Ammonia Combustion Stability: NOx Emissions and Mitigation Strategies

Abstract

Ammonia, as a carbonless carrier of energy, presents considerable potential for hydrogen storage and production, as well as for power generation, thanks to its high energy density and relatively easy transportability. However, the practical adoption of ammonia in combustion systems faces major stability challenges—chiefly its low reactivity, slow laminar burning velocity, narrow flammability envelope, and high ignition temperature. These attributes increase the risks of flame instability, misfire, and incomplete combustion, which, in turn, can elevate levels of unburned ammonia and greenhouse gas emissions such as NOx—posing significant health and climate concerns. Stable ammonia combustion demands optimization of several interrelated factors: the air–fuel equivalence ratio, flame temperature, flow regime, and combustor design are critical for maintaining reliable operation. Particularly pivotal is the control of the air–fuel equivalence ratio; excessively lean conditions can trigger flameout. Modern systems utilize real-time monitoring of flame and exhaust properties to diagnose and prevent instabilities. Advanced combustion strategies, such as transitioning to diffusion or flameless (MILD) regimes, substantially expand the stable operating window, especially under lean conditions. Overall, sustaining stable ammonia combustion is essential for maximizing efficiency and emission control, and integrating aftertreatment (deNOx) technologies is crucial for sustainable, clean-energy implementation.

1. Introduction

Combustion instability has been observed in most combustion processes like those of gas turbines, rocket motors, power generation, steel and cement plants, and food production. Unstable combustion conditions can lead to non-uniform thermal distribution, high pollutant emissions, low efficiency, furnace vibration, and safety [1].

Industrial boilers and furnaces sometimes develop low-frequency vibrations because of the way the burner interacts with the acoustic behavior of the combustion chamber or the ductwork connected to it. In simple terms, pressure waves created by acoustic resonances travel back to the burner, where they match up with natural fluctuations in the combustion process caused by turbulence and chemical reactions. When these pressure waves line up with the flame’s heat release fluctuations, the normal damping forces that would usually weaken the pressure waves can no longer do their job. As a result, a self-reinforcing feedback loop forms, causing the vibrations to grow stronger instead of disappearing. These oscillations, known as thermoacoustic oscillations or rumble, have been extensively studied due to their potential to negatively impact thermal efficiency and emissions and cause mechanical vibrations leading to structural damage [2].

Instabilities manifest in various fluid flows, such as water flowing from a tap, smoke emanating from an incense stick, and the flow between two concentric cylinders, as well as in a layer of heated liquid.

Combustion instability arises from tightly coupled interactions among fluid flow, chemical reactions, and heat transfer within the combustor. These coupled processes can reinforce one another and give rise to large-amplitude oscillations, so carefully designed studies that selectively isolate or “decouple” individual effects are needed to clarify the underlying mechanisms and their relative importance [1].

The scope of the present work encompasses premixed, partially premixed, and non-premixed ammonia combustion across laboratory-scale burners, engines, and gas-turbine-relevant configurations. The included studies are peer-reviewed experimental, numerical, and theoretical works that explicitly address ammonia-specific behavior flame stability, extinction, blow-off, flashback, oscillatory behavior, and pollutant formation (notably NOx and unburned NH₃) under varying equivalence ratios, pressures, temperatures, flow regimes, and combustor designs, as well as deNOx aftertreatment [1,3].

This study presents a comprehensive review of combustion instability in ammonia-fueled low-carbon combustion technologies, with a focus on applications relevant to NOx emission mitigation.

1.1. The Characteristics of Ammonia Combustion Instability

There are technological challenges associated with ammonia combustion that cause stability problems. Ammonia flames are especially prone to instability because they burn slowly and have low reactivity and narrow flammability limits, so small perturbations can strongly disturb heat release and coupling with combustor acoustics [1].

A stable flame is one that burns steadily and uniformly, whereas an unstable flame flickers, flaps or changes shape, and may eventually extinguish itself. The operating conditions of an ammonia combustion system can have a significant impact on its flame stability (A: stable flame, B: unstable flame, C: flameless condition) and the risk of extinction. Some of the key factors that affect flame stability (Figure 1) include pressure, temperature, the equivalence ratio, flow velocity, the exhaust gas recirculation ratio (k_v), air and fuel injection methods, and nozzle types, as well as combustion mode [2,4].

Combustion instability manifests in various ways, impacting system performance and combustion characteristics. Under specific conditions, unstable combustion may even enhance the system’s performance by promoting the mixing of fresh and exhausted gases. However, the occurrence of thermoacoustic combustion instabilities, marked by rising pressure and heat release oscillations, often results in detrimental effects [5].

In such instances, the flame front tends to be confined to the surfaces of large vortexes in the primary region of the burner quarl. This non-uniform thermal distribution leads to local thermal peaks, causing an increase in thermal NOx and a reduction in the reaction rate. Oscillations of significant amplitude signify undesired combustion behavior, resulting in abrupt reductions in system performance [5].

Dealing with rumble is a real challenge for combustion engineers because it is hard to predict and can be influenced by many different design and operating factors. Things like fuel quality, burner swirl and staging, the behavior of induction and draft fans, duct design, and even the shape of the combustion chamber can all play a role. These factors interact in complex ways, which is why two boilers or furnaces that look the same can behave very differently in practice—one may run smoothly while the other develops rumble problems. Identifying the causes of thermoacoustic vibrations in industrial burners and improving diagnosis and remediation methods is crucial to mitigating these effects [5].

Whether instability develops depends entirely on the phase and strength of this acoustic–heat release interaction, with constructive coupling driving sustained oscillations.

Combustion instability is basically a “conversation” between the flame and the acoustics in the combustor. Small changes in fuel or inlet air under lean premixed conditions make the heat release wobble in time, and this unsteady heat release behaves like a strong sound source that sends acoustic waves through the chamber [6]. Some of these waves reflect off boundaries and come back to the flame, which then makes the heat release even more unsteady, strengthening the oscillations in a feedback loop; nonlinear effects finally stop the growth and set a saturated amplitude, so whether instability appears depends on how strongly and in what phase the acoustics and heat release are coupled [6,7].

During the transition from a stable to an unstable flame, which is caused by changes in fuel composition, a novel methodology can detect multiple modes of combustion instability effectively in a gas turbine combustor, as well as dynamic pressure and flame images [6]. In this technique, spectral analysis of dynamic pressure is used to develop a new filter bank method. Sequential processing with a triangular filter with Mel-scaling and a Hamming window is utilized to enhance the accuracy of the method. By determining the magnitude of filter bank components, the instability criterion is determined. It has been shown that the filter bank method performs better than two conventional methods based on root-mean-squared dynamic pressure and temporal kurtosis. Based on the results, the filter bank method shows comparable performance in terms of detection speed, sensitivity, and accuracy to other methods. Moreover, the filter bank components enable the analysis of various frequencies and multi-mode frequencies. The filter bank method can, therefore, be considered as an additional prognosis tool for determining multi-mode combustion instability in a gas turbine combustion monitoring system [7].

1.2. The Challenges and Effective Factors Affecting Ammonia Combustion Instability

Ammonia is widely utilized as a carbon-free fuel in internal combustion engines because of its high energy content and broad availability. The low reactivity, slow burning ignition temperature of ammonia and NOx emissions are crucial challenges for combustion stability [1,8].

Ammonia has a low laminar burning velocity (LBV) (roughly one-tenth that of methane), making it difficult to maintain a stable flame, especially at high flow rates. It has a very narrow flammability range, which increases the risk of flameout (extinction) under lean conditions. It requires higher temperatures to initiate combustion compared to hydrocarbons. The high temperatures needed for efficient combustion can promote thermal NOx formation in addition to fuel-NO formation [3].

Key influencing factors include fuel properties (low burning velocity, long ignition delay, and strong fuel-NO formation), operating conditions (pressure, equivalence ratio, inlet temperature, and residence time), and the mixing/flow field, especially when swirl or hydrogen–methane co-fuels are used to modify flame speed and shape. These features create major challenges: maintaining stable ignition and flame holding over the load range, controlling coupled thermoacoustic instabilities and high NOx and ammonia slip, and dealing with uncertainties in detailed chemistry at high pressures and high hydrogen fractions, which complicate design and prediction [3].

Ammonia combustion instability occurs when the combustion process in an ammonia–air burner becomes unstable, leading to fluctuations in flame intensity and temperature. It is difficult to achieve stable ignition and flame holding over a practical load range with pure or high-NH₃ fuels and to simultaneously control thermoacoustic instability and very high fuel-NO while maintaining efficiency. There is mechanism uncertainty at high pressure and high H₂ fractions, which complicates predictive design of combustors and control strategies [3].

Combustor design and operating factors such as swirl strength, injector design, fuel nozzle modification, high pressure, lean premixing, inlet temperature, conditions and residence time all shift stability limits and mode structures. Mixing and co-fuels such as hydrogen or methane blending strongly affect flame anchoring, shape, and susceptibility to thermoacoustic modes [1].

The temperature in the combustion system impacts flame stability. Due to the ignition temperature of ammonia, below 800 °C there can be no combustion reaction. Higher temperatures (>800 °C) can lead to higher flame speeds and improved stability but can also cause the combustion reaction to become more exothermic, resulting in elevated flame temperatures and consequently increasing the likelihood of thermal runaway [8].

The pressure in the combustion system can also affect the flame stability by changing the density of the gases and the rate of heat transfer. Generally, a decrease in pressure will cause a decrease in flame stability, as the lower density of the gases makes it more difficult for the fuel and air to mix and for heat to transfer. Large pressure oscillations lead to liner/burner fatigue, hardware damage, and forced outages. Excess NOx and unburned NH₃ (“ammonia slip”) cause environmental and safety concerns. Flashback or blow-off occurs when using high-H₂ blends to stabilize NH₃ combustion [8].

The equivalence ratio, or the ratio of fuel to air in the combustion system, affects the flame stability by changing the rate of the reaction and the availability of oxygen. An increase in the equivalence ratio can improve flame stability by increasing the fuel concentration but can also increase the risk of over-firing and thermal runaway [8].

The flow velocity of the fuel and air in the combustion system affects the flame stability by changing the mixing characteristics of the fuels and the speed of the reaction. Higher flow velocities can improve flame stability by promoting better mixing but can also cause the flame to become more unstable and susceptible to extinction if the flow becomes too fast. The presence of inert species, such as nitrogen and carbon dioxide, in the combustion system can affect the flame stability by reducing the availability of oxygen and changing the heat transfer characteristics of the system [8].

The associated risks include large pressure oscillations that damage combustor hardware, excessive NOx and unburned NH₃ with environmental and safety impacts, and increased flashback or blow-off tendencies (when high hydrogen blending is used for stabilization) [9,10,11].

A high chamber temperature, a well reactive zone and a stable environment are required for the start of combustion. In the reaction zone, the fuel and oxidizer undergo a chemical reaction through convection and diffusion, whereby both the fuel and the oxidizer blend prior to the chemical reaction. The proportions of burnt gases are quite low. The rate of combustion is usually controlled by the transport and mixing mechanisms and not by the chemical reaction, which leads to the increased stability of the flame [12].

The impact of instability on unburned ammonia, NOx emissions (ppm) and % O₂ is to be expected. The reaction pathway involves the formation of intermediate species such as NO, N₂O, H₂, and OH. The concentrations of these intermediate species can impact the stability of the reaction and the formation of the final products [8].

1.3. Multiple Instability Modes

The stability of an ammonia combustion flame is an important factor that affects its efficiency and safety. Analyzing ammonia combustion instability is an important step in evaluating the safety of systems that use ammonia as a fuel or a reactant to prevent damage to equipment and ensure the safety of personnel. It typically involves monitoring the pressure and temperature of the combustion system [8].

The Lewis number, low laminar burning velocity (LBV), and Damkohler number are indicators that can be used to measure thermal stability. By understanding the effects of these parameters on flame stability, ammonia combustion systems that control flame stabilization can be designed and operated. In some cases, flame stabilizers, such as catalysts or flame rods, may be used to improve flame stability and reduce the risk of extinction [13].

Multiple instability modes, including thermal–diffusive, hydrodynamic, and thermoacoustic instabilities, are frequently intermixed, without their physical origins, time scales, and governing equations being clearly distinguished. Commonly, three general classes of instabilities are distinguished: chamber instabilities, intrinsic instabilities, and system interaction instabilities.

