dominated by turbulent stretch, while the effects of density ratio and differential diffusion can be ignored.     

Investigation of small-scale unintended releases of hydrogen: Buoyancy effects

Measurements were performed in small-scale hydrogen leaks to characterize the dimensional properties and flow characteristics of the resulting ignitable hydrogen cloud. The data are intended to provide a technological basis for determining hazardous length scales associated with the formation of ignitable mixtures due to unintended releases. In contrast to a previous study where momentum-dominated releases were considered, the present study focuses on smaller-scale releases at lower flow rates where buoyancy becomes important. A turbulent jet flow is selected as representative of releases in which the leak geometry is circular. Laser-based Rayleigh scattering is used to characterize the hydrogen concentration field downstream of the leak. Particle Image Velocimetry (PIV) is also used to characterize the flow velocity. Time-averaged mean and fluctuating hydrogen concentration statistics are presented and compared with results in momentum-dominated flows to elucidate the effects of buoyancy on the H₂ dispersion process. Over the range of Froude numbers investigated (Fr = 268, 152 and 99), increasing effects of buoyancy are seen as the Fr is reduced and at downstream locations where the influence of buoyancy increases relative to jet momentum. The primary effect of buoyancy is to increase the centerline decay rate of the time-averaged H₂ mass fraction relative to momentum-dominated flows. Acceleration due to buoyancy also results in a slower decay of the time-averaged axial velocity component along the centerline. Radial profiles of the time-averaged H₂ mass fraction also collapse onto the same curves as results in momentum-dominated flows when plotted against the same similarity/scaling variables. While buoyancy is found to have a negligible effect on centerline velocity fluctuations, the maximum H₂ mass fraction fluctuation intensity increases by 70 percent in the buoyant regime and the peak value shifts from the mixing region to the jet centerline. The database presented should provide a good test for the validation of CFD models being developed to predict unintended hydrogen releases under conditions where buoyancy is important.     

Experimental study on the premixed syngas-air explosion in duct with both ends open

Premixed flame of stoichiometric syngas-air mixture with various hydrogen volume fractions, 10% ≤ X (H₂) ≤ 90%, propagating in a duct with both ends open is experimentally investigated in this study. Two representative ignition locations, i.e., Ig-1, locating at the center of the duct, and Ig-2, locating at the right open end, are considered. Results show that the tulip flame is first attained in the duct with both ends open at 10% ≤ X (H₂) ≤ 50% as the flame is ignited at Ig-1. However, the flame maintains the convex shape with the cellular structure on the flame surface as the flame is ignited at Ig-2. The cellular structure results from Darrieus-Landau instability, but the Darrieus-Landau instability cannot invert the convex flame front. The flame tip and pressure dynamics have been examined. When the flame is ignited at Ig-1, the flame oscillates violently, and the overpressure profiles oscillate as a Helmholtz-type. When the flame is ignited at Ig-2, the left flame front propagates in an atmospheric pressure with a nearly constant speed. The prominent flame acceleration and oscillation are not observed at Ig-2 because of lacking flame acoustic interaction. What's more, the characteristic time of flame propagation has been compared. The time t_w is shorter while the time t_p is longer than the calculated value, and the time t_e has been delayed by both open ends. The flame propagation process is moderated as the flame propagates in the duct with both ends open.     

Hydrogen release from a high pressure gaseous hydrogen reservoir in case of a small leak

The dynamic blow-down process of a high pressure gaseous hydrogen (GH₂) reservoir in case of a small leak is a complex process involving a chain of distinct flow regimes and gas states. This paper presents models to predict the hydrogen concentration and velocity field in the vicinity of a postulated small leak. An isentropic expansion model with a real gas equation of state for normal hydrogen is used to calculate the time dependent gas state in the reservoir and at the leak. The subsequent gas expansion to 0.1 MPa is predicted with a zero-dimensional model. The gas conditions after expansion serve as input to a newly developed integral model for a round free turbulent H₂-jet into ambient air. Predictions are made for the blow-down of hydrogen reservoirs with 10, 30 and 100 MPa initial pressure. A normalized hydrogen concentration field in the free jet is presented which allows for a given leak scenario the prediction of the axial and radial range of flammable H₂-air mixtures.     

