1. Introduction
The aim of the European Commission of reaching an economy based on net-zero greenhouse emissions by 2050 [
1] may probably be an over-optimistic objective understood from the perspective of great concerns regarding high CO
2 concentration levels reached in the atmosphere and recent extreme climatic events experienced worldwide. However, the amount of energy needed to maintain industrial activity, employment, and healthcare is globally increasing despite efforts to attain higher process efficiency and reduce energy consumption. Two significant trends can be observed: one is the almost steady demand of European, North, and Latin American countries, and the other is the increasing energy demand of Asia, Pacific, African, and Middle Eastern countries [
2], with oil, gas, and coal being the main energy sources.
Efforts in reducing CO
2 emissions and compliance with environmental measurements should keep pace with the country’s economic resources without affecting population wellbeing. Thus, Nigeria has experienced a great population increase and showed an energy demand of 167 Mtoe (millions of tons of oil equivalent) with a population of 213.4 million inhabitants by year 2021, which is contrasting against countries such as Saudi Arabia with a population of 35.6 million and an energy demand of 254 Mtoe [
2].
The focus on reducing CO
2 emissions is usually set on fossil sources and their replacement by renewable energy. Reducing the usage of fossil fuels is contradictory to social development since energy consumption is expected to increase along with overall economic development unless a change in energy structure is accomplished along with improvements in energy efficiency [
3]. However, the solution is not simple. Renewable energies, such as wind and solar technologies, have an intermittent nature with low reliability under stand-alone configuration, thus requiring an excessive over-dimension of installations [
4,
5]. To these disadvantages it should be added the fact of the reduction in energy efficiency experienced by solar panels due to adverse climatic conditions, such as humidity, high temperature, and dust accumulation [
6,
7] and the effect of extreme climatic events, requiring special measures that allow grids with high penetration of wind power to withstand hurricane damage [
8].
Great efforts are being made to increase energy efficiency and economic circularity thus favoring the reuse of materials and waste valorization. However, several factors need to be considered when assuming a circular economic model, such as material dispersion, low quality of wastes, process feasible integration, and energy demand associated with treating recycled components [
9]. The transition to decarbonizing the economy must cope with a global increasing energy demand and attain a gradual substitution of fossil fuels with alternative fuels. Alexander and Floyd [
10] described an optimistic view of the economic transition, indicating that adapting to climate change via deep decarbonization may be challenging but manageable if negative emission technologies are implemented. Their approach uses fuel substitution, nuclear power, and carbon capture and storage.
One important matter to carefully evaluate is the fact that decarbonizing the economy is costly, and although several technological alternatives are available, not all of them are mature enough to consider that a fast and smooth transition can be made in a relatively short period. Currently, fuel substitutes and electrification bring an unavoidable increase in production costs, which the producer cannot assume and, if translated to the consumer in full, may cause excessive market distortion. In a globalized market, producing goods at higher prices may cause consumers to decant their preferences for more affordable products from other regions despite being produced from higher carbon footprint manufacturing lines.
It should be taken into account that not all process or fuel alternatives may be directly linked to a decrease in energy intensity. An example of this feature is the production of hydrogen. This carbon-free gas has become the preferred green fuel due to the exclusive emission of water from combustion reactions and the high energy contained per unit of mass. However, the current production of hydrogen comes from fossil sources. Steam methane reforming (SMR) presents an energy demand of 10.84 kWh/kg H
2, while the production of H
2 from water electrolysis, which seems to be more environmentally friendly, requires even more energy (48 kWh/kg H
2) [
11]. A similar situation is found when implementing carbon capture and storage: CO
2 capture, compression transportation, and further transformation will undoubtedly raise energy consumption [
12].
The present document examines the feasibility of using ammonia as a fuel for energy production and as a complementary alternative fuel in road transport. The production process and combustion features of ammonia, when used with various fossil fuels and biofuels, are discussed. The emphasis is placed on the advantages of ammonia as an alternative fuel to be used in turbines and internal combustion engines (ICEs). This feature may help expedite the move towards a decarbonized economy.
3. Ammonia Production
Ammonia is one of the chemicals with the highest production worldwide due to its great application in the chemical industry and agriculture. Although, it is a gas at room temperature conditions with a pungent smell and corrosive characteristics, there is wide experience in its handling [
56]. In fact, the strong smell may serve as an alarm, allowing a prompt awareness of leakages and facilitating safety precautions. The high hydrogen content of ammonia (17.7 %wt/wt) and low-pressure liquid state (8.58 bars) make this molecule a suitable hydrogen carrier or an energy storage molecule. Ammonia production can be thought of as a way of transforming excess energy from renewables and equilibrating the balance between energy production and demand, a fact that is favored by the vast existing network of available pipelines, thus offering a lower transport cost than that estimated for hydrogen [
5,
57]. Therefore, the production of this molecule is expected to increase exponentially, with a forecast of 290 million metric tons by 2030 for the production capacity [
58]. This value may experience a tremendous rise if expectations in the energy and transport sector are achieved.
