Review of Ammonia Oxy-Combustion Technologies: Fundamental Research and Its Various Applications

Abstract

The combustion of ammonia with oxygen presents a promising pathway for global energy transformation using carbon dioxide-neutral energy solutions and carbon capture. Ammonia, a carbon-free fuel, offers several benefits, owing to its non-explosive nature, high octane rating, and ease of storage and distribution. However, challenges such as low flammability and excessive nitrogen oxide (NO_x) emissions must be addressed. This paper explores the recent advances in ammonia oxy-combustion and highlights recent experimental and numerical research on NO_x emission traits, combustion, and flame propagation across various applications, including gas furnaces, internal combustion engines, and boilers. Furthermore, this review discusses the diverse approaches to overcoming the challenges of ammonia combustion, including oxygen enrichment, fuel blending, plasma assistance, preheating, multiple injections, and burner design modifications. By summarizing the advancements in ammonia oxy-combustion investigation, this paper aims to provide valuable insights that can serve as reference information for prospective ammonia oxy-combustion research and applications toward the transition to sustainable energy.

1. Introduction

Excessive reliance on fossil fuels, including propane, natural gas, coal, and petroleum, in the combustion processes of power plants, industrial processes, and transportation is the primary source of CO₂ emissions that significantly contribute to the current level of greenhouse gas (GHG) emissions [1,2]. Accelerating the transition to a reasonable, reliable, and sustainable energy framework aimed at achieving CO₂-neutral energy processes is crucial, as emphasized by the sustainable development goals outlined by the United Nations [3]. The energy industry will face significant challenges and opportunities as global awareness of climate change increases and zero-carbon targets are established. Transitions to eliminate fossil fuel dependence are essential for achieving these goals [4]. The Paris Agreement calls for a collective focus on reducing GHG emissions and climate change impacts by setting carbon neutrality targets [5]. Strategies for mitigating the effects of climate change include implementing an emission trading system, applying carbon credits, and establishing carbon taxes to encourage sustainable practices through innovation and investment in green technologies [6]. A tax of USD 40 per ton of CO₂, which covers 30% of total emissions, is estimated to reduce cumulative emissions by 4–6% [7], as depicted in Figure 1. Additionally, environmental taxes significantly affect green development goals and technological inventions in high- and middle-income countries. By learning from developed countries, developing countries can accelerate their transitions to advanced technologies [8].

The most important strategy to reduce carbon emissions involves utilizing renewable energy. Solar, wind, and hydroelectric energy are the most used renewable sources. However, due to the lack of continuity in terms of reliability, the performance of these renewable energy sources depends on natural conditions. It may not be easy to use directly for power generation because it will require large-scale batteries and challenging long-distance transportation [9]. Therefore, an energy transition is necessary to reduce GHG emissions and even to prevent the damaging effects of reliance on traditional fuels. By embracing green energy production and innovative technologies, the community can rapidly develop alternative energy sources, such as low-carbon fuels, hydrogen, and ammonia [10,11,12]. Carbon-free fuels promise to reduce fossil energy consumption and to contribute significantly to energy security by driving a well-established renewable energy infrastructure that encloses industrial decarbonization, supply chains, and regional development [13].

Among the three examples of carbon-free fuels mentioned above, ammonia has potential applications in the energy sector as a decarbonizing fuel, hydrogen vector, and energy storage medium. It has several desirable properties, such as a hydrogen capacity of 17.8% by mass and a high energy density per unit volume, establishing ammonia as an ideal hydrogen carrier and ensuring safety, ease of storage and transport, and zero carbon emissions by the end of the process [14,15,16]. The production process is relatively straightforward and efficient, and it utilizes the Haber–Bosch (HB) method for ammonia synthesis [17]. Ammonia is used globally as an energy source in gas turbines, industrial furnaces in thermal power plants, and internal combustion engines (ICEs) in marine transportation [18]. Compared with other fuels (hydrogen, methanol, dimethyl ether, gasoline, and diesel), ammonia has an extensive operating range and high power output [19]. Ammonia is the primary carbon-free fuel for decarbonizing in marine transportation. Ammonia and hydrogen contribute over 60% of generated energy, enabling the 2050 Net Zero Emissions scenario [20]. Ammonia can be mixed with other fuels to assist coal-fired power plants in reducing carbon pollution through co-combustion with up to 60% ammonia content in the fuel. The evolution to 50–60% ammonia co-firing in the 2030s was predicted to achieve full ammonia firing by the 2040s [18].

Although ammonia can be used directly or substituted with fossil fuels in combustion systems, several fundamental issues exist because of its low flammability, low laminar burning velocity (LBV), elevated nitrogen oxide (NO_x) emissions, and weak radiation intensity [14]. Furthermore, the presence of water can corrode fuel systems [21]. Research has been conducted to address these issues [22]. The corrosion problems can be addressed using advanced coatings, such as composite silane coatings, or materials resistant to ammonia corrosion at high temperatures, such as 15CrMoG steel [23]. Considerable efforts have focused on comprehending the mechanism of NO_x generation and advances in NO_x emission reduction technologies, such as staged combustion, oscillating combustion, selective catalytic reduction, moderate or intense low-oxygen dilution (MILD), and preheating [2,24].

A widely recognized approach to enhancing the combustion rate of ammonia involves reactivity stratification using a high-reactivity fuel such as hydrogen [25]. Oxygen-enriched combustion is used in combustion systems to tackle issues in the combustion characteristics of ammonia by increasing the LBV of ammonia and the adiabatic temperature due to rising oxygen levels in the combustion process [26]. This process is referred to as oxy-combustion, where pure O₂ is used as a substitute for air through CO₂ recycling [27]. The addition of oxygen affects the combustion system by reducing the mass flow of exhaust gases and increasing the adiabatic temperature, which significantly increases the efficiency [28,29]. Although ammonia oxy-combustion offers considerable advantages, challenges remain, including the boost in NO_x emissions due to the elevated maximum temperature in ammonia combustion [30,31].

The previous paragraph discussed the urgency of emission mitigation, which is a priority to be addressed along with the development of the industry. Along with this development, the solutions offered to suppress GHG emissions are becoming increasingly progressive. As is known, the involvement of ammonia in a combustion system still has several obstacles that need to be addressed. Most recent literature reviews have focused on ammonia combustion characteristics and applications [32,33,34,35,36,37,38,39,40,41,42] or a combination of ammonia and other fuel blends, especially with hydrogen, while often neglecting ammonia oxy-fuel combustion. This paper contributes significantly to recent research developments and provides crucial insights for future ammonia oxygen combustion studies. Currently, a full assessment of energy concerns focuses primarily on ammonia as a carbon-free fuel, as discussed in Section 1. Section 2 briefly summarizes the properties, advantages, and downsides. Section 3 examines the progress in research on ammonia oxy-fuel combustion. Section 4 highlights the characteristics of ammonia oxy-fuel combustion across various applications, and Section 5 highlights the progress and forecasts the prospective research and development patterns of ammonia oxy-fuel combustion.

2. Ammonia Fuel Combustion Characteristics

2.1. Advantages of Ammonia Fuel Combustion

Ammonia has been predicted as an ammonia energy storage system for hydrogen and other effective energy carriers, owing to its high gravimetric hydrogen density of 17.8 wt% [43,44]. Methanol is a strong competitor to ammonia because of its capacity to store significant amounts of energy and ability to facilitate long-distance transmission. Nonetheless, utilizing methanol for hydrogen storage has environmental consequences because it releases CO₂ when used directly or after degradation [45,46]. Meanwhile, ammonia is usually transported in liquid form because of its high density. It can be stored at −33 °C below ambient pressure or at room temperature at approximately 10 bar. Using refrigerants is effective for large-volume storage [15]. The infrastructure for handling and transporting ammonia is well established, enabling efficient global transmission using trucks, trains, barges, vessels, and pipelines [47].

Table 1. Comparison of ammonia with hydrogen, propane, and methane in terms of critical properties [14,50,51,52,53].

Property	Ammonia	Hydrogen	Propane	Methane
Chemical formula	NH3	H2	C3H8	CH4
Octane number	130	–	111	120
Density (kg/m3)	0.73	0.09	2.01	0.66
Molar mass (g/mol)	17.03	2.016	44.096	16.04
Boiling point (°C)	−33.4	−253	−42.1	−161
Adiabatic flame temperature (°C)	1800	2110	2000	1950
Flashpoint (°C)	132	−253	−104	−188
Auto-ignition temperature (°C)	650	520	450	630
Critical pressure (MPa)	11.3	1.29	4.25	4.6
Critical temperature (°C)	132.4	−240.21	96.9	−82.59
Lower heating value (LHV) (MJ/kg)	18.6	120	50	46.4
Higher heating value (HHV) (MJ/kg)	22.5	141.9	50.4	55.5
Explosive limits in air (%)	15–28	4–75	2.37–9.5	4.4–17
Minimum ignition energy (mJ)	8	0.011	0.26	0.28
Maximum laminar burning velocity (m/s)	0.07	2.91	0.43	0.37
Gravimetric hydrogen density (wt%)	17.8	100	18.2	25
Flammability limit (equivalence ratio)	0.63–1.40	0.10–7.1	0.51–2.5	0.50–1.7

provides a comparative overview, along with the chemical kinetics investigation results for the explosive properties of NH₃. Ammonia has a narrow explosive range and the lowest explosive reactivity among fuels such as hydrogen and methane [48]. It can be directly utilized in combustion without generating CO₂ emissions, leading to carbon-neutral energy systems. The reaction of an ammonia–air mixture under stoichiometric conditions can be expressed as follows [49]:

$${\mathrm{N}\mathrm{H}}_3+0.75({\mathrm{O}}_2+3.76{\mathrm{N}}_2)\to 1.5{\mathrm{H}}_2\mathrm{O}+3.32{\mathrm{N}}_2.$$

(1)

Table 1 shows that ammonia has an octane number of 130, higher than those of the other fuels. An engine fueled by ammonia has the potential to reduce the chances of knocking, operate effectively at higher compression ratios, and enhance thermal efficiency [54]. Investigations were conducted in a rapid compression engine to understand the effects of ammonia’s addition on isooctane knock combustion. The observed data imply that the intensity of knocking decreases as the mole fraction of ammonia increases to 0% (A0), 40% (A40), and 80% (A80) while maintaining consistent thermodynamic conditions, as illustrated in Figure 2. The effect of the initial pressure on the knock intensity was noted [55].

More than 98% of the ammonia produced globally in significant amounts is generated using the HB method, which involves the catalytic interaction of nitrogen and hydrogen at high pressures and temperatures [56]. Ammonia can be classified into several types based on its production process. “Brown ammonia” refers to the coal gasification method for producing hydrogen. “Gray ammonia” is ammonia delivered from hydrogen extracted from methane. “Blue ammonia” resembles gray ammonia, but it integrates carbon capture, utilization, and storage technology into its production process. These three production processes use fossil fuels that generate GHG emissions. However, more sustainable ammonia production can be achieved by using renewable energy as an electricity source, namely “green ammonia” [57,58,59]. The HB process is used to synthesize green ammonia by combining H₂ produced through the electrolysis of H₂O with N₂ extracted from the air in an air separation unit [60].