Study on hydrogen dynamic leakage and flame propagation at normal-temperature and high-pressure

Experiments of two nozzle diameters at three ignition positions under three initial pressure conditions were carried out. The dynamic leakage characteristics and the stagnation parameters of flame propagation under normal temperature and high pressure conditions were studied. Based on van der Waal's equation, a model for predicting stagnation parameters, jet velocity and flow rate of hydrogen leakage was proposed. Compared with the experimental results, it was found that the maximum error occurred when the initial pressure was 200 bar. Theoretical leakage time was 1.66 s, experiment leakage time was 1.84 s, the error was 9.8%. Background-Oriented Schlieren image technology was used to record the flame development and propagation process after ignition. For the same nozzle diameter and ignition location, the higher pressure caused the flame to propagate faster upstream and downstream. For the same initial pressure and ignition position, a flame with a large nozzle diameter propagated faster upstream and downstream. For the same initial pressure and nozzle diameter, the farther the ignition point was, the greater the slope of flame attenuation when propagating upstream. Due to the attenuation of hydrogen concentration and jet velocity, the flame propagation velocity to the downstream decreased linearly with the increase of distance from the ignition location.     

Laser raman measurements of temperature and species concentration in swirling lifted hydrogen jet diffusion flames

Simultaneous spatially and temporally resolved point measurements of temperature, mixture fraction, major species (H₂, H₂O, O₂, N₂), and minor species (OH) concentrations are performed in unswirled (S_g = 0), low swirl (S_g = 0.12), and high swirl (S_g = 0.5) lifted turbulent hydrogen jet diffusion flames into still air. Ultraviolet (UV) Raman scattering and laser-induced predissociative fluorescence (LIPF) techniques are combined to make the multi-parameter measurements using a single KrF excimer laser. Experimental results are compared to the fast chemistry (equilibrium) limit, to the mixing without reaction limit, and to simulations of steady stretched laminar opposed-flow flames. It is found that in the lifted region where the swirling effects are strong, the measured chemical compositions are inconsistent with those calculated from stretched laminar diffusion flames or stretched partially premixed flames. Sub-equilibrium values of temperature, sub-flamelet values of H₂O, and super-flamelet values of OH are found in an intermittent annular turbulent brush of the swirled flame but not in the unswirled flame. Farther downstream of the nozzle exit (x/D ≥ 50), swirl has little effect on the finite-rate chemistry.     

Experimental study of shock wave propagation and its influence on the spontaneous ignition during high-pressure hydrogen release through a tube

An experimental study of shock wave propagation and its influence on the spontaneous ignition during high-pressure hydrogen release through a tube are measured by pressure transducers and light sensors. Results show that the pressure behind a shock wave first increases, and subsequently remains near constant value with an increase of the propagation distance. That is, a certain propagation distance is required to form a stable shock wave in the tube. In the front of the tube, the minimum value of pressure behind the shock wave (P_shock) required for spontaneous ignition decreases with the increase in axial distance to the diaphragm. However, the minimum P_shock remains nearly a constant value in the rear part of the tube. Moreover, the critical values of shock Mach number (M_S) for spontaneous ignition decrease with the increase in tube length. And the ignition delay time decreases with the increase of the M_S. As the ignition kernel grows in size to a flame, it propagates downstream along the tube with velocity greater than the theoretical flow velocity of the hydrogen-air contact surface. The flame propagation velocity relative to tube wall increases with M_S. When the self-sustained flame exits from the tube, a rapid non-premixed turbulent combustion is observed in the chamber. The combustion-wave overpressure increases with the increase of the M_S.     