Instabilities inside combustion chambers are a natural part of how flames behave. For example, premixed flames can show hydrodynamic instability, known as Darrieus–Landau instability, and they can also become thermo-diffusively unstable when the Lewis number is not equal to one. These types of intrinsic instabilities arise because the combustion process interacts and couples with the acoustic behavior of the system, making the flame sensitive to pressure and flow disturbances. The resonant modes ensuring feedback typically have a planar nature, with wavelengths equivalent to the total system in the longitudinal dimension. System interaction instabilities involve interactions with the feed and exhaust and are generally characterized by low-frequency oscillations (subsonic) [5].

In the third category, combustion is still linked to acoustic modes, but these modes are tied to the chamber’s natural resonances. They often appear as vibrations or oscillations that move sideways (transversally) or around the chamber (azimuthally), rather than simply back and forth. The wavelength associated with the third category is determined by the chamber diameter, resulting in oscillations falling within the high-frequency range (supersonic) [5].

1.3.1. Hydrodynamic Stability Characteristics

Ammonia combustion is characterized by inherently low chemical reactivity and non-unity Lewis numbers, defined as the ratio of thermal to mass diffusivity. An overturning instability occurs as a result of thermal convection. Thermal instability occurs when there is a rapid change in the temperature of the flame, leading to a disturbance in the heat release rate and resulting in fluctuations in the combustion process. This type of instability can be caused by a variety of factors, including changes in the air–fuel ratio, variations in the heat transfer rate, and fluctuations in the flame temperature [14].

Instabilities are more likely to arise in weakly burning flames, making them particularly relevant for ammonia combustion, given its inherently low reactivity. Although instabilities in premixed NH₃ flames have been examined in a few studies, reported cellular flame fronts arising from the combined action of hydrodynamic and diffusive–thermal mechanisms, and the instability characteristics and stability limits of non-premixed NH₃ flames remain poorly documented [15].

In contrast to premixed flames, where the hydrodynamic (Darrieus–Landau) instability of intrinsic non-premixed instabilities is governed predominantly by diffusive–thermal effects, cellular structures tend to be produced at small Lewis numbers and pulsating behavior at large Lewis numbers [16].

The cellular instability of a premixed mixture causes diffusive–thermal instability and hydrodynamic instability. Premixed flames and diffusion flames exhibit diffusional–thermal instability [15].

Hydrodynamic instability occurs when there is a disturbance in the flow of the reactants and the combustion products. This type of instability is typically caused by turbulence in the combustion chamber, which can result in fluctuations in the velocity of the air flow, leading to changes in the mixing of the fuel and air [17,18].

The hydrodynamic instability demonstrated a monotonically increasing trend with rising pressure and exhibited non-monotonic variations with increasing stoichiometric ratios. On the other hand, thermal–diffusive instability experienced a significant increase with higher stoichiometric ratios but showed less sensitivity to pressure variations [19].

To characterize the dynamic changes in hydrodynamic instability modes during flame extinction, a frequency analysis and bluff body inflow velocity regimes are required [20].

The hydrodynamic flame instability or Darrieus–Landau (DL) instability of premixed flames, which is caused by the expansion of gas following combustion, causes hydrodynamic disturbances that enhance the perturbations of the flame front in turbulent flows [21].

Viscosity, heat conduction, and species diffusion all have significant effects on flame stability. These influences were highlighted in studies that sought to improve the classic Darrieus–Landau (DL) analysis by incorporating more realistic physical behavior, as well as in asymptotic analyses that revealed explicit relationships between stability and key physical parameters. Hydrodynamic instability, meanwhile, is an inherent feature of flame propagation and is directly connected to the expansion of combustion products as the flame moves [21].

In order to determine the hydrodynamic stability of a planar flame (deflagration), the complete system of equations must be solved, including thermal conduction and energy release due to chemical reactions. This theory provides a rigorous justification of the Darrieus–Landau assumption that the flame-front velocity is constant as a result of large-wavelength perturbations, which is the necessary supplementary condition in the discontinuous flame front model. Analytical solutions are obtained for an arbitrary activation energy for the suppression of flame-front instability. The obtained solution does not depend on the form in which energy is released. In addition to finding the perturbation growth rate numerically, the eigenvalue problem is also used to find the perturbation growth rate [22].

In simple terms, instability occurs when energy moves from a smooth, organized flow into small, disorganized disturbances. As these velocity disturbances grow and become large enough compared to the main flow, nonlinear effects start to dominate. This causes the orderly flow to break down and transition into turbulence, where the motion becomes irregular and chaotic. During this shift from a neat laminar flow to a turbulent one, the way mass, momentum, and energy transported through the system change dramatically [23].

1.3.2. Thermoacoustic Instability Characteristics of Ammonia Combustion

Thermal instability often occurs when a fluid is heated from below. If the temperature difference becomes large enough, the buoyancy forces created by the hot, rising fluid can overpower the stabilizing effects of viscosity and thermal conductivity. When this happens, the fluid becomes unstable and begins to circulate or form convection patterns.

Acoustic combustion instability can be thought of as a self-sustaining oscillation. Even though the combustion process itself is not periodic, it can feed energy into sound waves, allowing them to persist without dying out. This happens because the acoustic waves, in turn, influence the way the flame releases heat, creating a continuous feedback loop. In this situation, the characteristics of the oscillations—such as their amplitude, shape, and frequency—are determined by the system’s own internal properties rather than by any external forcing [7].

Thermal–acoustic flame instabilities influence gas turbine emissions. Engine components are susceptible to severe structural damage, reduced operability, and inefficiency due to the coupling mechanisms of combustor acoustics and combustion heat release fluctuations [24].

Analyzing combustion acoustic phenomena using large eddy simulations reveals that resonance pressure oscillations cause severe thermal load, resulting in the mechanical failure of a chamber in compression-ignition engines [25]. When the flame front and the acoustic field inside the chamber are coupled in phase, the oscillations of both variables are amplified. A coupling mechanism produces oscillations that self-sustain and cause combustion to become unstable, usually generating resonant acoustic modes in the combustion chamber or causing chaos in the combustion process [26].

When pressure fluctuations in the combustion chamber are coupled with the rate at which heat is released, thermoacoustic instability occurs. As a result of fluctuations in flame temperature, this type of instability can result in changes in the pressure waves generated by the combustion process [27].

During gaseous combustion, sound is emitted according to the classical set of mass, momentum, and energy conservation equations. An oscillation with a large amplitude is a common symptom of thermoacoustic instabilities. Instabilities are spontaneously generated and oscillations are maintained through a feedback loop between combustion and the acoustic field. Acoustic noise can be generated by unsteady combustion [28,29].

Acoustic waves are generated by the unsteady heat release of the flame, which are reflected at the boundaries of the system, resulting in standing waves. Acoustic fluctuations result in flow and mixture perturbations, which in turn affect the flame, resulting in a fluctuation in the heat release. Thus, the loop is closed. Depending on the phase between the heat release and the pressure, the oscillations will be amplified or damped. As opposed to combustion noise, thermoacoustic instabilities exhibit oscillations of distinct amplitudes and frequencies [29].

In a lean, premixed, pre-vaporized, multi-nozzle gas turbine model combustor, the impact of fuel variations on thermoacoustic instability characteristics, as well as flame/flow dynamics, was investigated experimentally with high-speed particle image velocimetry and flame OH chemiluminescence measurements [26,30].

2. Ammonia Combustion Stability Conditions and Limits

Identifying gaps in the understanding of ammonia combustion stability is essential for rigorously assessing how various combustion methods and burner configurations influence stable flame operation [31]. The flame stability limit of ammonia combustion is defined as the portion of operating conditions (the pressure, temperature, equivalence ratio and velocity of the flow) over which a flame is maintained. Current knowledge lacks detailed comparisons across different burner geometries, ignition schemes, and operational regimes—especially regarding how these factors affect flame propagation, blow-off limits, and sensitivity to the equivalence ratio (Figure 2) and temperature [1].

These are the limits that are of significance for the safe functioning of industrial processes that utilize ammonia as fuel, like those of boilers, turbines and engines [32,33]. Because of combustion stability, a flame remains alight over a wide operating range while burning smoothly [34].

A stability diagram and the flashback generation condition of NH₃/air flames based on the equivalence ratio are shown in Figure 3 [35]. The stability limits of ammonia–air flames are functions of the equivalence ratio [22,36]. For any combustion chamber, there are upper (rich) and lower (lean) limits to the air–fuel ratio, beyond which flame extinction occurs [3].

A stability map of the fuel staging tangential injection combustor and a diagram are shown in Figure 4. The premixed ammonia–air tubular flame (Figure 4) decreased as the global equivalence ratio (φ_global) increased for different bulk velocities (U_bulk = 40–160 cm/s). NO emissions also decreased rapidly with increasing pressure up to 10 bar [14].

The characteristics of premixed ammonia–air under elevated pressure and temperature conditions at various equivalence ratios are illustrated in Figure 5 [36], which depicts stability limits correlated with exit velocities corresponding to equivalence ratios. The lower limits of combustion stability arise from heat losses, leading to phenomena such as flashback or extinction [37]. On the other hand, the upper limits stem from inadequate residence times, causing blow-off of the injected mixture jet [38]. Flames lacking coflow exhibit extended stability limits with higher fuel-equivalence ratios, owing to increased unreacted fuel in the premixed flame zone. This unreacted fuel mixes with ambient air, undergoing combustion, resulting in non-premixed flames and suppressed blow-off. This highlights the asymmetric nature of the upper stability limits in terms of fuel-rich flames [16,17,34,39].

Lean blowout for CH₄-air flames happens at an equivalence ratio of ϕ = 0.53, while for NH₃-air flames it occurs at ϕ = 0.70. When blending NH₃ with CH₄, reducing the proportion of ammonia (X_NH3) leads to a lower lean blowout equivalence ratio (Figure 6). Even adding just 5% of either CH₄ or H₂ can noticeably improve the lean stability limit in ammonia combustion. Interestingly, increasing the ammonia content up to X_NH3 = 0.70 does not drastically affect overall flame stability [40].

The upper stability limit shows a different trend as the ammonia content increases (Figure 6). At high ammonia fractions (X_NH3 ≥ 0.7), rich blowout can replace flashback as the limiting mechanism. For low to moderate ammonia levels (X_NH3 ≤ 0.4), the equivalence ratio at flashback rises gradually with ammonia addition, which is consistent with the reduction in flame speed. Once the ammonia fraction exceeds about 50%, however, the flashback limit shifts much more rapidly, with the equivalence ratio nearly doubling as the ammonia increases from 0.5 to 0.7. Since the bulk flow velocity remains similar under flashback conditions, this sharp change highlights the strong influence of ammonia chemistry on flame stability at high ammonia fractions [40,41].

2.1. Effective Factors Affecting Ammonia Combustion Stability

Instability during ammonia combustion processes depends on various factors, such as air preheating temperature, reaction zone temperature, pressure, and reactant concentrations, involving several chemical species and mechanism reactions.

Furthermore, ammonia combustion stability is strongly affected by hydrodynamic and chemical parameters, such as flow velocity, the equivalence ratio, and ignition energy. Instability may arise under conditions of low excess air, fluctuations in O₂ and NO₂ concentrations in the exhaust stream, and inadequate residence time of the injected mixture jet, potentially resulting in flame blow-off. Because of the interplay between these factors and the effects of operating conditions on them, engineers and operators can design and operate ammonia combustion systems that are safe, efficient, and stable [42].

For a lean premixed flame stabilized behind a circular cylinder, hydrodynamic instability for a Lewis number (Le) > 1 is directly proportional to the equivalence ratio and pressure. Upon reducing the equivalence ratio (ϕ) at a fixed Reynolds number (Re), it is found that the flame transitions from a steady mode to a varicose mode and then to a sinuous mode [41].

Fluctuating flame intensity, which may appear in the combustion chamber as a flickering or change in size, indicates instability in the combustion process. The key parameters affecting the combustion process are ammonia and air flow and its distribution in the combustor; the temperature and pressure of the inlet air, which are influenced by ambient conditions; compressor performance; load; grid frequency; fuel nozzles; and combustion chamber design and condition [1].

Measuring factors associated with ammonia combustion instability, such as the pressure and temperature signals, can be used to detect fluctuations and combustion instabilities in real time, while exhaust gas analysis provides information about the complete combustion process and helps to identify any incomplete combustion that may lead to increased emissions of harmful byproducts such as Nox in addition to unburned ammonia, as well as the concentration of ammonia in the exhaust gases [37,39].

To perform a stability measurement, thermocouple sensors are placed in the combustion chamber to measure temperature fluctuations. The sensor signals are then analyzed to determine the frequency and magnitude of the fluctuations. This information can be used to identify areas of the combustion system that are prone to instabilities and to optimize the system design and operation to ensure safe and stable combustion. Methods for evaluating thermal instability in ammonia combustion can have serious safety consequences, including the potential for explosions and release of toxic gas. The high temperatures and fluctuations in the combustion process can cause damage to equipment and a risk to personnel working in the vicinity of the equipment. Additionally, the release of toxic gases, such as nitrogen oxides (Nox) and unburned ammonia, is harmful to human health in the workplace and the environment [1].