2.2. Challenges in Ammonia Fuel Combustion

Despite its advantages, ammonia utilization presents numerous challenges because of its unique characteristics. Ammonia toxicity is a crucial safety concern because ammonia can be hazardous if leaked. Ammonia poisoning represents the most significant health risk in the event of a leak [61].

Fuel-bond NO_x emissions and the amount of unburned ammonia in the exhaust gas are other critical challenges in ammonia combustion. NO_x has adverse effects on human health by contributing to the formation of acid rain and has substantial GHG consequences [24,62]. NO_x generation during ammonia combustion is a convoluted process controlled by several parameters, including the temperature, pressure, and oxygen ratio in the combustion zone [63]. During the burning process, the nitrogen atoms in the ammonia fuel react with oxygen to assemble NO molecules, which indicates that NO formation can be minimized by limiting the oxygen level available to participate in the reaction in Equation (2). The oxygen-to-fuel ratio decreases as the equivalence ratio increases [64]:

$${\mathrm{N}\mathrm{H}}_i+\mathrm{O}\mathrm{X}\to \mathrm{N}\mathrm{O}+{\mathrm{H}}_i\mathrm{X}$$

(2)

The low combustion efficiency of ammonia as a fuel presents an immediate challenge due to its lower heating value and higher heating value. Moreover, its LBV is significantly lower than those of other hydrocarbons, suggesting that the energy generated from burning ammonia is significantly less intense [65]. Additionally, the high auto-ignition temperature of ammonia (650 °C), which is higher than those of hydrogen (520 °C), propane (450 °C), and methane (630 °C), makes it challenging to ignite and results in low flammability. Nevertheless, the LBV of ammonia can be improved by integrating it with enriched oxygen, owing to the increased reactivity rates of OH, H, O, and NH₂ molecules in the reaction zone at a higher O₂ content [26]. Figure 3 plots the LBV at different O₂ concentrations.

3. Fundamental Research on Ammonia Fuel Combustion

This section is divided into two subsections. Section 3.1 reviews recent trends in experimental techniques, and Section 3.2 discusses modeling using chemical kinetic analysis.

3.1. Experimental Studies

Several studies on flame propagation in ammonia combustion have been conducted to investigate the fuel characteristics. Ichimura et al. [66] investigated the dynamics of turbulent flame propagation during ammonia combustion. The fan-stirred chamber was tested using a mixture of ammonia and air at diverse equivalence ratios (Φ) ranging from 0.6 to 1.3. Figure 4 presents Schlieren images of ammonia–air mixture burning at Φ values of 0.8, 1.0, and 1.2. Figure 5 indicates that a fuel-lean mixture at Φ = 0.9 can maintain propagation at the most remarkable turbulence intensity caused by Lewis’s influence, despite the LBV peaking at a Φ value of approximately 1.1, as shown in Figure 6 [66].

Karan et al. [67] reported experimental data on oxy-ammonia combustion at elevated temperatures and pressures in a constant-volume spherical chamber to evaluate the LBV under controlled conditions. Spherical flames spreading outward in a constant-volume chamber provide a method for measuring the LBV. Tests were performed using ammonia–oxygen mixtures with equivalence ratios of 0.8, 1.1, and 1.3. These ratios represent the lean, stoichiometric, and rich combustion states at an initial temperature of 300 K and pressure of 1–4 bar. The analysis revealed that the LBV obtained by Nakamura and Stagni closely matched the experimental results. The oxygen level in the mixture determines the LBV. Lean mixtures tend to have more efficient combustion processes, leading to LBVs higher than those of rich mixtures, with the maximum LBV observed at Φ = 1.1. Mei et al. [68] also analyzed the LBV of an ammonia/oxygen/nitrogen combustion mixture under oxygen enrichment and pressure consequences. They remarked that oxygen enrichment enhanced flame propagation and reduced buoyancy by increasing the adiabatic flame temperature and radical concentrations of H, OH, and NH₂. Additionally, they found that increased pressure could suppress the LBV, owing to reduced mixture reactivity.

Zhang et al. [69] experimentally and numerically investigated the LBV and emission characteristics of a Bunsen burner under oxy-ammonia combustion conditions (100% oxygen). Figure 7 illustrates the experimental set-up. They aimed to estimate the LBV in the surroundings of a constant-volume combustion bomb by utilizing a Bunsen burner with an output diameter of 6 mm and an area contraction coefficient of 178. This methodology assumes that the LBV remains uniform across the entire surface area of the flame. The LBV can be determined by applying the following mass conservation balance:

$$S_L={\displaystyle \frac{{\rho }_b{A}_b{S}_b}{{\rho }_u{A}_b}}={\displaystyle \frac{Q}{A_b}},$$

(3)

where Q is the total volumetric flow rate, A_b is the intended flame surface area, and ρ_b and ρ_u represent the densities of burned and unburned reactants, respectively. A_b was calculated from the NH* distribution image captured by an intensified high-speed CMOS camera system, which is a better method than the Schlieren technique. The impact of the preheating method was compared across a temperature ($T_u$, the temperature at 50 mm upstream of the outlet) range of 298–520 K and a Φ range of 0.7–1.6. Figure 8 shows that at ambient temperature and pressure, the highest possible LBV was 1.25 m/s, which is approximately 18 times greater than that of NH₃/air combustion, attributed to the accelerated reaction dynamics of OH, H, O, and NH₂ radicals within the reaction zone.

Co-firing ammonia with other fuels, particularly using oxy-combustion technology, is an innovative area of research and application. Okafor et al. [70] reported the LBVs of ammonia and methane in premixed air flames using experimental and computational methodologies. The tests were performed in a cylindrical combustion chamber with a constant volume. The chamber had an inner diameter of 270 mm and a length of 410 mm. Two quartz windows 60 mm in size were installed at opposite sides to provide optical access to the chamber. Schlieren photography was employed to observe and record the flame propagation over the quartz chamber windows. A high-speed digital camera was used while the equivalency ratio was changed from 0.8 to 1.3, and the ammonia concentration varied from 0 to 0.3 based on the heat fraction of the fuel. These findings demonstrate that the unstretched LBV diminished as the ammonia concentration increased. This highlights the significant effect of ammonia on the combustion characteristics of the combination.

Liu et al. [71] presented experimental and modeling findings on the laminar flame speed of a blending fuel of methane and ammonia under oxy-combustion conditions. They explored varying the NH₃/CH₄ ratio from 0.1 to 0.2, the CO₂ mole fraction from 45% to 65%, and the O₂ mole fraction from 35% to 40%, all within a two-burner counter flow arrangement. The bottom burner released a premixed combustible mixture of CH₄, NH₃, O₂, and CO₂, whereas the top burner emitted N₂. These experimental results were compared with the computational results from three chemical kinetic models: the HUST, Okafor, and Mendiara mechanisms. Notably, the Mendiara mechanism consistently yielded results that exceeded the experimental findings. In contrast, the Okafor mechanism tended to underpredict the laminar flame speeds at an NH₃/CH₄ (II) ratio of 0.1 when analyzing O₂/CO₂ mixtures with a 35:65 ratio. Regardless, the HUST mechanism demonstrated its effectiveness as a dependable predictor of all experimental outcomes.

Li et al. [72] inspected the LBV and flame propagation characteristics of two clean-energy blends, namely ammonia and hydrogen, during oxygen combustion under varying conditions. Their findings emphasized the importance of Φ and the initial pressure in defining combustion performance. The analysis demonstrated that the LBVs of the ammonia/hydrogen/oxygen mixtures increased with an increasing fuel ratio. The LBV increased significantly with the hydrogen content in the blend, emphasizing the role of hydrogen in enhancing the LBV. Increasing the hydrogen-to-ammonia ratio improved the flame stability and LBV. The highest LBV was approximately 0.8, which deviated slightly from the stoichiometric ratio, owing to the complex interactions between the flame temperature and diffusion. Li et al. [73] highlighted the flame propagation and inherent instability mechanisms in hydrogen–ammonia–oxygen combustion. When the fuel ratio was gradually increased, the Lewis number (diffusional–thermal instability) shifted from one, and the hydrodynamic instability was limited. In addition, an increase in pressure affected this instability. Moreover, the flame thickness of the hydrogen–ammonia combustion in air was higher than that of the oxygen combustion under the same operating conditions, indicating that the oxygen atmosphere significantly increased the hydrodynamic instability of the hydrogen–ammonia-blended flame.

Shi et al. [74] explored the LBV and combustion characteristics of co-firing reactive hydrocarbons and oxygenated fuels, namely ammonia with dimethyl ether (DME) (CH₃OCH₃), in an O₂/CO₂ atmosphere. The observed LBV of NH₃/O₂/CO₂ increased significantly, exceeding 65 cm/s when γ reached 70%. In contrast, higher oxygen levels significantly increased the LBV propagation rate of NH₃/DME, as shown in Figure 9. Furthermore, the LBV of (NH₃/DME)/(O₂/CO₂) typically increased with the DME concentration, despite its comparatively low diffusivity (Figure 10). Although chemical effects primarily enhance the LBV of NH₃ in an O₂/CO₂ environment with DME co-combustion, thermal and transport effects have a minimal impact.

Barbas et al. [75] conducted experimental and computational investigations on the formation and emission of CO and NO by NH₃ doping during oxy-methane combustion. Under the control conditions, the excess oxygen coefficient was evaluated to be 1.05, 1.1, 1.15, 1.2, and 1.25. The mass flow rate of CO₂ was adjusted accordingly for each excess oxygen coefficient. The tested oxidizer arrangements retained blends of 21% O₂ with 79% CO₂, 23% O₂ with 77% CO₂, 25% O₂ with 75% CO₂, 27% O₂ with 73% CO₂, and 29% O₂ with 71% CO₂, in addition to air. The research revealed that NH₃ significantly contributes to NO production through HNO intermediates, such as $\mathrm{H}\mathrm{N}\mathrm{O}+\mathrm{H}\to {\mathrm{H}}_2+\mathrm{N}\mathrm{O}$, which are essential to the process. During oxy-fuel combustion, NH₃ transforms the nitrogen bound in the fuel into NO. The study also revealed that NO generation is higher for oxidizers with a higher oxygen content in oxy-fuel applications. Nevertheless, it decreases slightly as the excess oxygen coefficient increases. The addition of ammonia to methane combustion under oxy-fuel conditions considerably impacts NO emissions while affecting CO build pathways, owing to altered radical dynamics and competition. A particularly elevated oxygen coefficient increases CO emissions while decreasing NO emissions, as shown in Figure 11 and Figure 12.

Wu et al. [76] provided further insights into the formation of NO from the co-firing of hydroxylated fuels, namely methanol (CH₃OH) and ethanol (C₂H₅OH), with NH₃ oxidation in diverse oxy-fuel atmospheres by simulating the formation and reduction of NO during the oxidation of CH₃OH/NH₃ and C₂H₅OH/NH₃ in O₂/CO₂, O₂/H₂O, and O₂/CO₂/H₂O atmospheres under oxy-fuel conditions. The authors found that the presence of H₂O and CO₂ affected NO generation. In particular, the addition of H₂O enhanced NO reduction more effectively than the addition of CO₂. At higher oxygen-to-fuel ratios, the reduction effects of H₂O and CO₂ were more pronounced and stronger for C₂H₅OH than for CH₃OH.