The industrial production of ammonia was dominated by the process developed by Fitz Haber in 1908 and industrialized by Carl Bosch thanks to solving material and pressure problems which allowed starting operation in 1913 of the first BASF industrial plant at Oppau site north of Ludwigshafen [
59]. Today, Casale S.A. is the ammonia leader company in plant design and revamping (up-grading and retrofitting). Between 1984 and 2021 over 200 ammonia converters have been retrofitted by Casale S.A. worldwide, increasing production yield by over 40% and developing a revolutionary iron-based catalyst called AmoMax®-Casale [
60].
The catalytic conversion of ammonia from N
2, previously separated from air, requires the production of H
2, which is currently obtained from fossil fuels (natural gas or coal), thus making this process one of the primary candidates for proposing the use of water electrolyzers using renewable energy as power supply. This way, a significant reduction in the carbon footprint is attained. Smith et al. [
61] estimated that a wind-powered ammonia process would have a carbon intensity of 0.12-0.53 t CO
2-eq/ t NH
3 against a value of 1.5-1.6 estimated for a modern highly efficient methane-fed process [
62]. However, the substitution is not so simple and estimation is not so straightforward.
The industrial market of water electrolyzers is dominated by alkaline water (AWE), proton exchange membrane (PEM) electrolyzers operating at low temperatures, and solid oxide electrolyzers currently under development, and with high operating temperatures [
63]. AWEs are preferred against PEM electrolyzers given their high installation costs and lower service life of the latter [
64]. Companies offering commercial units, such as Nel hydrogen (Oslo, Norway) [
65], offer atmospheric AWE with a production of 1.2 t H
2/day and power demand of 2.2 MW (3.8 kWh/m
3 (STP) H
2), whereas McPhy Energy Italia Srl (San Miniato, Italy) [
66] offers modules producing hydrogen at 30 bar with a consumption of 4.65 kWh/m
3 (STP) H
2. Additional energy is needed to increase H
2 pressure to the Haber-Bosch process (200-400 atm) and the availability of enough power from renewable source is also needed to guarantee a reliable supply.
Electrolyzers present several drawbacks which need an urgent solution. Inefficiencies in electrolyzer performance are associated with low load (25-45% of rated load), favoring the diffusion of hydrogen into the oxygen container, creating safety concerns regarding the explosive nature of the mixture [
67]. The flammability limits of the pure H
2-O
2 mixture are much wider than those expected for H
2-air mixtures, and the problem is aggravated with pressure increase. Fluctuations in renewable energy production would affect performance, leading to start-stop cycles. After shutting down electrolyzers, a reverse current phenomenon associated with redox reactions of the electrodes is observed, causing severe degradation (irreversible oxidation of cathodes), and adversely affecting the service life of electrocatalysts [
68,
69].
Chau et al. [
70] performed a study regarding different hydrogen production technologies evaluating safety, efficiency and infrastructure, concluding that the SMR process was the most desirable technology. Even if becoming fully optimistic by considering that solutions will be available in the short term regarding safety measurements that allow for technology scale-up, the challenges are still enormous. As an example, the ammonia plants of Fertiberia are expected to reduce their carbon footprint by introducing green hydrogen derived from the use of renewable energy. The plant located at Puerto Llano (Spain) has a capacity of producing 200 000 t/year, and the company intends to install 200 MW electrolyzers which will obtain electricity from photovoltaic panels [
71]. The intermittent nature of solar power and variations in irradiance and temperature deeply impact the reliability of electricity production [
72]. Thus, the increase of renewable in the energy mix comes inevitably with a concomitant risk of uncertainty due to the intrinsic dependence on sun irradiance and daily climatic conditions. An equivalent amount of energy as backup is then necessary to cover for unavoidable low/null energy production periods. A similar feature can be ascribed to wind power, causing an over-dimensioning of the power supply system. Therefore, interconnection with an electrical grid is needed as a warrant of steady energy supply to ensure electrolyzers’ energy requirements. The availability of other electricity sources, such as nuclear or large-scale energy storage systems, guarantees energy supply regardless of the intermittency of renewables.
An additional issue is water supply and quality. The water demand for electrolyzers is high, requiring pure water with low ion contents and low impurity levels. However, water scarcity and water distribution are among the greatest environmental concerns. During the past years, the global population faced changes in rain distribution patterns with extreme drought and flood episodes. This trend is expected to be maintained or even aggravated, requiring extraordinary measures to manage water availability for human consumption and ensure food supply due to the possible scarcity that will be probably faced in certain regions due to depletion of water reservoirs and incapacity of recharging [
73,
74]. Based on this scenario, it does not seem right to consider the production of H
2 as a feasible option in regions characterized by water scarcity problems.