2.2. Ammonia-Specific Physicochemical Mechanisms of Combustion and Instability

One of the main effects of composition and species on ammonia combustion stability is the impact on the ignition temperature and flammability levels. The presence of different species in a fuel mixture can alter the ignition temperature and flammability, leading to combustion instability. Another effect of composition and species on ammonia combustion stability is the impact on the heat release rate. The presence of different species in a fuel mixture can alter the heat release rate, affecting the combustion process [1,43].

The results in Figure 7 show how major reactants (NH₃) are progressively consumed and products such as H₂O are formed along the flame front, while intermediate and pollutant species, including NO, NO₂, and N₂O, emerge in distinct reaction zones [43]. Radical species such as NH₂, OH, H, and O peak within the high-temperature reaction region, reflecting intense chemical activity. The comparison among mechanisms highlights differences in predicted species evolution and temperature gradients, underscoring the sensitivity of ammonia combustion chemistry to the selected kinetic model under high-pressure conditions [43].

Figure 7 compares species and temperature profiles predicted by different reaction mechanisms. While all models show similar trends for the major species (NH₃) consumption and H₂O formation, noticeable differences appear for the Nox species and key radicals such as NH₂, OH, H, and O [43]. The detailed mechanisms (Okafor, Glarborg, and Shrestha) predict a steeper temperature rise near the unburned gas region, indicating faster heat release and, consequently, higher flame speeds. They also produce higher peak temperatures at the same spatial location, which helps explain the increased NO formation. In addition, these detailed models yield higher H and O radical concentrations and lower OH levels compared to reduced mechanisms, likely due to more comprehensive reaction pathways, particularly those involving HNO, NNH, and N₂O, that strongly influence NO predictions. Since NO is the dominant Nox species, its control is critical; operating under fuel-rich conditions can reduce NO but at the expense of increased unburned NH₃. Therefore, an optimal trade-off between NO and NH₃ emissions must be achieved to ensure effective and clean ammonia combustion [43,44].

A flame burning with a near-equimolar mixture of NH₃ and O₂ at low pressure (46.7 mbar) consists of both stable molecules, including NH₃, O₂, H₂O, H₂, N₂, NO, and N₂O, and more reactive and short-lived species, including H, O, OH, NH, and NH₂. There is also the concentration profile of HNO and N₂H radicals. The conversion of NO and NH2 and the creation of nitric oxide by the reaction of NH + O₂ →NO + OH play important roles when it comes to flame stability. The reactions between NH₂ and H and OH radicals are the primary means of NH production [24,25,45,46].

2.3. Effects of Combustion Methods on Ammonia Combustion Stability

Among the various combustion modes—including premixed, diffusion, and flameless combustion—conventional premixed combustion is most commonly employed in practical systems. In a diffusion flame, the fuel and mixture are combined in a furnace and ignite when they encounter each other. For safety reasons, many combustors operate in non-premixed mode. Because the fuel and the oxidizer are not premixed, a sudden combustion (explosion) is not possible. Flameless combustion involves the interaction of three essential components: fuel, air, and recirculated combustion products, which form a gas with a temperature exceeding the auto-ignition temperature. The combustion will take place in an oxygen-poor atmosphere, due to the watering down of the gas that burns [29].

Under non-flameless conditions, we can easily see temperature gradients, but we cannot actually see a reaction region, and there is a constant temperature in the furnace when a system is under flameless conditions of combustion. Unlike a classical diffusion flame, the flameless process involves the combustion of preheated air confined by a jet (s) of fuel (s). The fuel is mixed with the burnt gases, then mixed with preheated air.

In flameless combustion, there are no temperature variations. The most remarkable effect in flameless combustion is the appearance, which is created when the oxygen content of the air is minimal (less than 2%) [1,2,4,41].

The flameless combustion techniques have been examined under reaction conditions employed for the analysis. The power was set at 2.5 kW and the air flow rate ranged from 30 to 170 lpm, at an equivalence ratio of 0.2–1.2 and a pressure of 0.10 Mpa. Variations in equivalence ratio were set to fixed concentrations of ammonia NH₃, and then the air velocity was changed. However, flameless combustion is frequently portrayed as inherently stable, despite its stringent operational requirements, including high preheat temperatures, substantial dilution, and extended residence times [1].

Several burners utilize the technologies of air-preheating and air-preheating together, including regenerative burners, which operate in flameless combustion mode, as well as recuperative burners (in which the exhaust gases flow counter to the incoming air in a heat exchanger to recover enthalpy) and burners equipped with integrated heat exchangers [5,21,47,48].

To avoid combustion instabilities within a lean premixed gas turbine and establish suitable control mechanisms, the flame recirculation zone may adjust fluid flow rates and could also be a source of instability. The flame structure significantly influences the amount of heat emitted, flame oscillation, and flame stability. There is no precise definition of the reaction rate for flameless combustion of pure ammonia fuel, but it is volumetric [49].

The combustion noise in the conventional flame modes is significantly high as compared to the flameless modes due to the changes in pressure [5,13,47].

Temperature increments are also severely restricted as a consequence of this low oxygen fraction. It should be mentioned that the temperature profile is constant beyond the initial ignition stage, with no clear temperature peak [13,33].

The efficiency of these methods is largely governed by the temperature of the exhaust gases. The rate of air preheating varies among these burner types and is determined by the temperature ratio between the preheated air and the exhaust gases. The efficiency of an exhaust gas generator decreases dramatically if the air is not preheated. In addition, among all burner types, regenerative burners have the highest efficiency [5,50].

Figure 8 presents the variation in adiabatic flame temperature with the equivalence ratio for different NH₃ mole fractions, based on calculations using the Okafor mechanism. As the NH₃ content in the fuel increases, the adiabatic flame temperature decreases markedly, following a trend similar to that observed for flame speed, highlighting the strong influence of temperature on flame propagation. However, unlike the flame speed, the peak adiabatic flame temperature consistently occurs at φ = 1.0 regardless of the NH₃ fraction. This distinction suggests that while temperature plays a key role, the flame propagation behavior of NH₃–CH₄–air mixtures is also strongly governed by chemical kinetic effects within the flame [21,48,50].

Nox emissions from NH₃/air combustion change with temperature and time under different equivalence ratio conditions. This is a result of the fuel–Nox from the fuel-bond nitrogen in NH₃ and thermal-Nox from the oxidation of N₂ in air at very high temperatures [51,52].

Figure 9 shows the variation in the laminar burning velocity (LBV) with the initial pressure in ammonia/oxygen mixtures with different equivalence ratios. Thee numerical and experimental results for the stoichiometric ammonia/oxygen mixture differ considerably. The laminar burning velocity (LBV) in the stoichiometric ammonia/oxygen mixture increased from 1.02 m/s to its peak of 1.09 m/s as the initial pressure increased from 0.3 to 0.5 atm. As the initial pressure increased to 1.6 atm, the laminar burning velocity (LBV) decreased from 1.09 to 1.024 m/. As shown in Figure 9, the laminar burning velocity (LBV) increased with the initial pressure at the lower side of the initial pressure, and the maximum burning velocities (LBVs) of 0.78 m/s and 0.74 m/s occurred at the initial pressure of 0.7 atm for ammonia/oxygen mixtures with equivalence ratios of 0.5 and 1.3 respectively. With a further increase in the initial pressure to 1.6 atm, the laminar burning velocity (LBV) decreased to 0.76 and 0.74 m/s respectively [53].

Figure 10 indicates the time dependence of the combustion pressure of ammonia-oxygen flames with premixtures with different equivalence ratios. The findings show that pressure is highly dependent on various equivalence ratios, which depicts variability in flame-front behavior. The highest combustion pressure, as shown in Figure 10, occurred at equivalence ratios of 0.2 and 1.0. Within a comparatively lean range (0.2 < ϕ < 1.0), combustion was usually stable. But even at the equivalence ratio of 1.2, not all the experiments exhibited the initiation of combustion, which was an indicator of instability. The mixture was non-combustible in extremely lean (ϕ = 0.1) and rich conditions (ϕ = 1.4), indicating that it is hard to establish stable combustion in such regions because the flame propagation characteristics are unfavorable. However, local extinctions and re-ignitions occurred and caused combustion instability. The deformation of the flame structure caused by pressure fluctuation during combustion has a strong effect on combustion instability [14,35,54].

In a comparison of premixed ammonia–air laminar burning velocities (LBVs) at equivalence ratios of 0.8–1.2, combustion instabilities remained, even for a smaller tube with a 14 mm inner diameter, as presented in Figure 11 [55].

At elevated pressure, very lean ammonia–hydrogen–air flames can achieve notably lower NO emissions, making this regime attractive from a NO-control perspective, though it may come with other practical challenges such as stabilization and unburned fuel. With increasing ammonia content in ammonia–methane blends, there is a clear drop in CO₂ emissions from the combustion products, accompanied by only a moderate reduction in combustion temperature, highlighting a trade-off between climate-relevant emissions and flame reactivity [48,56,57,58].

Swirl stabilizers can be used—by tangential injection or bluff-body recirculation zones—to anchor flames effectively in lean combustion. However, intense swirling increases Nox emissions due to longer residence times. The swirl stabilization technique, which involves tangential injection reactants or forming bluff-body recirculation zones, enhances anchoring in lean combustion but increases Nox under high swirl intensities due to prolonged residence times [8].

In the operation of a 1.9 kW bluff-body combustion furnace, shown in Figure 11, a stable flame was observed only at a 0.7~0.9 equivalence ratio with pure ammonia, but when the hydrogen concentration was increased to 40%, a stable flame was generated at an equivalence ratio of 0.3 to 1.6. When the hydrogen concentration increased to 50%, the operable rich condition decreased to close to 1 due to backfire [22,36,55,57].

Increasing oxygen enrichment strongly intensifies flame instabilities in NH₃/CH₄/O₂/N₂ mixtures. This is mainly because higher oxygen levels raise the adiabatic flame temperature, increasing thermal expansion and strengthening the coupling between pressure fluctuations and flame curvature through thermal–diffusive instability. In contrast, under fuel-rich conditions with a higher methane content, these instabilities are noticeably weaker, as excess fuel dampens thermal–mass diffusion imbalances and limits temperature and thermal expansion increases [3,59].

2.4. Effect of Ammonia Laminar Burning Velocity (LBV) on Combustion Stability

The laminar burning velocity (LBV) of ammonia measures how fast a flame front moves through a uniform fuel–air mixture when there is no turbulence. This is an important factor for ammonia combustion stability, since it determines both how quickly the reaction front progresses and how much heat the flame sends into the surrounding gases.

A low laminar burning velocity (LBV) can result in a slower flame propagation rate and reduced heat transfer, which can increase the risk of flame extinction and instability. On the other hand, a high laminar burning velocity (LBV) can result in a rapid flame propagation rate and increased heat transfer, which can improve stability and reduce the risk of extinction.

The laminar burning velocity (LBV) of ammonia is influenced by several factors, including the composition of the fuel and air mixture, the temperature and pressure of the system, the mixing characteristics of the fuel and air, and the equivalence ratio. The composition of the ammonia mixture, specifically the fraction of ammonia in the mixture, can have a significant impact on the laminar flame velocity, and result in the stability of ammonia combustion [20,53,60].

Combustion stability in swirl premix ammonia–air burners is rather limited. As shown in Figure 12, increasing the burner flow velocity (corresponding to higher power) reduces the domain of the equivalence ratio over which stable combustion can be maintained [1,22,36,48].

To determine the burning velocity (S_u) of an NH₃/air flame at various ϕ values (0–1.4), 10 images were captured and analyzed for each experimental condition. A low laminar burning velocity (LBV) can result in a slower flame propagation rate and reduced heat transfer, which can increase the risk of flame extinction and instability. On the other hand, a high laminar burning velocity (LBV) can result in a rapid flame propagation rate and increased heat transfer, which can improve stability and reduce the risk of extinction [53,60].

Markstein length, a fundamental combustion parameter that describes the relationship between flame speed and the stretch rate, increases steadily as the equivalence ratio goes up, but decreases when the pressure rises from 1 atm to 5 atm. The existing burner technology can be operated by mixing ammonia and hydrogen, but the fast combustion speed of hydrogen causes backfire in the stagnant area in the combustion space, making the flame unstable. In particular, the boundary-layer flashback of an ammonia–hydrogen mixed flame in contact with the wall is an area where backfire is likely to occur because the flow velocity is slow. The flashback limit and the laminar burning velocity (LBV) of ammonia–air and substitute mixtures at various equivalence ratios (0.6–0.9), along with the maximum LBV at φ = 1.1483 K and 101 kPa, are shown in Figure 12 [36,48,58].