Under-expansion jet flame propagation characteristics of premixed H2/air in explosion venting

In premixed H₂/air explosion venting, an under-expansion jet may be caused by the pressure difference between the inside and outside of the explosion vent. Based upon the under-expansion jet, studying the structure of the under-expansion jet flame and the factors influencing its formation is essential to hydrogen safety in explosion venting. This study explored the basic characteristics of the under-expansion jet flame in premixed H₂/air explosion venting, and discussed the formation of two under-expansion structures (Mach disk and diamond shock wave) of such jet flames by conducting a premixed H₂/air explosion venting experiment. The influences of hydrogen fraction, explosion venting diameter, and duct length on the structure of under-expansion jet flames were evaluated. The results showed that after successful explosion venting, the under-expansion jet flame would be generated when the hydrogen fractions were 30–60 vol.%, and as the hydrogen fractions were 30–50 vol.%, the lengths of the venting duct were 30 and 50 cm. The duration of under-expansion jet flame was the longest when the hydrogen fraction was 40 vol.%. With the explosion venting diameter and hydrogen fraction increased, the spacing between under-expansion jet flame structures (S) increased. However, an increase in duct length led to the attenuation of the S. During the explosion venting, the under-expansion jet caused a pressure imbalance near the explosion vent and high-intensity convection forms on both sides of a jet, which can generate two or more explosions. Therefore, understanding the basic characteristics of under-expansion jet flame can aid the effective development of measures to prevent, mitigate, and protect against premixed H₂/air explosions.     

Propagation and stability characteristics of laminar lifted diffusion flame base

Numerical calculations of the flame propagation speed and the Damköhler number (Da) at laminar lifted flame base were carried out. The results are intended for further understanding the propagation and the Damköhler mechanisms for flame stabilization, with the former based on a tribrachial flame propagating against the local flow velocity and the latter based on the competition between the reaction time and local residence time of the peak reaction zone. Propane fuel without and with dilution (40% helium and argon, by volume) was used, while the reaction scheme adopted was the one-step irreversible Arrhenius kinetics (see Li et al., Combust. Flame 157 (2010) 1484–1495) which proved successful in predicting the flame lift-off height and effects of thermal expansion and multi-component diffusion. The results reported in this paper show that the flame base propagation speed is up to approximately four times of the one-dimensional stoichiometric flame speed of the fuels used, depending on where the propagation front is defined. These results are compared with previously published experimental and theoretical results from laminar and turbulent diffusion flames. It is found that the flame base propagation speed (V_p) increases in the downstream direction as a result of increasing jet velocity (V_o) under most flame conditions, providing a stabilizing mechanism. However, there exist conditions where V_p decreases while the flame stabilizes. The flame base Damköhler number (Da) always increases as the flame liftoff height increases (resulting from increasing jet velocity). Da is here defined as ratio of peak reaction rate of the reaction kernel (RR) to the flame stretch rate (k) determined at the intersection of the reaction kernel (approximately coinciding with the 2000 K isotherm) and the stoichiometric contour. The value of Da appears to be of the order of 10^?3 for the three fuels studied, and the increasing trend of Da with the lift-off height also helps to explain the flame stabilization.     

Flame dynamics analysis of highly hydrogen-enrichment premixed turbulent combustion

The highly hydrogen blended turbulent natural gas flames were stabilized on a nozzle-type Bunsen burner and measured with laser diagnostic technique. Flame topology characteristics and turbulent burning velocities for the lean turbulent combustion and uniform laminar flame speed of S_L ≈ 40 cm/s were investigated and compared. Hydrogen effect of high diffusivity on combustion properties was analyzed. The local flame structure parameters were obtained and analyzed. Results show that finer wrinkled structure is not only induced by increasing turbulence intensity u’/S_L, but also there is a significant enhancement due to the increasing hydrogen ratio. At large turbulence intensities for lean combustion, more elongated flame folds are formed and small scale structures are generated inducing flame pockets detaching from the main flame, which may largely due to the strong thermo-diffusive effect. However, when fixing S_L ≈ 40 cm/s, the flame front shows cusp structure with large negative curvature at high hydrogen ratio when u’/S_L is low, which mainly result from Darrieus-Landau instability in influencing the flame-turbulence interaction. Moreover, hydrogen addition apparently enhances turbulent burning velocity and the enhancement is more evident for higher intensities. S_T/S_L seems to follow the power law relation for lean flames while showing a quadratic relation for flame of S_L ≈ 40 cm/s. The PDF profile widens encompassing a larger range with increasing hydrogen ratio, indicating that the scale of wrinkled structure is getting smaller. This can be further verified by the profile of local radius of curvature. Hydrogen has an evident effect in enhancing flame surface density which may connect to turbulent burning velocity. And a slightly decreasing trend is found when ZH2 is beyond 0.6 at high u′/S_L.     