3.2. Numerical Study

Kinetic modeling was implemented to determine the unstretched LBV, species concentrations, and species production rates. Additionally, it contributed to the sensitivity analysis and emission generation during the combustion process. This method requires advanced process details and optimization. The experimental LBV results were validated to understand the chemical kinetics mechanisms. These mechanisms were utilized in Chemkin-Pro to model the LBV and to compare the modeling results with the experimental data. Various simulations were performed under isobaric and isothermal conditions. The results of these simulations verified that the mechanisms were driven by temperature rather than pressure because the change in LBV increased with the temperature but showed limited sensitivity to pressure shifts beyond 5 bar [67].

The experimental results were further analyzed using the PREMIX and OPPOSED code packages. Figure 13 and Figure 14 illustrate that the LBV exhibits a linear increase with an increasing O₂ concentration or decreasing CO₂ level under specified conditions [71].

In addition to analyzing the propagating flame sub-model, the nitrogen conversion pathway was examined using Chemkin-Pro. The analysis revealed that NO emissions from NH₃/O₂ combustion under stoichiometric conditions were approximately 7216 ppmv at 15% O₂. This level was approximately 1.5 times higher than that of NH₃/air combustion and 13 times greater than that of CH₄/air combustion. As illustrated in Figure 15, the mole fraction and N-normalized mole fraction of NO increased with Φ, and preheating also contributed to an increase in NO emissions [69].

Okafor et al. [70] conducted simulations to replicate experimental results and to analyze the LBV and species concentration. Various detailed combustion mechanisms were utilized for the numerical simulations, including GRI Mech 3.0, Mendiara Mech, and Tian Mech. The study revealed that the Tian Mech mechanism significantly underestimated the unstretched LBV, particularly at lower ammonia concentrations. In contrast, GRI Mech 3.0 provided predictions closer to the experimental results but lacked some crucial ammonia oxidation steps for high ammonia content flames. A new detailed chemical kinetic model was developed to address these limitations by combining elements from GRI Mech 3.0 and Tian Mech. This new model includes 59 species and 356 elementary reactions. The model was validated against experimental data and demonstrated improved accuracy in predicting the LBV and NO concentration in flames. Li et al. [72] highlighted the use of ammonia–hydrogen blends, stating that although the LBV increased with a higher hydrogen ratio, NO_x emissions, primarily NO emissions, decreased under rich combustion conditions. Figure 16 illustrates the mole fractions of reactants and products in a hydrogen/ammonia/oxygen flame under different Φ values, providing significant insights. H₂, NH₃, and O₂ were the primary reactants, whereas H₂O, N₂, and NO are the primary products. NH₃ was consumed the fastest among the reactants, and the concentrations of H₂ and O₂ stabilized after a certain amount was consumed. The most significant NO_x-producing substance in the mixed fuel was NO. The amount of NO produced decreased gradually with increasing Φ values. A higher Φ value reduced NO emissions, achieving maximum NO reduction under fuel-rich conditions. This characteristic makes ammonia–hydrogen blends promising carbon-free fuel alternatives for industrial and energy applications.

Table 2 summarizes the research developments through experimental methods and chemical kinetics simulations to describe the ammonia oxy-fuel combustion process in the above research.

Table 2. The recent literature on experimental and kinetic modeling studies of ammonia oxy-fuel combustion.

Authors	Research Method	Combustion Conditions	Approach (Injector Equipment, Methods, etc.)	Key Findings
Ichimura et al. [66]	Experiment	Initial pressure and temperature of 1 atm and 298 K, respectively Φ = 0.6–1.3 Ammonia air combustion	Constant-volume chamber Schlieren technique	The maximum intensity of turbulence occurred at Φ = 0.9. The highest LBV was achieved at Φ = 1.1.
Karan et al. [67]	Experiment and simulation	Initial pressure and temperature of 1–4 bar and 300 K, respectively Φ = 0.8, 1.1, and 1.3 Ammonia oxy-combustion	Constant-volume spherical chamber CMOS technique	The maximum LBV occurred at Φ = 1.1. The variation in LBV increased significantly with temperature.
Mei et al. [68]	Experiment and simulation	Initial pressure and temperature of 1–5 atm and 298 K, respectively Φ = 0.6–1.6 Ammonia/O2/N2 (oxygen concentrations of 25–45%) Premixed combustion	A high-pressure, constant-volume cylindrical combustion vessel Schlieren technique	Increased oxygen content led to a higher LBV while minimizing buoyancy effects. The LBV decreased with increasing pressure.
Zhang et al. [69]	Experiment and simulation	Preheating temperature (298–520 K) Φ = 0.7–1.6 Ammonia oxy-combustion	Bunsen burner Constant-volume chamber ICMOS technique	The LBV of a flame was 1.25 m/s, which is approximately 18 times that of NH3/air combustion. NO emissions increased with Φ, and preheating increased.
Okafor et al. [70]	Experiment and simulation	Initial pressure and temperature of 0.10 MPa and 298 K, respectively Φ = 0.8, 1.1, and 1.3 NH3, CH4, air combustion	Constant-volume chamber Schlieren technique	The LBV decreased as the ammonia concentration increased. GRI Mech 3.0 and the Tian Mech model improved the accuracy in predicting the LBV and NO concentration in flames.
Liu et al. [71]	Experiment and simulation	Initial pressure and temperature of 1 atm and 300 ± 2 K, respectively NH3/CH4 ratio: 0.1–0.2; CO2 mole fraction: 45–65%; O2 mole fraction: 35–40% CH4/NH3/O2/CO2 combustion	Two-burner counterflow DPIV technique	The HUST mechanism was considered practical for estimating the LBV. The LBV increased significantly with higher O2 or decreased CO2 levels.
Li et al. [72]	Experiment and simulation	Initial pressure of 0.1–1.0 atm Φ = 0.5–1.5 Fuel ratio: 0.5–2 NH3/H2 oxy-combustion	Cylindrical constant-volume chamber Z-arranged Schlieren technique	The maximum flame was achieved near Φ = 0.8. Maximum NO reduction was reached under fuel-rich conditions.
Li et al. [73]	Experiment	Initial pressure of 0.1–1.0 atm Φ = 0.5–1.5 Fuel ratio: 0.5–2 NH3/H2 oxy-combustion	Cylindrical constant-volume chamber Z-arranged Schlieren technique	The combustion in oxygen significantly enhanced the hydrodynamic instability.
Shi et al. [74]	Experiment and simulation	Initial pressure and temperature of up to 5 atm and 373 K, respectively Φ = 0.7–1.5 NH3/DME in O2/CO2 combustion	Constant-volume chamber Schlieren technique	Oxygen enrichment accelerated the LBV propagation. The LBV of the (NH3/DME)/(O2/CO2) mixture demonstrated a notable increase as the proportion of DME rose.
Barbas et al. [75]	Experiment and simulation	Operating at atmospheric pressure Excess oxygen coefficients (1.05, 1.1, 1.15, 1.2, and 1.25) NH3 doped in CH4 oxy-combustion	Water-cooled burner	NH3 significantly contributed to producing NO via HNO intermediates. A specific excess oxygen coefficient led to higher CO emissions and lower NO emissions.
Wu et al. [76]	Simulation	Oxygen concentration: 30% Oxygen-to-fuel ratio α: 0.8 and 1.2 CH3OH/C2H5OH/NH3 in oxy-combustion	Plug flow reactor	The NO decrease was more significant at high oxygen-fuel levels and with C2H5OH instead of CH3OH.

Table 2. The recent literature on experimental and kinetic modeling studies of ammonia oxy-fuel combustion.

Authors	Research Method	Combustion Conditions	Approach (Injector Equipment, Methods, etc.)	Key Findings
Ichimura et al. [66]	Experiment	Initial pressure and temperature of 1 atm and 298 K, respectively Φ = 0.6–1.3 Ammonia air combustion	Constant-volume chamber Schlieren technique	The maximum intensity of turbulence occurred at Φ = 0.9. The highest LBV was achieved at Φ = 1.1.
Karan et al. [67]	Experiment and simulation	Initial pressure and temperature of 1–4 bar and 300 K, respectively Φ = 0.8, 1.1, and 1.3 Ammonia oxy-combustion	Constant-volume spherical chamber CMOS technique	The maximum LBV occurred at Φ = 1.1. The variation in LBV increased significantly with temperature.
Mei et al. [68]	Experiment and simulation	Initial pressure and temperature of 1–5 atm and 298 K, respectively Φ = 0.6–1.6 Ammonia/O2/N2 (oxygen concentrations of 25–45%) Premixed combustion	A high-pressure, constant-volume cylindrical combustion vessel Schlieren technique	Increased oxygen content led to a higher LBV while minimizing buoyancy effects. The LBV decreased with increasing pressure.
Zhang et al. [69]	Experiment and simulation	Preheating temperature (298–520 K) Φ = 0.7–1.6 Ammonia oxy-combustion	Bunsen burner Constant-volume chamber ICMOS technique	The LBV of a flame was 1.25 m/s, which is approximately 18 times that of NH3/air combustion. NO emissions increased with Φ, and preheating increased.
Okafor et al. [70]	Experiment and simulation	Initial pressure and temperature of 0.10 MPa and 298 K, respectively Φ = 0.8, 1.1, and 1.3 NH3, CH4, air combustion	Constant-volume chamber Schlieren technique	The LBV decreased as the ammonia concentration increased. GRI Mech 3.0 and the Tian Mech model improved the accuracy in predicting the LBV and NO concentration in flames.
Liu et al. [71]	Experiment and simulation	Initial pressure and temperature of 1 atm and 300 ± 2 K, respectively NH3/CH4 ratio: 0.1–0.2; CO2 mole fraction: 45–65%; O2 mole fraction: 35–40% CH4/NH3/O2/CO2 combustion	Two-burner counterflow DPIV technique	The HUST mechanism was considered practical for estimating the LBV. The LBV increased significantly with higher O2 or decreased CO2 levels.
Li et al. [72]	Experiment and simulation	Initial pressure of 0.1–1.0 atm Φ = 0.5–1.5 Fuel ratio: 0.5–2 NH3/H2 oxy-combustion	Cylindrical constant-volume chamber Z-arranged Schlieren technique	The maximum flame was achieved near Φ = 0.8. Maximum NO reduction was reached under fuel-rich conditions.
Li et al. [73]	Experiment	Initial pressure of 0.1–1.0 atm Φ = 0.5–1.5 Fuel ratio: 0.5–2 NH3/H2 oxy-combustion	Cylindrical constant-volume chamber Z-arranged Schlieren technique	The combustion in oxygen significantly enhanced the hydrodynamic instability.
Shi et al. [74]	Experiment and simulation	Initial pressure and temperature of up to 5 atm and 373 K, respectively Φ = 0.7–1.5 NH3/DME in O2/CO2 combustion	Constant-volume chamber Schlieren technique	Oxygen enrichment accelerated the LBV propagation. The LBV of the (NH3/DME)/(O2/CO2) mixture demonstrated a notable increase as the proportion of DME rose.
Barbas et al. [75]	Experiment and simulation	Operating at atmospheric pressure Excess oxygen coefficients (1.05, 1.1, 1.15, 1.2, and 1.25) NH3 doped in CH4 oxy-combustion	Water-cooled burner	NH3 significantly contributed to producing NO via HNO intermediates. A specific excess oxygen coefficient led to higher CO emissions and lower NO emissions.
Wu et al. [76]	Simulation	Oxygen concentration: 30% Oxygen-to-fuel ratio α: 0.8 and 1.2 CH3OH/C2H5OH/NH3 in oxy-combustion	Plug flow reactor	The NO decrease was more significant at high oxygen-fuel levels and with C2H5OH instead of CH3OH.