A higher ammonia fraction will result in a higher equivalence ratio and a more fuel-rich mixture, which can improve flame stability but also increase the risk of over-firing and heat release (Figure 13) [55]. A higher ammonia fraction can also result in higher heat release during combustion and increased flame temperature, which can improve stability (Figure 13) [16,36].

The addition of hydrogen to ammonia has been considered for combustion safety because the higher the hydrogen concentration, the closer the equivalence ratio is to 1, and a higher oxygen concentration in the oxidizing agent results in a higher average velocity of flashback, and this rapidly decreases at an equivalence ratio of 1 or more. At an inlet temperature of 483 K, the flashback generation condition is closest to that of methane, with a hydrogen concentration of 38%, but the laminar combustion rate is only 60% of that of methane, so it is rather difficult to create a condition consistent with methane. These differences should be considered in design and operation, and a simple relational expression for the flashback generation rate was derived [40,45].

It is important to carefully control the ammonia fraction in ammonia combustion systems to achieve stable and efficient combustion. In some cases, additives such as oxygen or steam may be used to control the ammonia fraction and improve stability.

2.5. Effects of Flammability Limit and Ignition Properties of Ammonia on Combustion Stability

Investigation of the ignition properties of an ammonia flame under different temperature and pressure conditions demonstrates that there is a major difficulty in the use of ammonia fuel with a relatively low velocity of combustion compared to other more common hydrocarbons or hydrogen. To overcome this, a strategic solution is to improve combustion properties using reactive fuels. Suggestions to address ammonia’s low reactivity include blending it with more reactive fuels, using oxygen enrichment, increasing preheating temperatures, and applying techniques like heat regeneration and swirl burners [3,49].

The ignition sensitivity of ammonia in an air mixture is affected by the ammonia fraction and its narrow flammability range (vol. % fuel in air), which is typically around 15% at the lower flammability limit (LFL) and 28% by volume at the upper flammability limit (UFL). These limits vary depending on factors such as pressure, temperature, and the presence of other gases. A higher ammonia fraction can result in a mixture that is more difficult to ignite. The ignition temperature of ammonia plays a crucial role in determining combustion stability and is influenced by several factors, including the air–fuel ratio and engine speed [36].

The presence of diluents or contaminants such as water, sulfur, or carbon monoxide can alter the LFL and UFL of ammonia (Figure 14) [1,69], as can the presence of other gases, such as hydrogen, that can lower the ignition temperature, leading to combustion instability and reduced engine performance. However, hydrogen is a more dangerous material which has a wider range between lower explosive (LEL) and upper explosive (UEL) limits. Combustion stability can be affected by the concentration of hydrogen [70].

To maintain combustion stability, it is important to control the ignition temperature of ammonia, and controlling the concentration of ammonia in air by adjusting the air–fuel ratio is essential to ensure that it remains within the flammability limits, in addition to monitoring and controlling the presence of other gases [3].

2.6. The Lean Blow-Out, Extinction and Stability of Ammonia Combustion Flames

The lean blow-off limit is an important factor to keep in mind when designing and optimizing combustion systems. It essentially determines how little fuel you can use while still keeping a system running safely and efficiently, and it also sets the boundaries for the system’s operating range. Knowing where this limit lies is crucial for making sure that combustion systems work properly and reliably, especially in industrial settings where equipment like gas turbines, boilers, and furnaces are used. Understanding the lean blow-off limit means that you can avoid situations where the flame goes out unexpectedly and can make sure that the system runs as smoothly and economically as possible [71].

Lean combustion is used in gas turbines and aircraft engines to reduce Nox emissions. Burner designs that promote good mixing of fuel and air can also improve the lean blow-off limit by ensuring that the fuel is evenly distributed throughout the flame [10,11,72,73].

In general, a flame characterized by a higher burning velocity and a lower ignition temperature exhibits lower lean blow-off limits. Blowouts are more likely to occur under lean conditions. The area between the stable lifted flame line and the blow-off line is referred to as the lean blowout limit [23,74]. The flame stability limits, including lift-off and blow-out, can be established across a broad spectrum of air inlet flow rates by progressively elevating the air flow velocity while maintaining a constant fuel flow rate [75,76].

As the gap between the blow-off and lift-off limits decreases, the flame stability increases [74,77]. A decrease in air flow velocity is caused by an increase in the radius of the central channel, resulting in an increase in the radius ratio (Figure 15) [78]. For all flames, the Damköhler number correlation was found to collapse blow-off velocity data with a satisfactory level of accuracy [56].

Determining the blowout limits for a swirl-stabilized flame, a non-swirling diffusion flame, and a non-premixed flame involves the calculation of their respective Damkohler numbers [30,79].

Da = (S_L²/α_F)/(U_F/d_F)

(1)

where the key parameters include the inverse residence time (U_F/d_F), the fuel velocity at blowout (U_F), the fuel tube diameter (d_F), the heat release reaction rate (S_L²/α_F), the maximum LBV (laminar burning velocity) (S_L), and the thermal diffusivity (α_F). Introducing swirling enhances flame stability by making a flame five times more stable. Swirl flow contributes to improved flame stability by extending the blowout limits of the flame [79,80].

The addition of ammonia diminishes the reactivity of the fuel mixture. Stability is confined by flashback as well as lean/rich blowout. The inclusion of ammonia expands the stable range for this burner, aligning with measurements of extinction. Flames composed of NH₃-H₂ exhibit lower susceptibility to blowout compared to NH₃-CH₄ flames [40].

The presence of cracked NH₃ with N₂ diminishes the reactivity of the mixture and constricts the lean stability limit. The stable limit becomes leaner with increasing pressure, as illustrated in Figure 16 [31,40].

In high-speed air flows under non-premixed conditions, the Damkohler number serves as a scaling parameter to establish a correlation with flame blowout limits, as depicted in Figure 16 [81]. Flame blowout occurs when fuel is directly injected into a wall cavity and is influenced by the positioning of the fuel injector within the recirculation zone. Beyond estimating blowout limits by analyzing rich and lean limit branches, enhancing the model involves refining the understanding of entrainment into the recirculation zone and incorporating unsteady effects. This is achieved by preheating the shear-layer gases through exposure to hot products in the recirculation zone, thereby increasing the propagation speed of the flame [32,80].

The extinction of flames in ammonia combustion can impact the combustion stability of internal combustion engines. The extinction strain rate, which is the rate at which combustion becomes unstable and extinguishes, can impact the stability of ammonia combustion in internal combustion engines. The extinction strain rate is affected by several factors, including the air–fuel mixture, the size and shape of the combustion chamber, and the presence of contaminants in the fuel. If the extinction strain rate is too high, this can result in combustion instability and decreased engine performance. For example, if the air–fuel mixture is too rich (too much fuel), this can lead to a high extinction strain rate and combustion instability. Similarly, the presence of contaminants in fuel can increase the extinction strain rate and negatively impact combustion stability and performance. When the fuel/air ratio is beyond the rich or lean flammability levels, or when the strain rate applied to a diffusion flame is beyond its extinction level, the flame will become extinct.

As a result, the global equivalence ratio for a simple jet diffusion flame becomes unique, which makes this way of describing extinction comparable to the conventional strain-induced extinction limit—it essentially makes the usual lean and rich extinction limits less relevant. These extinction limits should not be confused with the flammability limits generally reported for premixed or homogeneous reactions [10,72].

The lean blowout process in propane turbulent premixed combustion was investigated using high-speed particle imaging velocimetry and CH chemiluminescence (Figure 17). A normalized time variable, t* (t* = [(t − ts)/(tb − ts)]), is introduced to enable direct comparison among the four test cases independently of their differing extinction durations. This parameter scales the physical time (t) between the onset of extinction (ts) and complete blowout (tb), providing a consistent temporal framework for analyzing extinction dynamics across all cases [82].

There is a significant relationship between flame extinction and flame–vortex dynamics, that is, an increase in downstream shear-layer vorticity is coupled with a decrease in flame-generated vorticity. The vorticity dynamics become more unstable as the equivalence ratio decreases.

Flame stability is generally determined by the fuel and air velocity. Different operating conditions may result in different mechanisms for extinction of flames. The existing regime diagrams, however, do not fully describe the physical mechanisms that control flame extinction. To characterize the flame stability limits, the equivalence ratio and strain rate are employed.

The unstretched flame speed, governed by the density ratio and burning rate, plays a crucial role in characterizing flame stability and validating kinetic models. The stability limits of premixed ammonia–air flames, as functions of the equivalence ratio (0.5–1.5) and mean inlet velocity (Uₙ), at 300 K and 0.1 MPa, are illustrated in Figure 18. The heat values, represented by dotted lines, were calculated using an equilibrium calculation [15,38].

Extinction limits are established based on the strain rate of an unstrained premixed flame and extend beyond the corresponding lean and rich conditions. The stability boundary narrows at Λ = 0 and closely aligns with that of an unstrained premixed flame [34,83].

The equations below explain the relationship of the strain rate to the lean and rich extinction limits.

The lean extinction limit:

ϕ = (Λ/U_a) (1 + φ) − φ

(2)

The rich extinction limit:

ϕ = U₀/[(Λ/U_f) (1 − φ) − 1]

(3)

where ϕ is the global equivalence ratio, Λ is the strain rate, U_f is the dimensionless fuel inlet velocity, U_a is the dimensionless oxidizer inlet velocity, and U₀ represents the oxidizer-to-fuel velocity ratio evaluated under reference conditions [83].

Figure 18 indicates the stability limits of diffusion flames in a counterflow setup. The curve (ϕ = 1, Λ = 1, U_f = 1, U_a = 1) at which ϕ = 1 is acquired by varying the velocities of both fuel and air simultaneously until the flame is extinguished and keeping the ratio of the fuel and air velocities and, accordingly, the global equivalence ratio constant. Similarly, the resulting line, L = 1. The extinction limit can be determined by correcting the fuel velocity, U_f = 1, with a slow variation in the air velocity (the red curve), increasing the fuel velocity incrementally (the blue curve) while maintaining the oxidizer velocity constant (U_a = 1). (Starting from the upper right octant, the results are numbered clockwise.) This has a direct effect on flame extinction in the second and third octants of the flame. With the increased velocity of the fuel and oxidizer, the strain rate slowly increases up to the point when the flame extinguishes [83].

In the second and third extinction limits, there are more constraints, which must be included in transient blow-out computations, particularly when you are manipulating the fuel stream or the oxidizer stream to induce blow-out. These conditions generate distinctively different flame stability behaviors, as will be explained in the following section [34,83].

A stability diagram presenting the various equivalence ratios (0–1.4) and strain rates (0–1.4) is shown in Figure 18. The velocity ratios for air (U_a = 1) and fuel (U_f = 1) and stability diagram boundaries are indicated by dashed–dotted lines. The circles indicate the flame conditions, and the triangles indicate specific blowout conditions [34,83].

The left stream provides fuel with a specific composition (Y_F), density (r_F), and velocity (u_F), while the right stream supplies an oxidizer with its own composition (Y_O), density (r_O), and velocity (u_O). The global strain rate (a_g) and the overall fuel–air equivalence ratio (ϕ_g) are determined by these velocities. The velocities of both the fuel (u_F) and the oxidizer (u_O) at the nozzle’s exit, along with their area-weighted mass ratios and their stoichiometric mass ratios (f_st), all play a role. The local maximum strain rate, as well as the fuel and oxidizer inlet velocities, can be described as functions of the global strain rate and the fuel/air equivalence ratio, which helps in providing consistent stability results.

The fuel/air equivalence ratios are slowly brought down using modulation of the flow conditions in the fifth and sixth octants (blue numbers). This has a direct influence on flame extinction in the second and third octants of the flame. The strain rate increases slowly with the increasing velocities of the fuel and oxidizer, and at some point, the flame extinguishes. The equivalence ratios of fuel and air are gradually decreased by adjusting flow conditions in the fifth and sixth octants (green triangles).

When the fuel/air equivalence ratio exceeds the rich flammability limit, however, the flame will extinguish. As a result, lean/rich flammability limits are considered the dominant extinction mechanism for Λ < 1 in quadrants 1 and 2, while strain-induced extinction occurs in quadrants 2 and 3 (green triangles).

The interplay between strain rate and equivalence ratio significantly influences the physical mechanisms of flame extinction in the first and fourth quadrants. The provided formulation adopts a global perspective, aiming to reconcile experimentally observed blowout limits in confined combustion chambers.

