Experimental and numerical study of NOx formation in a domestic H2/air coaxial burner at low Reynolds number
Introduction
Hydrogen is a potential alternative to fossil fuels in transportation, industrial, commercial and residential sectors [[1], [2], [3]]. Since hydrogen can be generated through different renewable energy sources [[4], [5], [6]], the so-called hydrogen economy could be an integral part of a renewable energy system, allowing the substitution of pipeline natural gas with renewable hydrogen to reduce carbon emissions as well as the dependence on fossil fuels [2,7,8]. The end use of this renewable hydrogen has been a research topic during the last decades. For instance, different studies focused on the use of generated renewable hydrogen as fuel in combustion devices such as internal combustion engines or burners [[9], [10], [11], [12]]. An extensive research literature on hydrogen flames can be found, since its combustion behaviour presents several particularities compared to other conventional gaseous fuels, due to its physical properties. Among others, hydrogen presents a wide flammability limit in air (4–75 vol), low ignition energy in air (0.019 mJ), low density (0.0899 kg/m3) and high adiabatic flame temperature (2380 K) [[13], [14], [15], [16]]. Hence, the variety of combustion devices in different sectors still requires an in-depth research work to re-adapt or re-design them so as to ensure safe and efficient operating conditions when using hydrogen as inlet fuel. In this context, the present study focuses on the use of gaseous hydrogen as fuel in domestic gas boiler burners. As discussed by de Vries et al. [17], while fundamental changes in combustion phenomena and exhaustive analysis of individual combustion devices has been widely studied, there is no clear conclusion concerning the maximum fraction of hydrogen that maintains the performance of installed domestic combustion appliances without adverse consequences.
The safety risk when burning hydrogen is closely related to the flame flashback phenomena due to the high flame front velocities in premixed and partially premixed flames, mainly caused by an imbalance in the local flame and flow speeds, when the former exceeds the latter [[18], [19], [20], [21]]. Recently, Zhao et al. [22] studied the influence of hydrogen addition to pipeline natural gas on the combustion performance of a partially premixed cooktop burner. They concluded that the burner could operate safely and efficiently with a maximum hydrogen concentration of 20. In addition, the measured NOx emissions were over 100 ppm, without meaningful variations when adding hydrogen. Similarly, Choudhury et al. [23] studied the admissible hydrogen percentage in natural gas in low-NOx and conventional water heaters, working with partially premixed and lean premixed conditions. They analyzed characteristics such as flashback, ignition delay, flame structure and emissions. The most interesting conclusion was that the maximum hydrogen tolerance in both low-NOx and conventional water heater was below 10 by volume.
On the other hand, many combustors operate in the non-premixed mode since they ensure safer operating conditions, avoiding flashback and explosion risks [15,24,25]. However, in hydrogen flames around atmospheric pressure as in the present study, where NOx formation can be correctly described by the well known Zeldovich mechanism or thermal mechanism [26,27], the high adiabatic flame temperature of hydrogen diffusion flames (peak temperatures above 2300 K in this study) can yield to high levels of NOx [[28], [29], [30]]. Therefore, the main technological challenge in the design of hydrogen burners is to maintain both a high hydrogen fuel content and acceptable levels of NOx emissions. This clearly requires new or modified existing combustion equipments for hydrogen. Non-premixed combustion can be considered to be the best option to design domestic hydrogen burners since they completely eliminate the flashback phenomena, and voluminous research has been published on turbulent non-premixed hydrogen burners. However, laminar non-premixed hydrogen burners have been hardly studied in the literature. We summarize here some relevant studies.
Igawa et al. [31] examined the effect of splitting hydrogen diffusion flames on NOx emissions for stove gas burners. The experimental setup had no confinement or combustion chamber for the flames, and hydrogen was mixed with the ambient air through diffusion. They recorded measurements at different nozzle diameters for individual and multi-flame burners (0.6, 1.2 and 2.6 mm) and at different nozzle intervals for multi-flame measurements. They also experimentally concluded that splitting the flame was an effective method to decrease NOx emissions in hydrogen diffusion flames, since they measured lower NOx emissions when dividing a single flame burner into three flames for the same total input power.
Chen and Driscoll [33] experimentally studied the effects of coaxial air and fuel jet Reynolds number on NOx formation in non-premixed jet flames. They aimed at deriving general scaling laws for NOx production (in methane/air and hydrogen/air flames, both with and without coaxial air, and both in laminar and turbulent flames). The results showed that coaxial air reduced NOx formation in hydrogen flames and had a negligible effect in methane flames at Re = 5000. In addition, the results for simple hydrogen jet diffusion flames without coaxial air showed a reduction of NOx levels when reducing the hydrogen inlet diameter from 0.37 to 0.16 cm in both laminar and turbulent regimes. Their attempt to derive general scaling laws is interesting in order to higlight what we will consider here as the three main physical aspects that play a key role, and may compete, in NOx production:
- 1.
Following the Eulerian approach of Peters et al. [34], they related the production of NOx to a reaction volume (proportional to , where is the flame length and δ the flame thickness);
- 2.