Both the experimental and simulation research methods reviewed provide important insights into the combustion characteristics of ammonia oxy-combustion, namely the flame stability, LBV, and NO_x formation characteristics. The experimental approach offers valuable insights into real flame observation, but this method is limited because it is only usable at small scales and has high costs. On the other hand, modeling approaches such as chemical kinetic methods or computational fluid dynamics (CFD) enable broader and more detailed parametric studies on a larger scale. Both methods have their respective advantages. Experiments have advantages in flame validation and chemical reactivity processes, whereas numerical analysis can predict and optimize combustion processes such as flame characteristics and emission production. Okafor et al. [70] and Li et al. [72] combined experimental and modeling methods and showed significant improvements in the accuracy of kinetic mechanisms and NO formation.

4. Ammonia Oxy-Fuel Combustion in Various Applications

Numerous studies, including experimental and numerical analyses like CFD investigations, have assessed the feasibility of employing ammonia energy in oxygen combustion. The numerical CFD method is widely used because it enables more complex analyses and can be applied to larger or even industrial-scale systems, unlike experiments or other numerical methods, which are typically limited to simpler scales and analyses. CFD enables tasks like optimizing or redesigning combustion processes on a much larger scale. CFD can produce high-precision results by employing a range of numerical equations. This section is divided into three subsections. Section 4.1, Section 4.2 and Section 4.3 review the trends in applying ammonia oxy-fuel combustion in gas furnaces, ICEs, and boilers, respectively, with approaches demonstrated according to the actual conditions. The end of each section summarizes the reviewed literature studies, presented in Table 3, Table 4 and Table 5. This paper aims to evaluate the research progress and assess the impacts of various methods on the characteristics of ammonia oxy-fuel combustion.

4.1. Ammonia Oxy-Fuel Combustion in Gas-Fired Furnaces

This subsection focuses on the research on combustion performance and emissions associated with ammonia oxy-combustion within gas furnaces to evaluate the research progress and analyze the advantages and disadvantages of the various methods employed in recent research, the summary is provided in Table 3. El-Adawy and Nemitallah [77] conducted a relevant study on gas furnace conditions and investigated the flame combustion performance and NO_x emissions against the inlet mixture temperature and velocity of ammonia oxy-combustion in a gas turbine combustor. They used a flamelet-generated manifold model to deliver an efficient and accurate simulation of complex combustion phenomena, enabling analysis of the temperature distribution, flame dynamics, and emissions. The initial temperature increased with the inlet mixture temperature under stoichiometric conditions because it determined the flame position. In addition, the turbulent flames of NH₃ and O₂ increased with the inlet mixture velocity, indicating that a higher inlet velocity can lead to enhanced mixing and combustion efficiencies. Meanwhile, the impact of the inlet velocity (V_in) on the NO_x emissions was minimal.

Cai et al. [78] systematically analyzed the thermal performance and NO emissions during ammonia and oxygen burning in a microplanar combustor. They examined the effects of various parameters, including Φ, the inlet pressure (P_in), and the inlet temperature (T_in), which affect the outer wall mean temperature (OWMT). The results indicated that NO production is susceptible to Φ. Specifically, decreasing Φ increased the NO emissions due to higher flame temperatures and oxygen atom concentrations. Optimizing Φ can result in drastically lower NO concentrations. T_in was essential in influencing the production of OWMT and NO. Increasing T_in decreased the concentrations of OWMT and NO. By contrast, an increase in P_in suggests that the OWMT also increases, owing to the intensified mixing process that drives the flame downstream at high temperatures. However, less NO was produced during combustion with increasing P_in values.

Zhang et al. [79] presented a concept for non-premixed ammonia/oxygen combustion in a dual-inlet microfluidic system. In their study, a 2D numerical model was used to compare premixed combustion using one channel and non-premixed combustion using two channels, in which oxygen and ammonia entered through different channels into the combustion chamber. Figure 17 shows that the non-premixed flame prevents flashback, particularly when NH₃ and O₂ are introduced through separate channels. This strategy improved the mixing and combustion stability, optimized the thermal performance, and reduced emissions. The aforementioned paper also discussed the impact of Φ on the combustion performance. The authors concluded that Φ = 0.9 produced the most optimal thermal performance and emission characteristics. Figure 18 and Figure 19 show that higher inlet velocities increased the wall temperature and reduced NO emissions, indicating that enhanced convective heat transfer improved the thermal performance of the combustor.

Cafiero et al. [80] examined the properties and NO_x emissions of ammonia and oxygen combustion within a 20 kW semi-industrial furnace operating under MILD conditions. The compliance measure in the study involved varying the O₂ enrichment, increasing the operational range from 21% to 50%, and controlling two oxidizer injectors to achieve average inlet velocities of 150 m/s for ID16 and 80 m/s for ID20. The authors observed enhanced reactivity in NH₃ combustion due to O₂ enrichment under conventional and MILD conditions. Nonetheless, higher O₂ levels augmented NO_x emissions. Optimizing O₂ enrichment with MILD combustion could minimize NH₃ destruction while controlling NO_x. O₂ enrichment of 37% was identified to be effective for MILD combustion, significantly reducing the ammonia drop while maintaining the NO_x emissions at an allowable level of 146 ppmvd. Moreover, the effect of the oxidizer velocity for two injector sizes (ID16 and ID20) was tested to determine its impact on the inlet combustion. An elevated inlet oxidizer velocity stabilized the combustion and affected the temperature distribution. NO_x emissions were lower with ID16 than with ID20 across all O₂ levels at the optimal Φ value of 0.95, which achieved the maximum temperature as the optimal compromise between NO_x emissions and ammonia slip. At Φ = 1.01, the minimum NO_x emissions were achieved, whereas the NH₃ slip was higher than the minimum limit that could be assessed during the experiment, reaching 4000 ppmvd before any corrections. These findings align closely with those obtained by Sato et al. [81], who found that unburnt ammonia became significant at Φ ≥ 1.05.

Plasma assistance is a strategy for improving ammonia ignition and flame propagation under challenging combustion conditions. This research group showed how plasma-enhanced systems can reduce ignition delay and stabilize combustion, which is critical for practical applications. Matveev et al. [82] conducted research on the plasma-assisted burning of ammonia in air and oxygen atmospheres. They utilized ANSYS Chemkin to determine the kinetic temperature and concentration of the combustion product components with pressure variations (1–10 bar) and an excess oxidizer coefficient. This technique uses plasma from a radio frequency plasma torch to stabilize a mixture of ammonia and oxygen, thus efficiently enhancing the combustion process. This stability is achieved by the formation of O, H, and OH radicals while reducing the ignition delay. Extending the combustion pressure reduces NO_x emissions during the ammonia combustion process in an air–oxygen setting. Matveev and Serbin [83] conducted follow-up research on plasma-assisted ammonia and oxy-combustion, using 3D modeling to analyze the combustion chamber. Their examination focused on the combustion properties of ammonia. The model employed the eddy dissipation concept, which integrates detailed chemical reactions from the Arrhenius equation with the turbulence in the flame. These findings indicate that the plasma torch is an essential starting point for ammonia ignition. The combustion remained constant even when the oxygen flow rate increased. Predominant NO formation occurred in the central region of the combustion chamber, where ammonia combustion was more intense and the highest temperature was achieved. In contrast, NO₂ was produced in areas with lower temperatures.

Advanced burner redesigns, such as bluff body designs, have been explored to assess their effectiveness in stabilizing the flame, modifying the temperature profile, and affecting pollutant formation. This deconstruction provides insights into how burner design affects ammonia combustion performance under oxygen fuel conditions. Cai et al. [84] investigated the incorporation of a bluff body within a microplanar combustor by utilizing premixed ammonia and oxygen. They focused on the effects of thermal efficiency and NO_x emissions, as depicted in Figure 20. They employed a 3D model integrating flow dynamics, combustion processes, thermodynamics, and heat transfer. The SIMPLE mode couples the pressure and velocity to evaluate the impact of a bluff body on the combustor’s performance. This analysis included an exploration of the dimensionless (h) influence and the bluff body dimensionless location (l) effect. The findings showed that the presence of a bluff body enhanced the outer wall temperature (OWT) around it, unlike in the absence of a bluff body. Regardless, the temperature at the combustor outlet tended to be lower in the presence of the bluff body. Bluff body application can also reduce NO formation under certain conditions because the ammonia flow rate is adjusted to three distinct values, as depicted in Figure 21. NO production had minimal variation at low fuel flow rates of ≤1000 mL/min. Regardless, at a flow rate of 1200 mL/min, the bluff body caused a notable reduction in the NO concentration. The presence of the bluff body significantly affects the temperature field distribution. The flow field and temperature profile show differences without the bluff body, especially around the bluff body region. Figure 22 implies that, compared to the case without the bluff body, the flame temperature downstream of the bluff body along the flow direction (axial) is substantially reduced. This decrease effectively reduced the rate of NO formation. Increasing h increases the OWT but can also inhibit NO formation in certain instances. However, a higher bluff body height increases the pressure loss. Increasing l can increase the OWT and reduce low NO concentrations, probably because a larger l value intensifies the heat transfer between the blend and the lower surface of the combustor.

The examination of NO_x emission performance in ammonia–oxygen premixed combustion with a new design in a microplanar burner that adds a perforated plate was introduced by Cai et al. [85], with the design shown in Figure 23. They used a 3D model validated with experimental data incorporating 22 species and 67 reactions to explore the flow area, impacts of preferential diffusion, and conjugate heat transfer. Several critical findings emerged from this study. Perforated plates expanded the downstream recirculation zone, lowered the temperature and flame rate, and decreased NO formation in the recirculation zone. The primary factors affecting NO_x emissions included the flow dynamics, conjugate heat transfer, and preferential diffusion. Conjugate heat transfer analysis considers the interaction between the combustion gases and the burner wall. This is essential because it affects the location of the flame and the rate of chemical reactions, thereby affecting NO formation. The quality of the burner wall material significantly affects the heat transfer, flame containment, and NO_x output. Three types of materials were selected: quartz, steel, and nickel. Materials exhibiting high thermal conductivities are particularly effective in facilitating heat transfer among solid products during combustion.