On the contrary, a local analysis based on a triple flame takes into account the equilibrium between burning velocity and convection. Consequently, both theories complement each other.

The current stability diagram does not take into account other factors like heat losses from flame/wall interaction, turbulence, swirl, or flow reversal. Accurately modeling these blow-out conditions will likely require enhancements to flamelet-based combustion models, which right now rely only on strain rate as the main factor. Future models will need to include both changes in strain rate and variations in equivalence ratio to fully capture the extinction regimes [34,83].

The ammonia extinction strain rate is a critical factor in determining the stability of ammonia combustion, and the extinction of flames in ammonia combustion can negatively impact its combustion stability in internal combustion engines. Proper control and management of factors that can lead to flame extinction or impact the extinction strain rate led to improved engine performance [35,42].

In the context of convective–diffusive scaling, flames that are near quasi-steady extinction are expected to show striping patterns. The wavenumber of these stripes is proportional to the cube root of the Zeldovich number. NH₂ is predominantly generated through reactions involving H- and OH-radicals. It is feasible to ascertain the rate constants for some of the reactions under consideration [24,45].

The extinction stretch rate is useful for predicting the stability of flames (Figure 19). The swirling, turbulence and blending of ammonia with a more reactive fuel affects extinction [33,42]. Flame extinction under these conditions can be attributed primarily to flame stretch induced by high strain rates [33,42].

2.7. Effects of Heat Release Characteristics on Ammonia Combustion Stability

The amount of heat released during combustion is directly proportional to the energy content of the fuel. A higher energy content leads to a higher heat release rate, which can impact the combustion stability. The heat released during the combustion of ammonia can affect the combustion process and stability in several ways.

Heat release during ammonia combustion primarily affects stability by elevating the temperature and pressure in the combustion chamber. This can lead to an increase in efficiency and performance, but if the temperature and pressure become too high, they can cause combustion instability, leading to engine knock and decreased performance. The heat release rate is influenced by various factors, including the air–fuel ratio, the compression ratio, the engine speed, and the presence of contaminants. A lean air–fuel mixture can increase the heat release rate, while a rich air–fuel mixture can decrease it [88,89].

Another effect of heat release on ammonia combustion stability is the impact on the air–fuel mixture. Heat release can cause the air–fuel mixture to expand, affecting its composition and flammability. To ensure ammonia combustion stability, it is important to control the heat release rate by adjusting the air–fuel ratio and the engine parameters. In addition, it is also important to decrease the contaminants in the fuel, as they can alter the heat release rate and impact combustion stability [11,73].

The heat released during the combustion of ammonia can have a significant impact on its combustion stability in internal combustion engines. Proper control and management of these factors, including the air–fuel ratio and the presence of contaminants, can ensure efficient and stable combustion, leading to improved engine performance [78].

The heat release rate is a key fuel parameter and an essential quantity in turbulent combustion. However, only a few studies have examined the NH₃–air flame structure, reaction mechanisms, and the heat release behavior of ammonia–air flames. However, in MILD combustion, direct numerical simulations have been carried out to evaluate the NH₃–air ammonia flame structure, reaction mechanism, and heat release characteristics and to compare these with those of CH₄–air and H₂–air combustion [2,4,50,51].

Lean mixtures of fuel and air also reduce the peak flame temperature and hence help reduce the NOx emissions (approximately 1 ppm with 15% excess O₂). Nonetheless, premixing may lead to unwanted effects such as spontaneous combustion, and an extremely lean premixed operation may adversely affect combustion stability by increasing sensitivity to blowouts. Velocities, mixing of species, heat release and flame structure were measured by planar laser-induced fluorescence, chemiluminescence imaging, particle image velocimetry and spontaneous Raman scattering measurements of OH radicals at atmospheric pressure [90].

As the inlet velocity increases, the combustion gas temperature rises because the heat dissipation rate remains essentially constant. Turbulence in the shear layer is high, which results in high product entrainment, which in turn increases the rate of the reaction, enhancing the stability of the combustion chamber. Characterization of the turbulent flame structure confirms that thin reaction zones dominate over the full axial extent of the combustion chamber. The measurements of fuel–air mixing under non-premixed operation show that the fuel is not in contact with the hot products until the moment of its full mixing with the air, and the near-premixed operation presents almost the same performance as the real premixed one, without the safety problems that the former has.

Recirculation of exhaust gas (EGR) in the combustor does not seem to have a major effect on the levels of NOx emissions. Thus, the decline in the emissions of NOx in the stagnation point reverse-flow combustor is mainly owed to the capability to offer steady operation at ultra-lean (and close to premixed) conditions in the combustor [17,34].

The key factor to achieve flameless combustion is the recirculation of product gases, which is defined as follows [3,7,51].

$$m_r={\iint }_{A_z}{\rho V}_{z}dxdy$$

(4)

k_v = m_rec/(m_air + m_NH3)

(5)

where k_v is the recirculation ratio, m_rec is the recirculation gas mass flow rate, m_NH3 is the ammonia mass flow rate, and m_air is the inlet fresh air mass flow rate. To calculate the recirculation ratio, the r density of the mixture gas, the negative axial velocity and the area with negative axial velocity (V_Z) in the simulation are needed.

Anticipating the optimal air diameter, combined with the existing combustor diameter, is crucial for attaining a minimum recirculation ratio conducive to sustaining flameless combustion. The effects of the preheated air temperature on the performance of flameless combustors deserve attention, and minimal influences on NOx emissions are realized at recirculation ratios of more than 3. With proper choice of the air diameter and cylinder diameter of the combustor, optimum recirculation ratio can be obtained, which is required to maintain flameless combustion [51]. The 73 chemical reaction mechanism can accurately predict the temperature and O₂ concentrations in most of the combustor [9,52].

2.8. Effects of Swirl Number on Ammonia Combustion Stability

Swirling induces an interaction between a jet and a vortex, with the recirculation vortex diminishing the velocity of the fuel jet along the centerline, thereby robustly stabilizing the lifted flame. Increasing either the fuel tube diameter or the reaction rate (via hydrogen addition) enhances the stability of the swirl flame, resembling the stability of a non-swirling flame. Additionally, swirling enables overall fuel-lean operation, as the flame in fuel-lean conditions lacks stability without swirling. With an increase in swirl number, the flame stability improves, measured by the maximum fuel velocity [71].

Depending on how the system is run, low-swirl flames can show three distinct combustion regimes. When using a low-swirl injector at atmospheric pressure, you will see an attached flame form at low inlet velocities. As the inlet velocity increases to moderate levels, the flame takes on a W-shaped appearance. With increasing mixture velocity, a bowl-shaped flame structure can be observed. Low-swirl flames exhibit local extinction and relight zones [23,74].

A diagram and flame regime maps of a 4.7 kW low-swirl vane burner for ammonia/air combustion under different air flow rates and different swirl numbers are shown in Figure 20 [74], along with flame shape evolution at different air mass flow rates, measured under atmospheric conditions with a constant fuel supply. Figure 20a–c shows how the shape of the flame changes as the air mass flow rate increases, while the fuel rate stays constant and atmospheric conditions are maintained. As the air mass flow rate increases under constant fuel flow and atmospheric conditions, the flame lifts off and is blown out at a vane angle of 37°, as shown in Figure 20a [74].

As the proportion of annulus air flow rises, the blockage ratio also increases, causing a subsequent elevation in air flow velocity. This escalation in velocity leads to lift-off and blowout, attributed to an amplified tangential velocity and swirl number.

Ammonia on its own has a relatively high lean blowout (LBO) limit of about 0.6–0.8, so it is harder to keep the flame stable at high inlet velocities or very lean conditions than for typical hydrocarbon flames. Once extinction starts, the flame does not gradually weaken; instead, it tends to collapse abruptly and the whole flame is suddenly blown out [3].

Flame stability is strongly influenced not only by fuel composition but also by swirl intensity. Without swirl, stable combustion is not possible even for pure methane or CO, whereas sufficient swirl enables stable flames across a wide range of ammonia fractions, including pure ammonia. Hydrogen proves especially effective in stabilizing ammonia flames, forming a clear flame root even at low swirl, while methane tends to produce lifted and less stable flames. Overall, these results highlight that although high-ammonia combustion is feasible, maintaining stability and controlling emissions—particularly N₂O and CO—require careful control of swirl strength, fuel blending, and operating power [3].

2.9. Effects of Residence Time on Ammonia Combustion Stability

The residence time is a critical factor in determining ammonia combustion stability. Residence time, or the time that the air–fuel mixture spends [3,91] in the combustion chamber before ignition and combustion, can impact the stability of ammonia combustion. Insufficient residence times of the injecting mixture jet cause blow-off. The residence time can affect the mixing of the fuel and air, the flame propagation, and the combustion efficiency. At lower temperatures and with longer residence times, the area after the flame (the post-flame zone) shows higher reactivity for the subsequent reaction [74].

Every structure on the flame’s surface gets carried toward the tip of the flame at its tangential velocity, meaning that each one has a specific residence time.

τ = L/U ≡ h/(U_o cos² (α)) ≡ h. tan(α)/(S_L cos(α))

(6)

where L represents the length of the inclined flame, h indicates the vertical height of the Bunsen flame, and α = cos⁻¹ (h/L) is the half-angle of the tip of the flame [92].

The residence time needs to be juxtaposed with the growth time (1/σ) of the instability, where σ represents the complex growth rate. When the growth time is significantly longer than the residence time, at the base of the inclined flame, small perturbations do not have sufficient time to amplify before being convected out of the flame. Conversely, if the residence time is too short, incomplete mixing of the fuel and air can occur, leading to combustion instability and reduced performance. On the other hand, if the residence time is too long, this can lead to excessive heat loss and decreased engine efficiency. To ensure ammonia combustion stability, it is important to control the residence time by optimizing the combustion chamber design and fuel injection system. For instance, adjusting the shape and size of the chamber, as well as the placement of fuel injectors, can help to control the residence time and ensure efficient and stable combustion [92].

Improvements in flame stability can be made by increasing the residence time using a two-stage combustion concept. This design addresses the strong cooling effect of liquid ammonia evaporation by creating a high-temperature environment, improving droplet dispersion and mixing and raising the residence time [93].

2.10. Effects of O₂ Concentrations on Ammonia Combustion Stability

Atmospheric air is used in most combustion chambers, and the air flow is measured to regulate the provision of oxygen. Excess air and O₂ supply are needed to attain high efficiency, low emissions and low noise.

If the air supply falls short of the stoichiometric requirement and the combustion becomes fuel-rich, this can affect the stability of ammonia combustion and allow unburned fuel to escape through the exhaust stack. This not only wastes fuel but also produces additional emissions and hazardous air pollutants. It also poses a possible safety concern if sufficient fuel then combines with O₂ and ignite. The laminar burning velocity (LBV) varies from 1.4 to 8.23 cm/s over an equivalence ratio range of 0.7–1.3, reaching a maximum of 7.9 cm/s at an equivalence ratio of 1.1. When the oxygen content rises from 0.27 to 0.30, the laminar burning velocity (LBV) increases from 27.5 to 33.9 cm/s. Normally, the proper burning velocity (LBV) of ammonia for practical applications will be at an O₂ volume concentration between 0.35 and 0.40 in an O₂/N₂ mixture [94,95].

This is mainly because there is an increase in the reaction rates of OH, H, O and NH₂ radicals in the presence of a higher O₂ level [96].

It has been demonstrated that enhanced flame propagation with oxygen enrichment (O₂ content of 0.21 to 0.45%) at pressures of 1 to 5 atm and equivalence ratios of 0.7 to 1.5 at 298 K is primarily due to an increase in adiabatic flame temperature, resulting in enhanced concentrations of radicals, such as H, OH, and NH₂ radicals. According to the calculated pressure-dependent coefficient, NH₃/O₂/N₂ flames are strongly influenced by pressure, whereas hydrocarbon and biofuel flames exhibit a weaker pressure dependence. The burning velocity (LBV) rises with a higher O₂ concentration but decreases as the initial pressure increases. At an O₂ content of 1.0, the laminar burning velocity (LBV) reaches a maximum of 125.1 cm/s, and the maximum value is reached at an equivalence ratio of around 0.9 [3,44,45,97,98].

Furthermore, flame velocity demonstrates an increase with rising temperature (Figure 21), particularly showcasing higher sensitivity to temperature changes under lower O₂ contents [37]. Excess air is problematic due to its 80% nitrogen content, which can increase NOx emissions with excess amounts of air–methane fuels [99].