From a Lagrangian point of view, they pointed out how the relevant dimension is a residence time (within );
- 3.
They also indirectly considered the chemical time of NOx production (i.e. the inverse of a reaction rate ).
Keeping this in mind, the following references are interesting since they give some insight on these factors in laminar diffusion flames.
Zhao et al. [35] studied the OH formation in a hydrogen diffusion co-flow burner through chemiluminescence images and CFD simulations. The flow conditions were fixed with and without co-flow air with a central hydrogen jet of D = 15 mm and low thermal powers (0.2 kW). The numerical results showed a flame height of 20 mm and were in good agreement with the experimental measurements. The numerical results showed slightly longer flames for the case with co-flow air as well as higher maximum temperature. This could be attributed to the higher inlet thermal power used in the flame with co-flow air.
Li et al. [36] experimentally and numerically investigated the combustion and heat release characteristics of a hydrogen laminar diffusion flame in a non-premixed micro-jet burner (1 mm diameter for H2 inlet and 3.16 mm for air) together with OH distribution measurements at various inlet powers (from 0.1 kW to 1.0 kW) and equivalence ratios. The flame structure was obtained by flame image detection and spectroscopic systems. However, NOx emissions were not measured in this work. Laminar flames were studied varying the inlet hydrogen Reynolds number between 14 and 144, while the Reynolds number of the inlet air flows varied between 2 and 52. Their experimental and numerical results showed longer flames for higher inlet powers, whereas different equivalence ratios would lead to similar flame volumes at a given inlet power.
Khan and Raghavan [37] investigated non-premixed laminar jet flames using carbon monoxide-hydrogen mixtures. Varying the inlet composition, they carried out a numerical and experimental analysis of the flame length, thermal plume width, temperature and flame luminosity. They concluded that the flame became shorter as hydrogen content was increased and that thermal plume width got wider. In addition, when adding hydrogen to the inlet mixture, the maximum temperature isotherm moved closer to the burner exit (moving the reaction volume to a zone of shorter residence times ), but also showing an increasing trend in the maximum temperature (and therefore possibly higher ).
In this paper, we present the results of experiments in a coaxial H2/air burner at low Reynolds number in order to study possible flame splitting strategies for NOx reduction in domestic boiler burners. The different flames are modelled numerically in order to further analyze the results observed experimentally and better understand the different trends in thermal NOx concentrations depending on inlet power and equivalence ratio. Following the ideas suggested by Peters et al. [34], the numerical results are used to extract the thermal NOx reaction volume and a weighted average indicating the competition between residence time and chemical reaction time within . In the light of this discussion of the measured NOx trends, different strategies are discussed for the design of low NOx hydrogen diffusion burners. Finally, the effect of coaxial air on NOx reduction compared to a simple jet flame is illustrated in the appendix.
Section snippets
Single-burner experimental configuration
The flame under study was defined as a single flame from an hypothetical array burner, as illustrated in Fig. 1. The conceptual design of such burner is based on the patent developed by IK4-Ikerlan [38]. Future steps in the design methodology would involve the geometry optimization of the inlet jets, the interaction between several flames and the design optimization for burner integration in real boiler operating conditions.
It can be noted that air is coaxially introduced to guarantee hydrogen
NOx emission measurements
The main components of the experimental setup are shown schematically in Fig. 5. Hydrogen is supplied from a high pressure steel container and compressed air is taken from the compressed air line. The flow rates are controlled by mass flow controllers, and can be varied between 0 and 10 L/min for hydrogen and 0–20 L/min for air. Those mass flow controllers are connected to a Brooks 0152 Microprocessor control&read out unit® in order to control the fixed mass flow values manually. Additionally,
Modelling
The steady two-dimensional axisymmetric flow field was obtained by solving the exact Navier-Stokes equations. Moreover, species transport equations including chemical reaction source terms and the sensible enthalpy transport equation including a radiative heat loss source term were solved. We give here some details on the modelling of these source terms, together with the treatment of heat transfer within the insulation and through the outer boundary. The post-processing of the thermal NOx is
General observations
The power range under study was from 0.2 to 1.0 kW and consequently the inlet Reynolds numbers increased with increasing inlet mass flows rates. No steady solution could be obtained numerically when increasing the inlet power above 0.6 kW, suggesting that laminar-turbulent transition was reached. This non-laminar regime is observed experimentally in the unconfined flames at 0.8 kW as shown in Fig. 8 for the case . Therefore, the computational study was limited to the range 0.2–0.6 kW
Discussion
Different trends have been observed in the results. NOx concentrations increase with inlet power in the laminar regime, while the opposite trend is observed experimentally when flames are not laminar. On the other hand, NOx concentrations are higher at lower equivalence ratios, even out of the laminar regime. Since the numerical results correctly capture the main features of the flames considered, we will use them in order to interpret the trends in the laminar cases.