Bektaş et al. [86] dissected the capacities of methane and ammonia under air–oxygen combustion conditions against the influence of different equivalence ratios in a two-inlet cyclone combustion chamber. They revealed that ammonia–oxygen combustion performed better at 1828 K, compared with air–methane at 1538 K or air–ammonia at 994 K. Employing oxygen as an oxidizer enhanced NO_x emissions, although fuel-rich conditions had the potential to mitigate NO emissions. Cyclone combustors have prominent characteristics for assembling a strong swirling flow, which can improve the mixing stability and long residence times within the combustion chamber, ultimately leading to more efficient combustion. As shown in Figure 24, during the air–ammonia combustion process, the maximum axial velocity was achieved while maintaining consistent thermal power across all instances. This finding implies that a higher flow rate is introduced into the combustor to correspond to an identical thermal power because ammonia has a lower calorific value than methane. The ideal LBV for ammonia is achieved under fuel-rich conditions, specifically at an equivalence ratio of Φ = 1.1 when using oxy-ammonia. However, combustion that can create a more stable flame condition occurs under stoichiometric conditions. According to a report by Davies et al. [87], unburned NH₃ emissions contribute to flame instability, where Φ > 1.1 causes the amount of unburned NH₃ to increase so much that the combustion efficiency cannot be calculated. Therefore, combustion under fuel-rich conditions should be reconsidered because it increases the risk of incomplete combustion, leading to hydrocarbon emissions from ammonia slip.

The studies reviewed in this subsection investigated various fuel-blending and MILD combustion techniques to overcome the low laminar combustion velocity of ammonia and its high potential for NO_x emissions. This approach is critical to developing optimal and efficient ammonia combustion systems. Zhao et al. [88] studied CH₄/NH₃ combustion with MILD oxy-combustion technology to improve the thermal efficiency, carbon technology, and NO_x reduction in a jet flame in a hot co-flow (JHC) combustor. Their initial study used 2D modeling to examine how CO₂ and H₂O affect NO formation. The simulation utilized an in situ adaptive tabulation algorithm to calculate the efficiency and stability of the complex heterogeneous chemical kinetics. This study revealed four major NH₃-to-NO conversion routes with HNO as the central intermediate and indicated that the significant addition of CO₂ facilitated NO formation through reactions involving NH and NH₂ radicals. H₂O also contributed to NO formation by increasing the number of OH radicals. Compared with the addition of CO₂, the addition of H₂O led to a smaller increase in the N conversion rate. As illustrated in Figure 25, the emission trends of the three dilutions differed, with MILD-N₂ exhibiting the highest emissions followed by disused MILD-CO₂, whereas MILD-H₂O achieved the lowest emissions.

In a subsequent study, Zhao et al. [89] investigated the effects of O₂ co-flow (X_O2) and temperature (T_cof) on the combustion system and fuel NO formation in a CH₄/NH₃ jet diffusion flame utilizing hot O₂/CO₂ co-flow. The differences in X_O2 ranged from 3% to 30%, whereas T_cof ranged from 1300 to 2100 K. The findings revealed that the MILD strategy was attainable when X_O2 ≤ 21% for all T_cof levels but required a higher T_cof for higher X_O2 values (e.g., T_cof ≥ 1500 K for X_O2 = 24%). However, increasing the maximum temperature above the auto-ignition temperature led to a transition of the combustion regime to high-temperature combustion. Increasing X_O2 and T_cof changed the radicals (H, O, and OH), reducing the O radicals, which shifted the NO formation route. Increasing the availability of radicals enabled efficient NH₃ oxidation and NO reduction through recombustion, thereby decreasing the NO emission index.

Zhao et al. [90] continued their research, focusing on investigating the mechanism of fuel and NO formation during MILD oxy-combustion of CH₄/NH₃ under different diluent environments: N₂, CO₂, and water vapor. The analysis revealed that MILD-N₂ combustion had the highest temperature and most intense reaction zone that caused NO emission. In contrast, in MILD-CO₂ combustion, the flame temperature decreased, and the flame zone expanded because the CO₂ absorbed heat. MILD-H₂O combustion produced a slightly higher flame temperature than CO₂ but yielded a minor reaction zone because H₂O enhanced the oxidation process by generating OH radicals.

Blending ammonia oxy-combustion with other fuels is highly desirable in gas combustors to achieve high fuel activity. Cai and Zhao [91] analyzed the micro-power system by investigating the impact of changing the fuel composition ratio (ε) and the hydrogen molar fraction in the ammonia–hydrogen mixture. In addition, they examined the influence of the wall thermal conductivity (WTC) on the thermal efficiency and NO_x pollution. The examination revealed that adding hydrogen reduced the NO concentration and changed the flame position. Figure 26 depicts the change in radiation efficiency (RE) with and without H₂ inclusion, clarifying that a higher ε ratio led to a reduced RE value. Nevertheless, an increased fuel flow rate resulted in an upward trend in the RE. However, beyond a certain point, any further increase caused a decline in RE. Hydrogen addition also impacted the thermal performance, reducing the OWT, which correlated with decreased NO_x emissions. Further analysis showed that a high WTC could improve the OWT at low flow rates. A high WTC also produced a more consistent temperature distribution across the outer wall.

Luo et al. [92] explored ammonia, hydrogen, and oxygen in premixed combustion within a segmented nozzle micro-combustor in a thermophotovoltaic system, as shown in Figure 27. This technique aims to enhance the stability of ammonia combustion and control NO_x emissions. The numerical model uses the pressure-based COUPLE algorithm because it accurately represents the interaction between the flame and wall, enabling more efficient convergence. Figure 28 shows that under identical inlet velocity conditions, the average wall temperature of the segmented structure exceeds that of the non-segmented structure. In addition, the segmented channel produces lower NO emissions than the non-segmented channel. As the hydrogen concentration increased from 10% to 30%, the efficiency of the burner radiation improved by 13.5%. However, this increase in hydrogen led to higher NO emissions. This study also revealed specific combustion conditions that minimize NO emissions while maintaining high RE. For instance, under Φ H₂ = 1.0 and Φ NH₃ = 1.2, the NO emissions could be diminished by up to 13.3% while maintaining RE above 60%.

Sun et al. [93] performed a 3D numerical analysis of a rotating detonation engine fueled by a blend of ammonia/hydrogen and oxygen-enriched air. The detonation wave structure, propagation mode, burner performance, and emission formation were analyzed under different hydrogen concentrations and equivalence ratios. Figure 29 depicts the temperature and pressure fields of the combustion chamber as the hydrogen concentration increased, demonstrating that an increase in the hydrogen concentration increased the temperature, pressure, and velocity. Meanwhile, the elevation of the detonation pulse was expected to decline, but an increase in the detonation velocity would increase the fuel consumption. The cycle efficiency is calculated as follows:

$$\eta =L_E/Q_{in},$$

(4)

$$Q_{in}=m_{h2}{H}_{h2}+m_{nh3}{H}_{nh3}$$

(5)

where m_h2 is the hydrogen mass flow rate and m_nh3 denotes the ammonia mass flow rate. Hydrogen and ammonia have low calorific values: H_h2 and H_nh3. The addition of hydrogen generates a constant detonation wave, thereby enhancing the cycle efficiency. However, it does not necessarily improve the effectiveness. As observed in Figure 30, the optimal cycle efficiency was attained when the hydrogen level was maintained at 0.3, with the Φ value remaining below 1. Furthermore, the NO_x emission was primarily influenced by the equivalency ratio of the fuel mixture instead of the hydrogen concentration, and in the range of Φ = 0.7–1.4, the NO emissions significantly decreased.

Ilbas et al. [94] reviewed the combustion characteristics and NO_x emissions of ammonia/kerosene fuel under air and pure oxygen combustion conditions in a swirl gas turbine combustor. The ammonia composition was determined for a blend of 70% kerosene and 30% ammonia in increments of 10% based on the heat fraction. Figure 31 shows no significant change in the maximum temperature of the kerosene fuel when ammonia was added to the combustor. However, when pure oxygen was used instead of air, the temperature of the combustion chamber significantly improved. Owing to the thermal NO_x mechanism, the high-temperature zone drove NO_x formation during oxygen combustion. However, the predicted NO_x emission levels at the combustion chamber’s exit were not significantly elevated.

Table 3. Recent research applications of ammonia oxy-combustion in a gas furnace.