Incomplete combustion is a common occurrence, leading to the release of unburned fuel and oxygen through the stack, even when the air–fuel mixture entering the burner is well-adjusted. The critical consideration is the surplus amount of oxygen beyond what is necessary to combust the fuel. However, assessing the total oxygen content in the flue gas stream can be misleading if operators lack a comprehensive understanding of the measurement’s implications. The challenge lies in determining the excess amount of oxygen in the flue gas compared to the stoichiometric requirement, and this varies depending on the fuel and combustion system. In cases where combustion is more challenging, a higher excess amount may be necessary. Oxygen analyzers prove invaluable in optimizing boilers and fired heaters situated in hazardous areas [101,102].

The flame size (volume) and color are not only affected by the chemical properties of the fuel, but also depend on air preheat temperature, oxygen concentration, the diluents and intermediate species. Different levels of air preheating and oxygen concentration result in distinct flame structures. At a fixed air temperature, increasing the oxygen concentration leads to a reduction in flame size, ranging from approximately 2% to 21% [1].

The flame stability and emissions of oxygen-enriched ammonia flames in a single-swirl combustor were studied using large-eddy simulations validated against experiments. The results show that adding oxygen is an effective way to stabilize ammonia flames, as higher oxygen levels significantly increase heat release and resistance to extinction, outweighing the destabilizing effect of increased flame stretch. Oxygen enrichment also improves axial flame stability during blow-off, although temporary spikes in N₂O and NO₂ emissions can occur, especially at higher oxygen levels. In terms of emissions, increasing oxygen raises NO and NO₂ while reducing N₂O; below 30% oxygen, this increase is driven by both thermal and fuel NO pathways, whereas above 30% it is mainly governed by fuel-NO formation [103].

An oxygen-enriched ammonia flame under conditions of Standard Temperature and Pressure (STP) and an elevated pressure of up to 5 atm showed an improvement in burning velocity with an increase in the oxygen concentration in the O₂/N₂ mixture to 0.35 [40,102].

The level of NOx emissions directly depends on the oxygen content of the incoming air as well as the combustion air temperature. The NOx emissions changed over time at various oxygen concentrations (5, 10, 21, 30, and 70%), with the maximum levels observed at 1000 °C and 800 °C constant temperatures respectively. The more the O₂ concentration (70%) increased, the more the concentration of NOx increased, and the highest concentration was observed at the 70% O₂ concentration [104].

Oxygen enrichment from 21% to 29% broadens the stable operating range for pure NH₃, with an equivalence ratio of 0.45–1.6, accompanied by a 60:1 power turndown ratio at φ = 0.9. Higher O₂ concentrations allow more fuel-rich primary combustion in an air-staged configuration, creating potential for further reducing NOx emissions. As the ammonia blending ratio increases under oxygen-enriched conditions, the optimal secondary air ratio decreases. Pure ammonia combustion with both oxygen enrichment and staging achieves high stability along with low nitrogenous emissions (~185 ppm NO, ~41 ppm NH₃, and undetectable N₂O). This performance is substantially better than single-stage, air-fed pure ammonia combustion at 1 kW, which produces 300 ppm NO, 100 ppm NH₃, 300 ppm H₂, and 20 ppm N₂O. These results confirm that oxygen-enriched staging simultaneously suppresses all major nitrogenous pollutants, demonstrating strong potential for industrial application [105]. However, further oxygen enrichment beyond about 29% leads to flashback in the present configuration, highlighting a practical upper limit imposed by safety considerations [106].

2.11. Effects of Fuel Composition and Species on Ammonia Combustion Stability

To achieve stable pure ammonia combustion, blending ammonia with more reactive fuels such as hydrogen, methane, diesel, gasoline, or propane is an effective approach, as it increases the laminar burning velocity (LBV) and improves flame stability. NOx formation in these fuel blends is strongly influenced by ammonia content, as well as the operating conditions, including air temperature, pressure, and the composition of a fuel mixture (the ratio of the various fuel components), and the species (the chemical elements present in the fuel) [8].

Increasing the ammonia content generally reduces NO at the outlet, with peak NO occurring near stoichiometric conditions for ammonia fractions above 30%. As the ammonia fraction increases, flame stability deteriorates markedly. At 50% ammonia, non-combustion becomes common, particularly when the CO/H₂ ratio is high, because slow reaction kinetics and weak stabilization cause flame extinction under lean and near-stoichiometric conditions. At 60% ammonia, a flame forms only under very rich conditions, while most operating points show distributed or no combustion. However, stable flames are limited to equivalence ratios above about 0.8, and even then NO levels remain high, exceeding 2000 ppm. Beyond 60% ammonia—including pure ammonia—stable flames cannot be sustained in this burner, highlighting the need to blend ammonia with syngas at mole fractions below 50% to ensure stable combustion [9,106].

Operating the burner under fuel-rich conditions is effective in lowering NO emissions, but it introduces a trade-off when ammonia–syngas blends are used. Although rich operation significantly suppresses NO formation, it results in high levels of unburned CO at the outlet, since CO is both a fuel and a regulated pollutant. As a result, an selective catalytic reduction (SCR) system is still needed to control residual NO emissions. However, the substantially reduced NO levels allow for a smaller and lower-maintenance SCR unit. In addition, the average outlet temperature of about 1339 K remains suitable for stainless-steel components and favorable for high SCR efficiency, making rich operation a practical compromise for emissions control [106].

The experimental data also indicate that dilution with H₂O enhances flame stability relative to CO₂ dilution at a given NH₃–CH₄ composition. For a 50–50% NH₃–CH₄ mixture, the onset of instabilities is shifted to higher strain rates when H₂O is used as the diluent instead of CO₂. To interpret this behavior, four artificial inert species (X_NH3, X_CH4, X_CO2, and X_H2O) were introduced into a one-dimensional non-premixed flame model. These placeholder species reproduce the transport and thermodynamic properties of their real counterparts but are excluded from chemical reactions, allowing a clean separation of inert, diffusive–thermal, and chemical effects.

Using a 50–50% X_NH3–X_CH4 diluent as a baseline, the model predicts extinction at a global strain rate of approximately 11%, which reflects the purely inert consequence of reactant dilution. Replacing this mixture with X_CO2 shifts the extinction point to about 7.5%, highlighting the additional diffusive–thermal influence of CO₂, while dilution with X_H2O markedly delays extinction to roughly 18.25% and increases the flame temperature relative to the inert baseline. When real CO₂ is allowed to participate chemically, extinction is further advanced to around 6.75%, whereas enabling reactions for H₂O produces only a negligible change in the extinction limit and flame temperature compared to the X_H2O case. Taken together, these findings indicate that the superior stability of H₂O-diluted flames over CO₂-diluted ones arises from a combination of diffusive–thermal and chemical effects, in agreement with previous observations in non-premixed syngas flames diluted with CO₂ and H₂O. Although radiation absorption can, in some situations, broaden the flammability limits of CO₂-diluted mixtures, this mechanism appears negligible under the present experimental conditions, as evidenced by the narrower flammability range measured for CO₂-diluted blends and its consistency with predictions from the one-dimensional adiabatic numerical model and the corresponding temperature measurements [9,106].

The stability limits of ammonia-containing flames depend strongly on the co-fuel and diluent. For NH₃–CH₄–CO₂ mixtures, flame instability is predicted to occur in the form of pulsations under conditions of high fuel and oxidizer Lewis numbers, consistent with earlier observations of CH₄–CO₂ flames. The pulsation frequency increases as the methane fraction rises, in agreement with theoretical predictions that associate higher effective Lewis numbers with faster oscillations. In contrast, adding hydrogen leads to a fundamentally different instability behavior: owing to its very low Lewis number, H₂ promotes cellular diffusive–thermal instabilities, in which an initially smooth flame breaks into localized hot spots separated by reactant-depleted regions. While these cellular structures are well understood for pure hydrogen flames, the instability mode that emerges when hydrogen is blended with ammonia—whose chemistry favors pulsating behavior—remains unclear, highlighting a gap in the current understanding of diffusive–thermal instabilities in mixed-fuel ammonia systems [106].

When ammonia is blended with natural gas at fractions above about 30%, clear flame instabilities appear, accompanied by significant changes in pollutant emissions. As the ammonia content increases, stable operation becomes more difficult and can only be maintained at reduced power levels. While NOx emissions generally follow expected trends and can be minimized at high ammonia fractions, operation near the stability limit leads to a sharp rise in N₂O emissions, reaching very high levels even when NOx is low. At the same time, incomplete combustion becomes more pronounced, with CO emissions increasing rapidly as soon as flame instabilities occur [93,106].

Compared to methane, ammonia has a narrow stable flame range, as shown in Figure 22 [107]. Lean blowout occurs at ϕ = 0.70 for NH₃–air and at ϕ = 0.53 for CH₄–air flames. Decreasing X_NH3 reduces the equivalence ratio at the lean blowout. Just 5% of CH₄ or H₂ improves the lean stability limit of NH₃ combustion. Ammonia addition up to X_NH3 = 0.70 does not drastically modify the stability. N₂ in cracked NH₃ reduces the mixture reactivity and narrows the lean stability limit. The higher the pressure, the leaner the stable limit. Consistent with extinction measurements, NH₃-H₂ flames are less susceptible to blowout than NH₃-CH₄ flames [40].

Measurements of a spherical ammonia-oxygen flame in a constant-volume chamber focus on key parameters, including laminar burning velocity (LBV), Markstein length, laminar flame thickness, and the critical radius of flame instability. Utilizing different initial pressures (0.5 to 1.6 atm) and equivalence ratios (ϕ) (0.5, 0.75, 1.0, 1.3, and 1.75), flame propagation was examined using a high-speed digital Schlieren photograph system (Figure 22). The investigation revealed a maximum laminar burning velocity (LBV) of 1.09 m/s in the ammonia/oxygen mixture [1]. Notably, as the initial pressure increases, the flame thickness decreases. The Markstein length increases with higher equivalence ratios, whereas it decreases with an increase in initial pressure. The minimum critical radius was measured at 1.8 cm in ammonia/oxygen, and it decreased with the increase in initial pressure [53,108].

3. Stability Limits and NOx Emissions

NOx formation pathways in ammonia flames are strongly influenced by flame stability limits, which in turn affect the levels of NO emitted during combustion.

Table 1. The key factors governing flame stability and NOx emissions.

Factor	Effect on Flame Stability	Effect on NOx Emissions
Temperature	Improves stability at higher than 1200	Reduces NOx at (0.8 > ϕ > 1.2)
Pressure (5–15 atm)	Improves stability and completeness	Reduces NOx by up to 40%
Hydrogen addition (10–40%)	Widens stability, increases burning velocity	Moderate NO increase in lean conditions and reduction in rich conditions
Rich operation (ϕ > 1.1)	Reduces stability slightly	Decreases fuel-NO via NHx radicals
Swirl/tangential staging	Enhances stability	Reduces NOx to ~50% with proper flow staging
Plasma-assisted combustion	Significantly extends lean limit	20–40% NOx reduction
Heat-recirculating (Swiss-roll)	Broadens stable regime of pure NH3	Non-monotonic NO vs. ϕ, lower rates in lean/rich conditions
Flameless combustion	Broadens stable regime of pure NH3 at high temperatures	More than 40% NOx reduction

summarizes the key factors that influence this behavior. NOx (NO and NO₂) emissions from ammonia involve three dominant pathways which include mainly fuel-NO, from N radicals generated during NH₃ oxidation through to HNO, NH, and NH₂ intermediates [2,4,8]. Two other factors have less influence on these emissions, namely, thermal-NO (the Zeldovich mechanism) derived from N₂ oxidation at T > 1800 K and prompt-NO (the Fenimore mechanism) generated via CH/O radicals (negligible in NH₃).

At equivalence ratios (ϕ) = 0.8–0.9, ammonia–hydrogen flames create a NOx peak region, driven by radical interactions [4,8]:

NH + O → NO + H and HNO + H → NO + H₂.

(7)

Richer mixtures (ϕ > 1.1) suppress NO formation, as NH and NH₂ pathways dominate, producing N₂ instead of NO [4,9].

There are some NOx reduction strategies, such as fuel composition optimization, that increase the ammonia fraction in NH₃/H₂ blends, lower the peak temperature and suppress NO through enhanced NHx production [4]. Rich-mixture operation (ϕ = 1.1–1.3) reduces thermal-NO but risks incomplete burning; staged combustion helps maintain oxidation efficiency [1,4].

Also, pressure enhancement that elevates pressures (5–15 atm) increases collisional deactivation of NO-forming radicals (H, O, and OH), decreasing total NOx by up to 40% while improving combustion completeness [7].