Some qualitative ideas on
Conclusions
Thermal NOx formation in laminar H2/air coaxial diffusion flames was analyzed through experimental measurements and CFD simulations. The influence of inlet power and equivalence ratio was experimentally and numerically analyzed varying the power from 0.2 to 1.0 kW at φ = 0.77, 0.83 and 0.91. Numerical results were compared to axial temperature measurements for the unconfined flame, and to NOx concentration measurements for the confined flame, showing satisfactory agreement in both cases. Using
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors are grateful to the Basque Government for funding this research through projects IT781-13 and IT1314-19. Part of this work is also supported at Ciemat by the project #PID2019-108592RB- C42/AEI/10.13039/50110 0 011033.
References (56)
Chapter 1 - hydrogen energy: sustainable and perennial
- et al.
A comparative overview of hydrogen production processes
Renew Sustain Energy Rev
(2017) - et al.
Hydrogen as an energy vector
Renew Sustain Energy Rev
(2020) - et al.
Performance enhancement of an integrated system with solar flat plate collector for hydrogen production using waste heat recovery
Energy
(2019) - et al.
Investigating the effect of hydrogen injection on natural gas thermo-physical properties with various compositions
Energy
(2019) - et al.
Introducing Power-to-H3: combining renewable electricity with heat, water and hydrogen production and storage in a neighbourhood
Appl Energy
(2020) - et al.
Experimental study the effects of various compression ratios and spark timing on performance and emission of a lean-burn heavy-duty spark ignition engine fueled with methane gas and hydrogen blends
Energy
(2019) - et al.
Seasonal energy and environmental characterization of a micro gas turbine fueled with H2NG blends
Energy
(2020) - et al.
Effects of flow rate and fuel/air ratio on propagation behaviors of diffusion H2/air flames in a micro-combustor
Energy
(2019) - et al.
Controlled autoignition of hydrogen in a direct-injection optical engine
Combust Flame
(2012)
Recent advances in understanding of flammability characteristics of hydrogen
Prog Energy Combust Sci
The impact of natural gas/hydrogen mixtures on the performance of end-use equipment: interchangeability analysis for domestic appliances
Appl Energy
Flashback, burning velocities and hydrogen admixture: domestic appliance approval, gas regulation and appliance development
Appl Energy
Experimental and computational study on stability characteristics of hydrogen-enriched oxy-methane premixed flames
Appl Energy
Chapter 6 - flashback due to combustion instabilities
A study of NOx formation in hydrogen flames
Int J Hydrogen Energy
Generation characteristics of thermal NOx in a double-swirler annular combustor under various inlet conditions
Energy
Numerical study of a laminar hydrogen diffusion flame based on the non-premixed finite-rate chemistry model; thermal NOx assessment
Int J Hydrogen Energy
Concept of hydrogen fired gas turbine cycle with exhaust gas recirculation: assessment of process performance
Energy
A novel configuration of natural gas diffusion burners to enhance optical, thermal and radiative characteristics of flame and reduce NOx emission
Energy
Nitric oxide levels of jet diffusion flames: effects of coaxial air and other mixing parameters
Structure and similarity of nitric oxide production in turbulent diffusion flames
Experimental and numerical study of OH chemiluminescence in hydrogen diffusion flames
Combust Flame
Hydrogen combustion as a thermal source
Energy Procedia
Structure and reaction zones of hydrogen – carbon-monoxide laminar jet diffusion flames
Int J Hydrogen Energy
NOx emission characteristics of confined jet nonpremixed flames
Proc Combust Inst
Effects of H2O and CO2 diluted oxidizer on the structure and shape of laminar coflow syngas diffusion flames
Combust Flame
Accurate multicomponent Fick diffusion at a lower cost than mixture-averaged approximation: validation in steady and unsteady counterflow flamelets
Combust Flame
Cited by (13)
An assessment of the operating conditions of the micromix combustion principle for low NOx industrial hydrogen burners: Numerical and experimental approach
2024, International Journal of Hydrogen EnergyExperimental and numerical characterization of hydrogen combustion in a reverse-flow micro gas turbine combustor
2024, International Journal of Hydrogen EnergyCFD study of flameless combustion in a real industrial reheating furnace considering different H<inf>2</inf>/NG blends as fuel
2024, International Journal of Hydrogen EnergyNumerical and experimental analyses of a novel natural gas cooking burner with the aim of improving energy efficiency and reducing environmental pollution
2023, EnergyCitation Excerpt :The strategy pursued in the present work was to solve the steady-state flow, combustion and heat transfer in the region outside of the vessel wall. In this regard, Eqs. (1)–(8) were solved by the finite-volume method using a commercial code [27, 38]. A two-step reaction mechanism involving three reactions was used for methane.
Investigation of a novel combustion stabilization mechanism and combustion characteristics of a multi-nozzle array model combustor
2022, FuelCitation Excerpt :The multi-nozzle array combustion with high-speed and without swirl [5,6], which is similar to micromix combustion [7], can effectively reduce the risk of flashback caused by hydrogen enriched fuel. The reaction zone of the combustor is relatively short, which shortens the residence time of flue gas in the combustion chamber, thus reducing NOx emissions [2,8,9]. Basing on the development of the ultra-low emission lean premixed combustion technology for industrial gas turbines, this paper focused on the experimental study of the combustion performance of a novel microscale multi-nozzle array combustor.