Authors	Research Method	Combustion Conditions	Approach Equipment	Key Findings
El-Adawy and Nemitallah [77]	Computational fluid dynamics (CFD) simulation	Tin = 298–498 K Vin = 5.2–9.0 Φ = 0.6–1.2 Ammonia oxy-combustion	Swirl gas turbine combustor	Flamelet-generated manifold models for complex turbulent combustion. Higher Tin and peak combustion temperatures. Higher Vin resulted in adequate mixing and homogenous temperature profiles.
Cai et al. [78]	CFD simulation	Initial pressure of 1–2 atm Initial temperature of 300–500 K Φ = 0.8–1.2 Ammonia oxy-combustion	Micro-planar combustor	Φ = 0.9 yielded the optimal heat transfer performance. Increasing Tin decreased the outer wall mean temperature (OWMT) and NO. Increasing Pin increased the OWMT but reduced NO formation.
Zhang et al. [79]	CFD simulation	Initial pressure and temperature of 1.0 atm and 300 K, respectively Φ = 0.6–1.1 Velocity: 0.1–1.1 m/s Non-premixed combustion	Micro-planar combustor	Non-premixed combustion prevented flashback and provided improved heat and emission efficiency. Φ = 0.9 yielded the optimal heat transfer performance. Increasing the velocity increased the temperature.
Cafiero et al. [80]	Experiment	Inlet velocity: 80–50 m/s Φ = 0.9–1.01 Ammonia oxy-combustion Oxygen concentration: 21–50%	20-kW semi-industrial-scale furnace	O2 enrichment of 37% is optimal for moderate or intense low-oxygen diffusion (MILD) combustion. A higher inlet oxidizer velocity stabilized the combustion and temperature distribution. At Φ = 0.95 and the maximum temperature, the optimal compromise between nitrogen oxide (NOx) emissions and ammonia slip can be achieved.
Matveevet al. [82]	Chemkin simulation	Average temperature at the plasma torch outlet: 4950 K RF plasma torch frequency: 3–7 MHz Pressure: 1–10 bar Fuel and oxidizer temperatures: 288 K and 409 K, respectively	Perfectly stirred reactor	Plasma assistance significantly stabilized the combustion. Increasing the pressure in combustion decreased the NOx emissions.
Matveevet al. [83]	CFD simulation	Pressure: 0.3 MPa Initial flow rate of ammonia: 53.76 g/s; temperature: 288 K Plasma oxygen flow rate: 3 g/s; average temperature: 4990 K Oxidizer flowrate: 400–720 g/s; temperature: 409 K	Radial swirl combustion chamber	The eddy dissipation concept model was implemented. Plasma assistance trigger factor in ammonia ignition. NO formed in a central zone, whereas NO2 was produced at reduced temperatures.
Cai et al. [84]	CFD simulation	Initial temperature: 300 K Flow rate: 600–1200 mL/min h = 1/5–4/5 l = 1/5–4/5 Ammonia oxy-premixed combustion	Micro-planar combustor with a bluff-body	A bluff body can increase the outer wall temperature (OWT) and reduce NO generation. The ideal dimensions of the bluff body are h = 4/5 and l = 4/5, which optimize the maximum OWT while minimizing the NOx emissions.
Cai et al. [85]	CFD simulation	Initial pressure and temperature: 1.0 atm and 300 K, respectively Φ = 0.8–1 Ammonia oxy-premixed combustion	Micro-planar combustor with perforated plates	Perforated plates reduced the LBV and NO formation. High-thermal-conductivity materials were effective in conjugate heat transfer.
Bektas et al. [86]	CFD simulation	Φ = 1.1–1.68 Ammonia oxy- and air-premixed combustion Methane air-premixed combustion	Two-inlet cyclone combustor	The performance in oxy-ammonia combustion was outstanding compared with that in air–methane and air–ammonia combustion. The maximum flame occurred at Φ = 1.1 for oxy-ammonia.
Zhao et al. [88]	CFD simulation	Fuel inlet: CH4/NH3 at 39.9 m/s with a temperature of 305 K Co-flow inlet: 3.2 m/s with a temperature of 1700 K XCO2 = 0–41% XH2O = 0–50% MILD-oxy condition	Jet flame in hot co-flow (JHC) burner	Increasing the CO2 and H2O contents increased the formation of NO.
Zhao et al. [89]	CFD simulation	Fuel inlet: CH4/NH3 Re 9482 (39.9 m/s) and temperature 305 K Co-flow inlet: 3.2 m/s and temperature 1300–2100 K XO2 = 3–30% XCO2 = 97–70% MILD-oxy condition	JHC burner	MILD combustion occurred at lower XO2 levels or higher Tcof levels. The NO emission index decreased as XO2 and Tcof increased.
Zhao et al. [90]	CFD simulation	Fuel inlet: (CH4/NH3) Re 39.9 m/s and temperature 305 K N2, CO2, and H2O dilution inlet 39.9 m/s and temperature 305 K MILD-oxy condition	JHC burner	MILD-H2O was most suitable for achieving lower NOx emissions while maintaining combustion stability.
Cai and Zhao [91]	CFD simulation	ε = 0.1–0.5 The wall thermal conductivity (WTC) was increased from 1 to 100 W/mK Ammonia/hydrogen–oxygen combustion	Micro-planar combustor	Adding hydrogen reduced the NO concentration and changed the flame position. High WTC improved the temperature uniformity and reduced the NO emissions. The RE increased with the WTC.
Luo et al. [92]	CFD simulation	Φ H2 and NH3 = 1–1.2 Hydrogen content: 10–30% Ammonia/hydrogen–oxygen combustion Premixed combustion	Segmented nozzle micro-combustor in thermophotovoltaic system	The segmented constructs enhanced the thermal efficiency and minimized NO emissions. Increasing the H2 content resulted in enhanced NO emissions. ΦNH3 = 1.2 and ΦH2 = 1.0 were the effective conditions.
Sun et al. [93]	CFD simulation	Initial pressure and temperature: 1.0 atm and 300 K, respectively Φ = 0.6–1.4 Hydrogen concentration: 0.1–0.5 Ammonia/hydrogen-oxygen-air combustion Premixed combustion	Rotating detonation engine	A sustained detonation wave requires a hydrogen concentration of at least 0.2. Φ must be maintained between 0.9 and 1.3. Φ is an effective feature for managing emissions.
Ilbas et al. [94]	CFD simulation	Ammonia content: 10–30% Ammonia/kerosene–oxygen combustion Non-premixed combustion	Swirl gas turbine combustor	An ammonia addition did not significantly alter the maximum temperature of kerosene combustion.

Table 3. Recent research applications of ammonia oxy-combustion in a gas furnace.

Authors	Research Method	Combustion Conditions	Approach Equipment	Key Findings
El-Adawy and Nemitallah [77]	Computational fluid dynamics (CFD) simulation	Tin = 298–498 K Vin = 5.2–9.0 Φ = 0.6–1.2 Ammonia oxy-combustion	Swirl gas turbine combustor	Flamelet-generated manifold models for complex turbulent combustion. Higher Tin and peak combustion temperatures. Higher Vin resulted in adequate mixing and homogenous temperature profiles.
Cai et al. [78]	CFD simulation	Initial pressure of 1–2 atm Initial temperature of 300–500 K Φ = 0.8–1.2 Ammonia oxy-combustion	Micro-planar combustor	Φ = 0.9 yielded the optimal heat transfer performance. Increasing Tin decreased the outer wall mean temperature (OWMT) and NO. Increasing Pin increased the OWMT but reduced NO formation.
Zhang et al. [79]	CFD simulation	Initial pressure and temperature of 1.0 atm and 300 K, respectively Φ = 0.6–1.1 Velocity: 0.1–1.1 m/s Non-premixed combustion	Micro-planar combustor	Non-premixed combustion prevented flashback and provided improved heat and emission efficiency. Φ = 0.9 yielded the optimal heat transfer performance. Increasing the velocity increased the temperature.
Cafiero et al. [80]	Experiment	Inlet velocity: 80–50 m/s Φ = 0.9–1.01 Ammonia oxy-combustion Oxygen concentration: 21–50%	20-kW semi-industrial-scale furnace	O2 enrichment of 37% is optimal for moderate or intense low-oxygen diffusion (MILD) combustion. A higher inlet oxidizer velocity stabilized the combustion and temperature distribution. At Φ = 0.95 and the maximum temperature, the optimal compromise between nitrogen oxide (NOx) emissions and ammonia slip can be achieved.
Matveevet al. [82]	Chemkin simulation	Average temperature at the plasma torch outlet: 4950 K RF plasma torch frequency: 3–7 MHz Pressure: 1–10 bar Fuel and oxidizer temperatures: 288 K and 409 K, respectively	Perfectly stirred reactor	Plasma assistance significantly stabilized the combustion. Increasing the pressure in combustion decreased the NOx emissions.
Matveevet al. [83]	CFD simulation	Pressure: 0.3 MPa Initial flow rate of ammonia: 53.76 g/s; temperature: 288 K Plasma oxygen flow rate: 3 g/s; average temperature: 4990 K Oxidizer flowrate: 400–720 g/s; temperature: 409 K	Radial swirl combustion chamber	The eddy dissipation concept model was implemented. Plasma assistance trigger factor in ammonia ignition. NO formed in a central zone, whereas NO2 was produced at reduced temperatures.
Cai et al. [84]	CFD simulation	Initial temperature: 300 K Flow rate: 600–1200 mL/min h = 1/5–4/5 l = 1/5–4/5 Ammonia oxy-premixed combustion	Micro-planar combustor with a bluff-body	A bluff body can increase the outer wall temperature (OWT) and reduce NO generation. The ideal dimensions of the bluff body are h = 4/5 and l = 4/5, which optimize the maximum OWT while minimizing the NOx emissions.
Cai et al. [85]	CFD simulation	Initial pressure and temperature: 1.0 atm and 300 K, respectively Φ = 0.8–1 Ammonia oxy-premixed combustion	Micro-planar combustor with perforated plates	Perforated plates reduced the LBV and NO formation. High-thermal-conductivity materials were effective in conjugate heat transfer.
Bektas et al. [86]	CFD simulation	Φ = 1.1–1.68 Ammonia oxy- and air-premixed combustion Methane air-premixed combustion	Two-inlet cyclone combustor	The performance in oxy-ammonia combustion was outstanding compared with that in air–methane and air–ammonia combustion. The maximum flame occurred at Φ = 1.1 for oxy-ammonia.
Zhao et al. [88]	CFD simulation	Fuel inlet: CH4/NH3 at 39.9 m/s with a temperature of 305 K Co-flow inlet: 3.2 m/s with a temperature of 1700 K XCO2 = 0–41% XH2O = 0–50% MILD-oxy condition	Jet flame in hot co-flow (JHC) burner	Increasing the CO2 and H2O contents increased the formation of NO.
Zhao et al. [89]	CFD simulation	Fuel inlet: CH4/NH3 Re 9482 (39.9 m/s) and temperature 305 K Co-flow inlet: 3.2 m/s and temperature 1300–2100 K XO2 = 3–30% XCO2 = 97–70% MILD-oxy condition	JHC burner	MILD combustion occurred at lower XO2 levels or higher Tcof levels. The NO emission index decreased as XO2 and Tcof increased.
Zhao et al. [90]	CFD simulation	Fuel inlet: (CH4/NH3) Re 39.9 m/s and temperature 305 K N2, CO2, and H2O dilution inlet 39.9 m/s and temperature 305 K MILD-oxy condition	JHC burner	MILD-H2O was most suitable for achieving lower NOx emissions while maintaining combustion stability.
Cai and Zhao [91]	CFD simulation	ε = 0.1–0.5 The wall thermal conductivity (WTC) was increased from 1 to 100 W/mK Ammonia/hydrogen–oxygen combustion	Micro-planar combustor	Adding hydrogen reduced the NO concentration and changed the flame position. High WTC improved the temperature uniformity and reduced the NO emissions. The RE increased with the WTC.
Luo et al. [92]	CFD simulation	Φ H2 and NH3 = 1–1.2 Hydrogen content: 10–30% Ammonia/hydrogen–oxygen combustion Premixed combustion	Segmented nozzle micro-combustor in thermophotovoltaic system	The segmented constructs enhanced the thermal efficiency and minimized NO emissions. Increasing the H2 content resulted in enhanced NO emissions. ΦNH3 = 1.2 and ΦH2 = 1.0 were the effective conditions.
Sun et al. [93]	CFD simulation	Initial pressure and temperature: 1.0 atm and 300 K, respectively Φ = 0.6–1.4 Hydrogen concentration: 0.1–0.5 Ammonia/hydrogen-oxygen-air combustion Premixed combustion	Rotating detonation engine	A sustained detonation wave requires a hydrogen concentration of at least 0.2. Φ must be maintained between 0.9 and 1.3. Φ is an effective feature for managing emissions.
Ilbas et al. [94]	CFD simulation	Ammonia content: 10–30% Ammonia/kerosene–oxygen combustion Non-premixed combustion	Swirl gas turbine combustor	An ammonia addition did not significantly alter the maximum temperature of kerosene combustion.

Ammonia oxy-combustion in gas furnaces can be improved without a drastic redesign but requires system modification in the ignition process. Ammonia in oxygen combustion can improve combustion efficiency, although flame stability and NO_x emissions remain significant challenges. NO_x formation is significantly influenced by flame temperature and oxygen availability factors that can be overcome by MILD combustion. Most researchers use fuel blending to overcome such ammonia combustion problems, incorporating fuels such as methane, hydrogen, and kerosene, which positively impacts the flame speed and flame stability. Although hydrogen improves combustion efficiency and significantly reduces the amount of unburned ammonia, a high proportion of blended hydrogen increases NO_x emissions; we observed NO_x emissions under fuel-rich conditions. This will provide an opportunity for a practical strategy, and blending ammonia and kerosene (up to 30% ammonia by heat fraction) does not significantly affect the combustion performance. The main changes occur in the maximum temperature level, which increases slightly, and the combustion distribution. In addition, integrating fuel mixing with advanced burner configurations, such as by performing a redesign with non-premixed combustion, incorporating a bluff body, and conducting micro-segmentation, is also effective in improving the combustion performance and reducing NO_x. However, it must be studied thoroughly, especially on an industrial scale.