The safety of ammonia–hydrogen–air premixed flames is improved by partial NH₃ substitution at normal temperatures and pressures. A reduction in the stability limits and NOx emissions of NH₃–H₂–air flames is observed with NH₃ substitution, nitrogen (N₂) coflow, and mixture injection velocity. Although the absolute value of NOx emissions increases with enhanced NH₃ substitution, the NOx emission index remains almost constant. Increased mixture injection velocity reduces NOx emissions under fuel-rich and coflow conditions. NOx emissions from fuel-rich flames are reduced through the thermal deNOx process in the post-flame region [19,20,44,53,108,109].

Evaluating different degrees of ammonia cracking to enrich the hydrogen content in the fuel influences several critical parameters, including the overall burning velocity, the lean blow-off limit, NO and NO₂ concentrations, and the flame’s response to acoustic disturbances. The results indicate that although ammonia cracking enhances the lean blow-off limit and overall burning velocity, its influence on pollutant emissions and flame stability becomes unfavorable even at relatively low cracking levels, such as 20% [11,73].

The stagnation-point reverse-flow (SPRF) combustor is specifically designed to achieve low NOx emissions, even when the fuel and air are not premixed prior to entering the combustion zone.

Furthermore, staged and swirl combustion that use tangential injection combustors achieve up to 50% NOx reduction using two-stage air delivery while maintaining stability through internal recirculation [1]. The Dimethyl ether (DME) fuel staging/secondary injection technique with multi-zone NH₃/DME (Dimethyl ether) flames achieve over 55% NO removal efficiency at optimized residence times (≈950 °C, ϕ = 0.9) with secondary ammonia use in post-flame regions [4,21,48]. Exhaust gas recirculation (EGR) lowers the oxygen concentration and flame temperature, though excessive dilution (>25%) weakens flame stability [4,110].

The stability of and reduction in the NO emissions of two-stage NH₃/air premixed flames influenced by secondary air injection at various equivalence ratios (ϕ_pri) at 0.5 Mpa are shown in Figure 23 [37].

An efficient low-NOx combustor for liquid ammonia spray was developed and successfully tested, showing major improvements in flame stability and strong control of NO, N₂O, and unburned NH₃ emissions using a two-stage combustion concept. Better temperature uniformity in the primary zone was achieved through controlled secondary air dilution, while a multi-nozzle, twin-fluid atomizer proved more effective than a single-pressure swirl nozzle. A longer combustor further enhanced fuel utilization and emissions performance. As a result, stable combustion of pure liquid ammonia was achieved over a wide equivalence ratio range (0.6–1.3), with near-zero N₂O and unburned fuel emissions and NO as low as 282 ppm. This work represents a practical pathway to stable, low-emission liquid ammonia spray combustion in gas turbine combustors [93].

Moreover, plasma-assisted and catalytic systems for plasma-assisted combustion (PAC) substantially extend lean blow-off limits while cutting NOx emissions by 20–40% through controlled radical chemistry and lower flame temperatures [10]. Finally, flameless combustion can broaden stable regimes of pure NH₃ at high temperatures, achieving more than a 40% reduction in NOx. There was a small NOx reduction in the combustion of ammonia in the diffusion swirl stabilizer (Figure 24) compared to lean flame conditions which produce less than 100 ppm of NOx due to the thermal deNOx mechanism [102,111,112].

Under stable operating conditions, methane largely controls the overall reactivity of the system, helping to stabilize the oxidation process and enhance ammonia reactivity. Most major species scale linearly with the carbon and nitrogen supplied by methane and ammonia, respectively, indicating limited interaction between the two fuels. The main exception is NOx formation, which shows a strong nonlinear response when ammonia and methane are burned together.

The main oxidation pathway is controlled by O₂ through the following set of reactions: NH + O₂ = HNO + O and HNO + O₂ = NO + HO₂, or NH + O₂ = NO + OH.

CO emissions increase almost linearly with methane content and are largely unaffected by ammonia, while hydrogen formation is generally insensitive to ammonia addition, except near extinction. In contrast, NOx emissions rise sharply even with small amounts of ammonia, jumping from near zero to hundreds of ppm, and show a strong dependence on equivalence ratios. Beyond low ammonia additions, further increases have a limited effect on NOx levels. Even at low ammonia fractions, OH radicals formed during methane oxidation consume NH₂ radicals, reducing the availability of key deNOx species and consequently increasing NOx emissions [113].

3.1. NOx and N₂O Mitigation from Industrial Streams

Flameless combustion entails the burning of pre-heated air in the presence of a jet of fuel or a jet of fuel mixture, which combines with burnt gases to reduce the concentration and then is combined with preheated air [69,75]. This approach results in a more distributed and slower reaction. The characteristics of flameless combustion can vary widely based on the specific burner and combustion chamber, which have yet to be standardized and are under research and development [58,78,114]. However, it is encouraging to see that flameless combustion has been performing well in certain domains, and thus it can be regarded as a combustion technology that could use ammonia fuel in the future [3,114,115].

Flue gas recirculation and changing the combustion method effect reductions in NOx emissions. However, despite the increase in temperature, NOx formation remains significantly restricted due to the low oxygen concentration. Reductions in NOx emissions and improvements in the efficiency of burners are achieved by limiting peak temperatures and lowering oxygen concentrations in a flameless combustion mode [3,58].

Figure 25 shows NOx emissions from a swirl flame and flameless combustion of a CH₄/NH₃ mixture (99/1%) over equivalence ratios ranging from 0.68 to 0.87. A notable increase in CO emissions can be observed at equivalence ratios above 0.8 [103,116].

Flameless combustion is distinguished by the uniformity of the reaction zone, reactant mixtures, and reduced temperature peaks resulting in a uniform temperature and species distribution. During flameless combustion, the combustion zone is filled with three components—fuel, air and recirculated burnt gas—to generate a gas with an auto-ignition temperature greater than that of the air and fuel mixture. The recirculation of exhaust gas does not have a substantial impact on NOx emissions [114,117].

Flameless combustion is founded on the fact that air is mixed with flue gas and then reacted to reduce the rate of the reaction by reducing the concentration of oxygen. There are significant restrictions on temperature increases because of the low oxygen fraction. It is important to point out that the temperature profile does not change following the initial ignition stage and does not display a sharp rise in temperature. The burning of the hot gas is diluted and, as a result, auto-ignition occurs in a low-oxygen environment. The most peculiar property of flameless combustion is the fact that it does not show a flame when there is a low concentration of oxygen in the air, which is usually below 2 [4,37].

Flameless combustion takes place at temperatures above the auto-ignition point and under low-oxygen conditions, with the reactions occurring within a diffusion layer. This causes high gradients in the concentration of fuels and oxygen following injection; hence, the concentration of fuels and oxygen drops rapidly. Also, the temperature stabilization and low oxygen that is typical of flameless combustion contribute to the prevention of some byproducts of combustion, including NOx and unburned ammonia [4,47,89].

The most common approach to promote flameless combustion is to preheat the air or employ a heat exchanger to recover the exhaust gas enthalpy. This method enhances burner efficiency, raises the final mixture temperature, and can lead to increased NOx emissions. Despite the elevated temperature, NOx production is significantly constrained by the low oxygen content [88].

NOx emissions are influenced by the preheated air temperature in two distinct scenarios. One involves nitrogen molecules embedded in the fuel matrix interacting with excess oxygen, while the other involves a direct reaction of nitrogen and oxygen molecules [89]. NOx emissions are reduced in a flameless combustion mode, improving burner efficiency and limiting temperatures and oxygen and residence times [21,48].

Using a suitable air nozzle diameter and geometry for a combustor, the recirculation ratio needed to maintain flameless combustion can be achieved [88]. The geometry scaling rules of a stagnation-point reverse-flow combustor have been developed with a simplified coaxial and counter-jet, leading to NOx reduction due to stable combustion even under ultra-lean conditions [42,118].

One of the most crucial aspects of the flameless combustion system design is the control of the thermo-fluid pattern in the combustion chamber by means of air and fuel nozzles. Flameless combustion has a high-speed air jet that generates strong flue recirculating gas that leads to a better convective heat transfer, and this leads to a uniform temperature distribution. The distinctive configuration of revers Air injection flameless combustion used to inject fuel at the back of a powerful air jet is a distinctive characteristic of the design when compared to the conventional design of a combustor [89,91].

MILD (moderate or intense low-oxygen dilution) combustion, as a form of flameless combustion, offers the unique advantage of simultaneously reducing emissions while improving efficiency, especially for low-reactivity fuels such as ammonia. This highlights the need for further investigation into ammonia’s behavior under MILD conditions, which remains insufficiently understood. The peculiarities of MILD combustion, which is one of the forms of flameless combustion technology, are the fact that, along with the ability to decrease emissions, it can also increase efficiency, in particular that of low-reactive fuels like ammonia. Nevertheless, the properties of pure ammonia MILD combustion have yet to be determined. MILD combustion of ammonia–air has been numerically simulated to assess the mechanism of all the reactions, flame structure and heat release, compared with CH₄-air and H₂-air [42,88]. Ammonia–methane under MILD combustion results in increased NOx emissions compared to pure combustion of ammonia–hydrogen mixtures. When combusting ammonia and methane in MILD conditions, NOx emissions are nearly less than a third of those in conventional combustion, indicating unique NOx mechanisms in MILD and conventional combustion [89,119].

When using a 70/30 blend of a NH₃/H₂ mixture, the lowest emissions of NOx were 450–654 ppm under lean conditions (ϕ = 0.5–0.8) and 344–211 ppm under rich conditions (ϕ = 1.0–1.2). The unburnt NH₃ and H₂ emissions are still low in lean conditions. Flameless combustion tends to show similar or even better emissions performance in lean and rich conditions than conventional combustion methods [102,118].

The best way to control NOx in NH₃-containing fuel is by increasing the equivalence ratio above 1.0. The rate of decrease in NOx is reduced above an equivalence ratio of 1.2, which is the optimal condition for lowering fuel NOx [3,13].

3.2. Post-Treatment of NOx Emission

Ammonia, as a reduction agent, is injected into the exhaust gas for NOx reduction and generates only nitrogen and water. The most popular means of achieving post-NOx reduction is selective catalytic reduction (SCR) technology, which employs an aqueous urea solution that, when injected into an SCR catalytic converter in a specific manner, transforms into ammonia. Various technologies have been developed to decrease the NOx emissions from combustion exhaust gas after the SCR system. In SCR in gas turbine power generation systems, a novel deep learning model network is employed to predict the current combustion state [89,102,119].

As a reducing agent, ammonia itself is used in selective non-catalytic reduction (SNCR) of NO, also referred to as thermal deNOx, which is commonly used in the fuel combustion of biomass and waste [96,120]. During ammonia-based SNCR processes, an air or steam stream of ammonia vapor is injected into the flue gas at the desired temperature range, 870–1200 °C, and results in a reaction that converts NOx to nitrogen and water. The injection of ammonia into the flue gas causes a complex of intermediate chain branching reactions that are summarized in the equations below:

2NO + 4NH₃ + 2O₂ → 3N₂ + 6H₂O

(8)

4NH₃ + 5O₂ → 4NO + 6H₂O

(9)

The NOx reduction reaction (first equation) takes place at a temperature of 870–1200 °C (1600–2200 °F) through injection of ammonia only. The reduction in NOx efficiency can be improved even up to 700 °C (1300 °F) through injection of hydrogen (H₂ and NH₃). Nevertheless, injection of NH₃ into the high-temperature flue gas will lead to an increase in the formation of NOx, as shown in the second equation, and is therefore counterproductive to the process. At initial NOx concentrations of 200 ppmvd or lower (NH₃/NOx), molar ratios of 1.5 are usually employed [102,118,119,120].

Recent research on ammonia as a combustion intermediary or SNCR agent for NO reduction (

Table 2. Comparison of typical SNCR and SCR systems [101,120,122].

Design Criteria	SNCR	SCR
NOx reduction efficiency	40–75%	60–90%
Temperature window	870–1200 °C	165–600 °C
Reactant	Ammonia or urea	Ammonia or urea
Reactor	None	Catalytic
Waste disposal	None	Spent catalyst
Thermal efficiency debit	0–0.3%	0%
Energy consumption	Low	* High I.D. fans
Capital investment costs	Low	High
Plot requirements	Minor	Major
Maintenance	Low	3–5 years (typical catalyst life)
Ammonia/NOx (molar ratio)	1.0–1.5	0.8–1.2
Urea/NOx (molar ratio)	0.5–0.75	Not applicable
Ammonia slip	5–20 ** ppmvd	5–10 ppmvd
Retrofit	Easy	Difficult
Mechanical draft	Not required	Required

* Induced draft fans (I.D. fans) often handle air under challenging conditions, such as high temperatures, acidic airstreams, and other extreme exhaust gases. ** ppmvd means parts per million by volume, dry basis.