The role of low temperature fuel chemistry on turbulent flame propagation

A new high temperature, high Reynolds number, Reactor Assisted Turbulent Slot (RATS) burner has been developed to investigate turbulent flame regimes and burning rates for large hydrocarbon transportation fuels, which exhibit strong low temperature chemistry behavior. The turbulent flow characteristics are quantified using hot wire anemometry. The turbulent flame structures and burning velocities of n-heptane/air mixtures are measured by using planar laser induced fluorescence of OH and CH₂O with reactant temperatures spanning from 400 K to 700 K. It is found for the first time that for n-heptane/air mixtures there are four unique turbulent flame regimes, a conventional chemically-frozen-flow regime, a low-temperature-ignition regime, a transitional regime between the low- to high-temperature-ignition regimes, and a high-temperature-ignition regime, depending on the initial reactant temperature and heated flow residence time prior to the flame. The turbulent burning velocities have been measured for the first two regimes, chemically-frozen-flow and low-temperature-ignition regimes, in order to quantitatively address the role of low temperature ignition on the turbulent burning velocity. In the latter case, large amount of CH₂O formation has been observed in the pre-flame zone, signaling a significant change in the reactant composition and chemistry. At a given reactant temperature and turbulent intensity, the normalized turbulent burning velocities can be varied depending on the extent of low temperature fuel oxidation by varying the heated flow residence time and reactant temperature. The present results suggest that contrary to the previous studies, the turbulent flame regimes and burning velocities for fuels with low temperature chemistry may not be uniquely defined at elevated temperatures.     

Turbulent burning velocity of stoichiometric syngas flames with different hydrogen volumetric fractions upon constant-volume method with multi-zone model

Syngas has been widely concerned and tested in various thermo-power devices as one promising alternative fuel. However, little is known about the turbulent combustion characteristics, especially on outwardly propagating turbulent syngas/air premixed flames. In this paper, the outwardly propagating turbulent syngas/air premixed flames were experimentally investigated in a constant-volume fan-stirred vessel. Tests were conducted on stoichiometric syngas with different hydrogen volumetric fractions (X_H2, 10%–90%) in the ambience with different initial turbulence intensity (u'_rms, 0.100 m/s~1.309 m/s). Turbulent burning velocity was taken as the major topic to be studied upon the multi-zone model in constant-volume propagating flame method. The influences of initial turbulent intensity and hydrogen volumetric fraction on the turbulent flame speed were analysed and discussed. An explicit correlation of turbulent flame speed was obtained from the experimental results.     

A parametric study of autoigniting hydrogen jets under compression-ignition engine conditions

This study examines the flame evolution of autoigniting H₂ jets with high-speed schlieren and OH1 chemiluminescence optical methods in a constant-volume combustion chamber over a wide range of simulated compression-ignition engine conditions. Parametric variations include the injector nozzle orifice diameter (0.31–0.83 mm), injection reservoir pressure (100–200 bar), ambient temperature (1000–1140 K), density (12.5–24 kg/m³) and O₂ concentration (10–21 vol.%). The jet ignition delay was found to be highly sensitive to changes in ambient temperature while all other parameter variations resulted in minor ignition delay changes. Optical imaging reveals that in most cases, the reaction front of the H₂ jet initiates from a localised kernel, before engulfing the entire jet volume downstream and recessing towards the nozzle. The flames attach to the nozzle, except at the lowest ambient oxygen condition of 10 vol.% O₂ for which a lifted flame is observed. The H₂ diffusion flame length shows a dependence on both the mass flow rate and the level of O₂ entrainment that follows the same correlations as previously established for atmospheric H₂ jet flames.