4.2. Ammonia Oxy-Fuel Combustion in Internal Combustion Engines

The prior studies related to ammonia oxy-fuel combustion in internal combustion engines are further discussed and summarized in Table 4. Park et al. [95] examined the potential of ammonia energy in an ICE and determined the characteristics of ammonia-enriched oxygen combustion. They observed that adding oxygen reduced the ignition delay under low-speed, low-load conditions, as evidenced by a reduction in the coefficient of variation for the indicated mean effective pressure. Figure 32 shows that increasing the oxygen concentration can improve stability and thermal efficiency, which increases significantly with the addition of 25% oxygen. Nonetheless, if the oxygen addition was greater than this value, then the increase in efficiency was no longer significant. Oxygen enrichment significantly decreased the amount of unburned NH₃ and N₂O emissions but increased NO_x emissions.

Nonavinakere et al. [96] explored how diverse oxygen levels influence the ammonia combustion sparked in a constant-volume combustion chamber. Figure 33 shows a series of high-speed photographs capturing the varying oxygen levels observed at various times following ignition. These findings reveal that increasing the oxygen concentration renders the flame velocity sensitive to the air–fuel ratio. According to Figure 34, the highest flame velocity recorded was 112.7 cm/s at 40% oxygen with a Φ value of 1.1 under 10 bar of pressure. Lower oxygen levels also impacted the flame velocity by causing flame instability, which resulted in asymmetric flame development triggered by buoyancy effects within the burner. The study also highlighted other practical effects of increasing the oxygen concentration, such as improved chemical reactivity and combustion completeness. As shown in Figure 35, these variables increased the peak heat release rate by approximately 2.5 times. Furthermore, they decreased the ignition delay and prolonged the flames.

Wang et al. [97] dissected the ignition and combustion enhancement processes of an NH₃/water mixture with auxiliary oxygen using a turbulent jet ignition (TJI) system. They analyzed the effects of various parameters, including the pre-chamber equivalence ratio (ΦPC), primary chamber equivalence ratio (ΦMC), and orifice diameter (d), on the combustion performance. The study revealed that auxiliary oxygen injection significantly improved the ignition performance, resulting in better flame propagation and stability. ΦPC had no significant effect on the jet power, but increasing ΦPC could decrease the ignition delay due to the increased availability of O₂. Although ΦMC controls the ignition delay and combustion duration, combustion was most efficient at ΦMC values between 1.2 and 1.3, balancing the mixture reactivity and oxygen availability. This study also highlighted that decreasing the orifice diameter increased the jet velocity, as shown in Figure 36, thereby advancing the flow turbulence. However, the turbulent flow caused a lack of ignition performance because NH₃/air cannot be ignited under high-turbulence velocity conditions.

Hong et al. [98] examined the effect of oxygen additions on the performance of an ammonia–hydrogen dual-fuel engine by applying a high compression ratio using hydrogen direct injection. They explored the impacts of single- and dual-hydrogen injection modes on the flame effect, combustion stability, and thermal efficiency. In addition, the effects of oxygen enrichment levels on burning and flame propagation were analyzed. Some injection strategies decreased the primary combustion duration, cyclic variation (CoV_Pmax), BMEP, and brake thermal efficiency (BTE). These improvements were due to the development of a hydrogen jet and mixture distribution for the optimal combustion process. In addition, increasing the oxygen concentration could enhance the heat release, shorten the ignition delay, and improve the combustion stability. However, higher oxygen levels reduced the BTE because of the increased adiabatic flame temperature, which caused more significant heat transfer losses. The dual hydrogen injection mode was more suitable when combined with oxygen addition strategies, resulting in improved engine performance. Figure 37 shows a hybrid power system design utilizing an ammonia–hydrogen dual-fuel engine as a future technology development based on the results of this and previous studies.

Park et al. [99] obtained significant insights through experiments on spark ignition engines using ammonia as preliminary fuel and the role of selective catalytic reduction (SCR) in reducing NO_x and unburned ammonia emissions. The effect of increasing the oxygen concentration in the intake air was also evaluated. The authors found that increasing the amount of O₂ in the intake air enhanced the combustion rate of ammonia and consequently improved the stability and efficiency of combustion. The increase in combustion rate with the addition of O₂ also affected the concentrations of the components in the exhaust gas, and as the O₂ concentration increased, NO_x increased, whereas the amount of unburned NH₃ decreased. The SCR post-treatment system showed that the O₂ concentration in the intake air could maximize the conversion efficiency of low NO_x and unburned NH₃ at 25.5% under high-load conditions. In contrast, under low-load conditions, the ideal O₂ concentration was 23% for NO_x reduction and 24% for unburned NH₃ reduction. Therefore, the operating conditions could be utilized to maximize the conversion efficiency in the SCR post-treatment system by adjusting the ratio of NO_x and unburned NH₃ via changing the O₂ concentration.

Table 4. Recent studies on the application of ammonia oxy-combustion in an internal combustion engine.

Authors	Research Method	Combustion Conditions	Approach Equipment	Key Findings
Park et al. [95]	Experiment	Operating conditions: 1500 rpm low load and 2000 rpm high load Oxygen content: 20–30% Ammonia oxy-combustion	SI engine	Applying oxygen to the intake air improved the spark timing, burning speed, and thermal efficiency.
Nonavinakere et al. [96]	Experiment and Chemkin simulation	Initial pressure and temperature: 10 bar and 373 K, respectively Φ = 0.9–1.15 Ammonia oxy-combustion (oxygen concentrations of 15–40%)	Constant-volume chamber Schlieren technique	Increasing the oxygen concentration positively impacted the ammonia combustion, flame velocity, ignition delay time, and HHR.
Wang et al. [97]	Experiment	ΦPC = 0.6–1.2 ΦMC = 1.1–1.4 d = 3–5 mm Ammonia oxy-combustion	TJI Constant-volume combustion bomb	O2 boosted the ignition capability of the unburnt fuel in the primary chamber. The orifice diameter controlled the jet velocity and turbulent intensity.
Hong et al. [98]	Experiment	Hydrogen direct injection For multiple injections, a secondary injection ratio of 0.5 Oxygen content: 21–34.29% Ammonia–hydrogen oxy-combustion	Dual-fuel engine	The engine performance metrics were enhanced using multiple injection techniques. Oxygen enrichment improved the combustion, but the thermal efficiency was compromised.
Park et al. [99]	Experiment	Operating conditions: low and high loads of 1000 and 1500 rpm, respectively Oxygen content: 20–30% Ammonia oxy-combustion	SI engine	Adjusting the O2 concentration in the selective catalytic reduction post-treatment system can enhance the conversion efficiency by optimizing the ratio of NOx to unburned ammonia.

Table 4. Recent studies on the application of ammonia oxy-combustion in an internal combustion engine.

Authors	Research Method	Combustion Conditions	Approach Equipment	Key Findings
Park et al. [95]	Experiment	Operating conditions: 1500 rpm low load and 2000 rpm high load Oxygen content: 20–30% Ammonia oxy-combustion	SI engine	Applying oxygen to the intake air improved the spark timing, burning speed, and thermal efficiency.
Nonavinakere et al. [96]	Experiment and Chemkin simulation	Initial pressure and temperature: 10 bar and 373 K, respectively Φ = 0.9–1.15 Ammonia oxy-combustion (oxygen concentrations of 15–40%)	Constant-volume chamber Schlieren technique	Increasing the oxygen concentration positively impacted the ammonia combustion, flame velocity, ignition delay time, and HHR.
Wang et al. [97]	Experiment	ΦPC = 0.6–1.2 ΦMC = 1.1–1.4 d = 3–5 mm Ammonia oxy-combustion	TJI Constant-volume combustion bomb	O2 boosted the ignition capability of the unburnt fuel in the primary chamber. The orifice diameter controlled the jet velocity and turbulent intensity.
Hong et al. [98]	Experiment	Hydrogen direct injection For multiple injections, a secondary injection ratio of 0.5 Oxygen content: 21–34.29% Ammonia–hydrogen oxy-combustion	Dual-fuel engine	The engine performance metrics were enhanced using multiple injection techniques. Oxygen enrichment improved the combustion, but the thermal efficiency was compromised.
Park et al. [99]	Experiment	Operating conditions: low and high loads of 1000 and 1500 rpm, respectively Oxygen content: 20–30% Ammonia oxy-combustion	SI engine	Adjusting the O2 concentration in the selective catalytic reduction post-treatment system can enhance the conversion efficiency by optimizing the ratio of NOx to unburned ammonia.

Research on ammonia oxy-combustion in ICEs has shown that NO_x is highly sensitive to the combustion phase, ignition timing, and oxygen concentration. An adaptive ignition control strategy is needed to achieve optimal performance without lowering emission standards. In addition, hydrogen injection can stabilize ammonia combustion. However, analyzing other additives such as methanol, ethanol, and propane is worthwhile, as they can offer better performance improvement between the initial ignition and flame speed [100,101]. This research is mainly conducted experimentally at a lab scale. In practical research, studies using a machine-scale approach with real load conditions are needed.

4.3. Ammonia Oxy-Fuel Combustion in Coal-Fired Furnaces

Another application of ammonia oxy-fuel combustion that is in coal-fired furnaces is discussed in this subsection and summarized in Table 5. Ghadi et al. [102] conducted a simulation analysis of a pilot-scale furnace encompassing ammonia–coal oxy-co-firing with variations from pure coal to 50% ammonia in a swirl burner. They investigated the consequences of ammonia oxy-co-firing on the fluid motion and heat transfer, species concentration, volatile combustion char burnout, and flue gas emissions. The authors presented a simulation using Ansys Fluent 22.0 for gas and particle modeling using the Lagrangian method. Additionally, the turbulent flow in combustion was analyzed using the realizable k-ε modeling method, which is particularly effective at predicting swirling flows. In addition, the effect of radiation on the heat transfer process was considered using the weighted sum of gray gases and discrete ordinate models. The authors found that the swirling motion improved the integration of the fuel and oxidizer, thus affecting the combustion efficiency. Increasing the ammonia content intensified the formation of internal recirculation zones. The ammonia ignited sooner and burned faster than the coal, leading to increased oxygen consumption near the burner. CO₂ emissions were reduced with increasing additions of ammonia in the fuel mixture compared with pure coal. In contrast, the NO emissions in Figure 38 and Figure 39 show a sharp increase when 10% ammonia is added. However, the NO formation rate stabilizes when the amount of ammonia is increased to 50%, where the NO level for the 50% coal-50% ammonia mixture (shown by the red line) is about 30 ppm higher than the NO levels for 100% coal combustion (indicated by the purple line).

Pan et al. [103] utilized an experimental approach and CFD-DEM numerical modeling to analyze the interaction between ammonia and oxygen bubbles in a fluidized bed combustion process. The mechanics of diffusion combustion between ammonia and oxygen bubbles in fluidized beds have been studied, emphasizing the effects of mass transfer on combustion reactions. Moreover, the combustion process occurs in three stages, as shown in Figure 40: the bubble injection process, bubble approach process, and merging or bursting of bubbles. Mass transfer occurs during the bubble approach process. The oxygen is repositioned from the bottom to the top of the bubble. At the same time, NH₃ from the upper bubble is absorbed by the lower bubble through the emulsion, which promotes combustion. The interval affects the bubble coalescence and combustion efficiency, with a shorter interval time increasing the oxygen transfer and combustion.