) focuses on modeling its oxidation and role in thermal deNOx, chemical kinetics-targeted high-temperature chemistry, and oxidation at temperatures of 800–1600 K and higher pressures (>2 MPa) relevant to modern combustion devices [101,121,122]. Considering its roles as an SNCR agent and fuel, ammonia’s conversion chemistry integrates its interactions with NOx and elementary reactions like the hydrogen atom abstraction and cross-reactions in binary fuel systems [60,95].

The effectiveness of SCR and SNCR processes in reducing NOx is sensitive to a few factors that include: flue gas temperature; uniformity of temperature at the reaction zone; time of residence; rate of ammonia/urea injection; ammonia/urea distribution and mixing; initial concentration of NOx; and the geometry of the heater, which influences the positioning of the nozzle and its design. SNCR operates only over a limited temperature range (950–1150 °C), and its performance decreases rapidly beyond that range, with a typical performance of 30–60% NOx collected at optimum power. SCR is catalyst-based and operates at lower temperatures (290–400 °C) and consistently attains 80–95% NOx reduction and will perform better over a wide load range. The two technologies are residence time-sensitive, reagent concentration-sensitive, and mixing quality-sensitive, and the appropriate location of the nozzle and the design of the heater are vital to maximizing the efficiency with minimum slipping of ammonia. As summarized in Table 2, these characteristics include the trade-offs between SNCR and SCR systems in terms of efficiency, costs, and complexity [120,122].

Another NOx reduction technology is NOx traps or NOx absorbers (Figure 26), typically integrating basic oxides like barium oxide, which react with NOx and store it as nitrate in a lean mode [104].

Periodically switching to a rich mode while injecting a small amount of fuel into the exhaust leads to reversal of the reaction, releasing the stored NOx. Finally, a downstream three-way catalyst converts it back to nitrogen and water [119].

Figure 27 is a schematic of a combined lean NOx trap (LNT) and selective catalytic reduction (SCR) aftertreatment system. During lean operation (Figure 27b), NOx is oxidized and stored on the LNT catalyst, while periodic rich operation regenerates the catalyst and produces NH₃ [123]. The downstream SCR catalyst utilizes the generated NH₃ to further reduce NOx slip, resulting in near-complete conversion to N₂ and H₂O conditions [119,123].

In the subsequent catalytic reduction phase, ammonia reacts with NOx, yielding nitrogen and water. While SCR systems are increasingly prevalent in larger vehicles, ongoing efforts are focused on developing systems suitable for smaller diesel vehicles [102,118,119].

Lean NOx trap (LNT) technology, also known as NOx adsorber catalysts, is an effective aftertreatment approach for reducing NOx emissions from lean-burn gasoline and diesel engines operating under highly oxidizing conditions. A typical LNT catalyst comprises precious metals (Pt, Pd, and Rh), a NOx storage component such as BaO, and a high-surface-area support (e.g., Al₂O₃, CeO₂, or ZrO₂) and functions under transient lean–rich cycling. During lean operation, NO—predominant at high combustion temperatures—is oxidized to NO₂ over Pt and stored on BaO as nitrates, with Pt dispersion and particle size strongly influencing NOx uptake. In the subsequent rich phase, injected fuel-derived reductants (CO, hydrocarbons, and H₂) convert the stored nitrates to N₂, thereby regenerating the catalyst. Owing to limitations in conversion efficiency over wide operating conditions, LNT systems are often integrated with selective catalytic reduction (SCR) catalysts to achieve enhanced and more robust NOx control [124]. Additionally, optimizing the combustion conditions, such as temperature, air and ammonia flow rates, and the NO₂/NO ratio, can improve the efficiency of the SCR process by mitigating NOx formation [60,102,119,121,125].

The implementation of high preheating and dilution levels, facilitated by robust internal recirculation, results in a distinctive combustion regime, as shown in Figure 28 [121].

NOx reduction is achieved through ammonia utilization over catalysts (CuO/CeO₂–TiO₂), a low-NOx burner design, air preheating, and configuration adjustments involving reburning exhaust gas, recirculation, low air excess combustion, staging combustion, reduction in flame temperature and stability [60,119,121].

Vanadium-based catalysts have long been employed in ammonia-based selective catalytic reduction (NH₃-SCR) systems to effectively reduce NOx emissions from various stationary sources (Figure 29a,b), including power plants, chemical manufacturing facilities, incinerators, and steel mills, as well as large ships and cars [101,122].

Low-temperature NOx removal is improved by increasing the tungsten content, with a tungsten (10 wt%) catalyst demonstrating the highest efficiency at (GHSV = 60,000 h⁻¹) (Figure 29a). However, beyond 13 wt%, efficiency declines due to reduced dispersibility, aggregation, and crystallization. N₂ selectivity remains stable at ~80–87% at 450 °C, and N₂O emissions are largely unaffected by tungsten content (Figure 29b) [122].

Reactor loading is shown in the kinetic equation as the gas hourly space velocity (GHSV). The base feed refers to the actual volumetric gas flow per hour divided by the catalyst volume in the reactor. In Figure 29a,b, the GHSV is 60,000 h⁻¹; in Figure 29c, it is 52,000 h⁻¹; and in Figure 29c, the relationship is given as a function of temperature.The concentrations shown are based on 1000 ppm NOx, with varying NH₃ amounts [102,118,126].

N₂O forms during low-temperature SCR, and doping MnOx/Ti catalysts with small amounts of Ce and V demonstrates an effective strategy to suppress it, maintaining high NO removal (Figure 30). Ce catalysts enhance NO conversion and shift the optimal activity to lower temperatures, while V markedly inhibits N₂O formation. Mechanistic analysis reveals that N₂O is mainly produced through direct NH₃ oxidation and, more importantly, through reactions between NO and adsorbed NH₃, with the latter dominating below 240 °C. Although the strong redox activity of MnOx/Ti promotes high NO conversion, it also generates abundant NH (ad) species that favor N₂O formation. Introducing V increases Lewis acid sites on the catalyst surface, which effectively suppresses N₂O formation without sacrificing NO conversion, and performs better than Brønsted acid sites in this respect. In Figure 30, the optimized Mn (0.4), Ce (0.1) and V (0.01)/Ti catalyst achieves high NO conversion with minimal N₂O emissions owing to improved surface acidity, a favorable redox balance, and enhanced textural properties at low temperatures [104].

It is important to improve the dispersion properties of the co-catalyst and the main catalyst to enhance the performance. Vanadium-based commercial catalysts are typically used with co-catalysts of tungsten and molybdenum, which increase structural and thermal stability whilst giving sulfur resistance. The disadvantage of molybdenum, however, is that it produces N₂O at high temperatures [126].

Ammonia/urea-based SCR remains the workhorse for NOx control, with urea offering a safer means of handling and delivering ammonia, while relying on a mechanism according to which NHx species are adsorbed on the catalyst surface and gaseous NO (g) reacts with the adsorbed NHx, or both NH₃ and NOx are adsorbed prior to the SCR reactions, or relying on hybrid surface pathways. Recent advances in V₂O₅/TiO₂ and related VOx, Mn-, CeO₂− and zeolite-based catalysts have refined acid and redox cycles, but issues such as N₂O formation and urea-derived byproducts still limit truly “clean” operation [127].

Hydrocarbon SCR, especially using propane and propene, can achieve effective NOx reduction below 400 °C, provided that the catalyst composition and operating variables (C₃H₆/NOx ratio, O₂ level, pH, and preparation route) are carefully tuned, as shown for systems like Cu/SAPO-34 and Mn-zeolites. Hydrogen SCR pushes the temperature window even lower (below 200 °C) and produces only water, yet it demands high H₂/NO ratios and suffers from sensitivity to CO₂/H₂O and side reactions. CO-SCR offers a compelling option for CO-rich exhausts by converting NOx to N₂ with concurrent CO oxidation, and Cu-based catalysts illustrate the promise and the challenge of achieving high low-temperature activity [121,127].

Overall, a deeper understanding of these SCR routes and their reducing agents is central to designing more durable catalysts with fewer byproducts, ultimately enabling cleaner exhaust treatment and supporting air-quality and public-health goals in a practical, scalable way.

4. Conclusions

This study addresses the identification of gaps in understanding of ammonia combustion stability. A systematic examination of stable pure ammonia flameless combustion and emission characteristics has been lacking thus far, the literature on this subject being limited.

Ammonia, as an energy carrier that is free of carbon, has great potential in hydrogen storage and production. However, ensuring efficiency and safety are important elements of combustion stability. Combustion stability refers to the fact that a system can maintain a constant, regulated flame over various operating conditions, which is vital not only in the maximization of energy production and reduced release of detrimental emissions, but also in avoiding such dangers as flameouts and explosions.

Ammonia has low reactivity, a low burning speed, a narrow flammability range, and a high ignition temperature, which are major impediments to attaining stable combustion, resulting in instability and low performance.

Ammonia combustion is drawing major attention as a carbon-free energy pathway, but its practical use faces two main challenges: flame instability and high NOx emissions. Recent studies show substantial progress in understanding and controlling these phenomena through burner design, blending strategies, and chemical/physical approaches.

The stability of ammonia combustion can be influenced by various factors, such as combustion method, flame temperature, pressure, equivalence ratio, flow rate, oxygen concentration, and heat release rate. Besides these, fuel quality, combustion chamber design, and insufficient residence time are also important. The air–fuel ratio is particularly critical—if it is too lean, the flame may fail to propagate smoothly, leading to instability. Real-time indicators such as flame oscillations, pressure fluctuations, and gas composition provide insight into system behavior.

Moreover, the combustion method significantly affects stability, with improvements observed in the progression from laminar flames to diffusion flames and ultimately to flameless combustion, which is stable even under lean conditions. Therefore, precise control and adjustment of the air–fuel ratio are essential for the stability of the combustion procedure. The stability of ammonia flameless combustion technology has multiple benefits over conventional combustion. Flameless combustion exhibits a distinct sound behavior compared to conventional combustion, resembling more of a normal air noise. Another notable feature of flameless combustion is its significantly reduced temperature fluctuations and lower noise generation compared to traditional combustion regimes. It possesses a less concentrated field of temperature, which promises a great alternative for lower NOx pollutant emissions and higher efficiency and economic benefits. This is why it is an applicable technology in the steel, glass and ceramic industries in the industrial sphere.

Ammonia’s low reactivity (laminar burning velocity (LBV) ≈ 0.07–0.15 m/s) and high ignition energy make its flames prone to extinction and blow-off under lean conditions.

The equivalence ratios (ϕ) of pure ammonia–air flames typically stabilize within the range of 0.8–1.2, which is narrower than hydrocarbon ranges. Flame extinction occurs below ϕ ≈ 0.7 under atmospheric pressure, and blow-off limits improve with elevated temperature, pressure, and preheating.

The general tendency of the laminar burning velocity (LBV) decreases with increasing initial pressure. Pressure effects such as high-pressure combustion (5–20 atm) enhance stability and NO reduction simultaneously by accelerating intermediate NHx formation and promoting complete oxidation.

Heat-recirculating systems through Swiss-roll or excess enthalpy burners extend pure ammonia stability limits by promoting thermal feedback, reducing quenching risks under lean conditions.

Ammonia extinction can adversely affect combustion stability. Ensuring efficient and stable combustion necessitates control and management of factors that could lead to flame extinction, such as the air–fuel equivalence ratio.

Adding other fuels such as hydrogen blending by adding 10–40 vol% H₂ widens lean stability from ϕ = 0.7 down to 0.4 and increases burning velocity by up to sixfold, creating self-sustaining operations comparable to methane–air systems.

Altogether, ammonia’s stability window can be broadened through hydrogen enrichment, high-pressure operation, staged combustion, and heat-recirculation designs. These factors can induce alterations in flame temperature and pressure, resulting in fluctuations in combustion and the generation of unsteady, reactive intermediate species.

The stability of ammonia combustion is influenced by NH₂ and other species generated during ammonia molecule decomposition. These species may not be present in the flue gases but can impact the activity and energy efficiency of ammonia combustion.

As oxygen concentration decreases, the activation energy and potential for spontaneous combustion and explosion decrease. Therefore, reducing oxygen concentration proves to be an effective method for inhibiting or even potentially preventing spontaneous combustion. Despite this, excess air poses fewer challenges to efficiency and is safer to operate, although it may lead to increased NOx emissions.

To sum up, combustion stability has a direct influence on the safety, efficacy, and environmental performance of hydrogen production systems that burn ammonia. Flameless combustion technology, in this case, was deemed to be an appropriate technology that can enhance pure ammonia combustion performance and NOx emission characteristics.