The role of oxygen and char nitrogen oxidation mechanisms in ammonia and coal combustion on the formation of NO_x emissions was investigated by Zheng et al. [104] using density functional theory. This examination revealed three primary reaction pathways involved in ammonia–coal combustion: the formation of N₂, NO₂, and NO. These findings indicate that oxygen promotes the oxidation of NO to NO₂, which effectively reduces the activation energy of this process. As the temperature increased from 1000 K to 2000 K, the NO reduction rate by ammonia consistently exceeded the NO production rate.

Chen et al. [105] studied the dynamics of NO production during the combustion of ammonia and coal using experimental studies and quantum chemistry computations under O₂/CO₂ atmospheric conditions. The results revealed that elevated temperatures and oxygen concentrations promoted NO formation, which peaked earlier. Meanwhile, CO₂ significantly inhibited NO formation when the temperature increased. Using quantum calculations, the authors discovered that O₂ reduced the energy barrier for ammonia–nitrogen oxidation and promoted its conversion into nitrogen oxide while increasing the energy barrier for coal–nitrogen oxidation. The combustion of oxygen fuel along with flue gas recirculation is an effective method for suppressing NO emissions in combustion processes involving ammonia and coal. Lei et al. [106] investigated the NO pollution characteristics of coal and ammonia co-firing under O₂/N₂ and O₂/CO₂ atmospheres using experiments and reactive force field molecular dynamics simulations (ReaxFF MD). The findings revealed that burning oxygen, especially in an O₂/CO₂ environment, suppressed NO emission generation compared with combustion with conventional air (O₂/N₂). Nevertheless, NO formation increased with the temperature and NH₃ blending ratio, whereas the CO₂ concentration inhibited the transformation of NH₃ into NO at low temperatures. In contrast, elevated temperatures accelerated the reaction, increasing the number of OH radicals in the burning process and encouraging the conversion of NH₃ into NO.

Xia et al. [107] conducted an experiment in a constant-volume chamber stirred by a fan to examine the flame propagation during the combustion of ammonia and coal under varying NH₃/O₂/N₂ Φ values. They observed that the flame propagation rate during the combustion of ammonia and coal exceeded that during pure ammonia combustion under lean conditions. Regardless, under ammonia-rich conditions, the speed of flame propagation diminished for the coal and ammonia mixture due to heat being absorbed by coal particles and an elevated fuel Φ value, which reduced the LBV under stoichiometric conditions. The flame propagation velocity was determined to be similar for co-combustion and pure ammonia combustion.

Table 5. Recent studies on the application of ammonia oxy-combustion in a boiler.

Authors	Research Method	Combustion Conditions	Approach Equipment	Key Findings
Zare Ghadi et al. [102]	CFD simulation	LHV of ammonia: 18.6 MJ/kg; LHV of coal: 20.9 MJ/kg Ammonia content: 0–50% Ammonia/coal oxy-combustion	Swirl burner	Increasing the ammonia content improved combustion efficiency and reduced CO2 emissions, whereas the NO emissions rose initially but stabilized as the ammonia content increased further.
Pan et al. [103]	Experiment and CFD-DEM simulation	Initial temperature: 200 °C Pressure: 3 bar Particle diameter: 0.5 mm Ammonia–oxygen	Fluidized bed furnace	The mass transfer process of bubbles significantly impacted the combustion efficiency. Shorter interval times enhanced oxygen transport and combustion.
Zheng et al. [104]	Quantum chemistry calculations	NH3 content: 0–100% NH3/coal/O2 combustion	A single-layer graphene structure	Oxygen promoted the conversion of NO into NO2 during combustion.
Chen et al. [105]	Experiment and quantum chemistry calculations	Ammonia content: 15% Pulverized coal: 119 mg Oxygen content: 21–40% Ammonia/coal oxy-combustion	High-temperature tube furnace Constant-temperature fixed bed	Oxy-fuel combustion with flue gas recirculation significantly prevented NO emissions.
Lei et al. [106]	Experiment and ReaxFF MD simulation	Bituminous coal Blending ratio of NH3: 0–60% Coal/NH3/O2/CO2/N2 combustion	Drop tube furnace	The NO in an O2/CO2 atmosphere exceeded that in an O2/N2 atmosphere.
Xia et al. [107]	Experiment	Φ ammonia/oxygen/nitrogen in co-combustion: 0.4–1.4 Turbulence intensity: 0–1.29 m/s Coal/ammonia/oxygen/nitrogen combustion	A fan-stirred constant volume chamber Schlieren images	The flame propagation rate in co-combustion rose under ammonia-lean conditions. Stoichiometry indicated the same rate for both combustion types.

Table 5. Recent studies on the application of ammonia oxy-combustion in a boiler.

Authors	Research Method	Combustion Conditions	Approach Equipment	Key Findings
Zare Ghadi et al. [102]	CFD simulation	LHV of ammonia: 18.6 MJ/kg; LHV of coal: 20.9 MJ/kg Ammonia content: 0–50% Ammonia/coal oxy-combustion	Swirl burner	Increasing the ammonia content improved combustion efficiency and reduced CO2 emissions, whereas the NO emissions rose initially but stabilized as the ammonia content increased further.
Pan et al. [103]	Experiment and CFD-DEM simulation	Initial temperature: 200 °C Pressure: 3 bar Particle diameter: 0.5 mm Ammonia–oxygen	Fluidized bed furnace	The mass transfer process of bubbles significantly impacted the combustion efficiency. Shorter interval times enhanced oxygen transport and combustion.
Zheng et al. [104]	Quantum chemistry calculations	NH3 content: 0–100% NH3/coal/O2 combustion	A single-layer graphene structure	Oxygen promoted the conversion of NO into NO2 during combustion.
Chen et al. [105]	Experiment and quantum chemistry calculations	Ammonia content: 15% Pulverized coal: 119 mg Oxygen content: 21–40% Ammonia/coal oxy-combustion	High-temperature tube furnace Constant-temperature fixed bed	Oxy-fuel combustion with flue gas recirculation significantly prevented NO emissions.
Lei et al. [106]	Experiment and ReaxFF MD simulation	Bituminous coal Blending ratio of NH3: 0–60% Coal/NH3/O2/CO2/N2 combustion	Drop tube furnace	The NO in an O2/CO2 atmosphere exceeded that in an O2/N2 atmosphere.
Xia et al. [107]	Experiment	Φ ammonia/oxygen/nitrogen in co-combustion: 0.4–1.4 Turbulence intensity: 0–1.29 m/s Coal/ammonia/oxygen/nitrogen combustion	A fan-stirred constant volume chamber Schlieren images	The flame propagation rate in co-combustion rose under ammonia-lean conditions. Stoichiometry indicated the same rate for both combustion types.

Integrating oxy-fuel ammonia combustion into coal-fired furnaces is an essential and fascinating topic to research in the effort to decarbonize legacy thermal power systems, which are widely used for electricity generation. The combustion strategy uses ammonia with coal in oxygen combustion to decrease CO₂ significantly. However, controlling NO_x under this combustion condition remains a concern because of the fuel-bound nitrogen in both ammonia and coal.

5. Conclusions

Ammonia oxy-combustion has emerged as a technological development that offers renewable energy and carbon-neutral fuels. It addresses the global challenges in energy supply and climate change issues generated by carbon emissions from traditional fuels. Using ammonia as fuel has several advantages, such as not generating pollution during combustion, not being easily explosive, and being easy to store in a mature supply chain infrastructure. On the production side, ammonia can be produced on a large scale by the HB method. However, ammonia also poses challenges, particularly due to its toxicity, making stringent safety measures to prevent leakage a critical concern. Another challenge with ammonia is the LBV, which makes ignition difficult and reduces its efficiency compared with conventional hydrocarbons. Additionally, its combustion generates fuel-bound NO_x and unburned ammonia, but oxygen enrichment in ammonia combustion can promote combustion efficiency.

Research on ammonia oxy-combustion has been conducted comprehensively, particularly focusing on turbulent flame propagation and LBV. The effects of Φ as well as the oxygen content, initial temperature or pressure, and dual and trinary fuel modes have been analyzed. In addition to chemical kinetic analysis, examinations have been conducted to validate experimental discoveries, establish the performance of sensitivity analyses, and identify the emission pathways. Previous research on ammonia–oxygen combustion revealed that the maximum LBV occurs at a Φ value of approximately 1.1 when the combustion is fuel-rich. However, this condition must be considered because it significantly increases the ammonia slip in the exhaust gas. This gain raises concerns about environmental pollution due to hydrocarbon emissions from unburned ammonia, resulting in decreased combustion efficiency. Moreover, other factors also influence the combustion process, including an elevated ammonia concentration and a higher starting pressure, which can decrease the LBV. Meanwhile, regarding NO emission formation, scholars have found that preheating enlargement caused an increase in NO. However, under conditions of higher pressure and fuel-rich combustion, the potential exists for NO emissions to be reduced.

This review examined the advancements in ammonia oxy-combustion for diverse applications in gas furnaces, ICEs, and boilers. Experimental and computational studies, such as CFD investigations, have been used to observe combustion phenomena. Miscellaneous technical innovations have also been employed to enhance combustion stability and decrease NO emissions. In gas furnaces, approaches such as MILD combustion, plasma assistance, non-premixed combustion, bluff body integration, WTC, and segmented nozzle design have been implemented. The dual-fuel modes include ammonia–methane, ammonia–hydrogen, and ammonia–kerosene. For ICE applications, the advancements focus on the multiple-injection method, redesigning the orifice diameter, and utilizing dual-fuel mode ammonia–hydrogen. Similarly, in boiler systems, these strategies use flue gas recirculation and the dual-fuel ammonia–coal mode.

In summary, ammonia oxy-combustion offers a promising route to improving combustion efficiency. Despite the observed improvements, this approach provokes issues primarily because it increases NO emissions. However, this review shows that the development of ammonia oxygen combustion is still focused on the laboratory and experimental scales, with most researchers concentrating on parameters such as the equivalence ratio and oxygen enrichment. Therefore, a comprehensive study on applying various methods in the ammonia oxy-combustion process is necessary to ensure stable combustion and control NO_x generation.

The general future research directions that could accelerate the transition to ammonia oxy-combustion systems, such as a staged combustion method and the application of a staged combustion system for ammonia oxy-combustion in gas furnaces, gas turbines, and boilers, are less explored. Future research is needed to analyze the effectiveness of the combustion stages, including the fuel and air stages, on the combustion performance and NO_x formation. Furthermore, mixing ammonia oxy-combustion with other fuels, such as hydrogen, methane, alcohol, and DME, requires further analysis under various oxygen combustion conditions in all applications, such as gas furnaces, ICEs, and boilers. Advanced burners with plasma assistance are important to analyze further because they can reduce ignition delay in ammonia combustion. More widely, research should be conducted on different combinations of pressures, temperatures, and burner designs. Ammonia slip must also be mitigated, especially under fuel-rich conditions such as those present in SCR applications. Overall, persistent research, invention, and partnerships are necessary to develop practical applications of this technology and support the global energy transition.