Using seawater or wastewater would offer a great advantage to the process since avoids the treatment and transport of high-quality water to the hydrogen production site. Unfortunately, the presence of anions, such as chloride and bromide, and cations (Na, Ca, Mg) creates a severe decrease in performance [
75]. El-Shafie [
76] reviewed the technical barriers regarding the use of water sources different than pure water. These authors reported that the presence of organics increases cathode fouling rates, halide oxidation reactions occur under the presence of anionic impurities (chloride and bromide) instead of the oxygen evolution reaction [
77], and cationic poisoning causes a complete electrolyzer failure. Therefore, finding a practical solution to all these problems seems far away, so the immediate alternative to overcome these difficulties may be the use of desalinated water.
The energy demand for water desalination is about 3.2-4.5 kWh/m
3. However, water desalination for human consumption is considered a costly option and too intensively energy-demanding [
78]. Based on the previous statement, it does not seem appropriate to think that the use of desalinated water would be an adequate solution for H
2 production, particularly if 9 t of water are required for producing 1 t of H
2 [
79], but in contrast, is as an unadvisable choice when meeting the demand for human consumption. Despite this controversy, this alternative is being categorized as a feasible option by several authors based on the currently high costs of hydrogen production systems [
80,
81,
82].
It should be remembered that many current hydrogen projects are receiving public funds as an incentive for facilitating technological penetration. The objective should be to pursue economically viable activities along with climate protection. As a result, the costly installation and energy-intensive nature of the hydrogen production process should not be overlooked, and any additional equipment included in the process should not contribute to increasing the already high energy demand.
Saygin et al. [
83] analyzed the energy demand of the ammonia producing process. These authors indicated that the rise in oil and gas prices directly affected that of fertilizers since most of ammonia plants are based on the SMR process and were incapable of assuming such an increase in feedstock and energy prices, generating a curtailment in European production. The increase in ammonia prices also had a negative impact on food production consequently rising food prices. The interconnection between fuel prices and the economy is a factor that needs careful assessment. At its beginning, the Ukrainian war caused a great disruption in the cereal market; this, along with an increasing trend in inflation and oil prices, created food security risks [
84].
The future increase in ammonia prices may be caused by a greater implementation of green ammonia production worldwide given the higher cost of this technology [
85] or because of a higher demand if ammonia is used either as a hydrogen carrier or alternative fuel. Therefore, a careful deployment of green ammonia technology is necessary to avoid adverse perturbations in the food supply chain. An interesting review of the industrial ammonia production using water electrolysis was performed by Rouwenhorst et al. [
86]. These authors described the expansion of the process in different countries, highlighting the relevance of the Lougi Casale process, indicating that the main factor for the installation of these plants was the availability of cheap hydropower electricity for producing hydrogen. The lower costs of fossil based plants and better scale-up relegated electrolyzer based ammonia plants to a shut-down with the only plant still operating located in Cuzco, Perú (Cachimayo plant, Enaex) [
87].
Other technologies for producing ammonia include electrochemical synthesis which can be carried out at different temperature ranges using liquid electrolytes, molten salts, composite membranes and solid-state electrolytes [
79]. Ammonia can be also produced by the use of plasma catalysts and non-thermal plasma methods; in this latter case an electric discharge generates the ionization stage [
88,
89]. However, the commercial ammonia production is still dominated by the SMR.
4. Use of Ammonia as Alternative Fuel: Ammonia Combustion
The search for alternative fuels complying characteristics regarding carbon dioxide emissions and renewable sources has set the focus on carbon-free molecules. The immediate candidate to think of is hydrogen, but given the difficulties of attaining adequate volumetric densities at a reasonable cost, the attention has moved forward to hydrogen-rich molecules such as methane and ammonia. In the case of methane, this gas may be produced from the biological degradation of wastes under anaerobic conditions, or the Sabatier process using syngas derived from biomass or CO
2 absorbed directly from the atmosphere. The Sabatier reaction requires a special Nickel-based catalyst and a source of hydrogen, transforming one-mole unit of CO
2 into methane by the use of 4-mole units of hydrogen [
90,
91], making the process complex, with only one plant built until now. The e-gas Audi plant was built in Werlte, Germany, and uses CO
2 derived from an anaerobic digester, thus complying with the requirement of producing synthetic methane from renewable sources [
92]. The plant was transferred to Kiwi AG and recently acquired by Hy2Gen [
93].
Given the scarce experience in producing synthetic methane at an industrial scale, synthetic ammonia production seems a better option if a fast transition to a carbon-free economy is to be achieved. Ammonia has a higher energy volumetric density than hydrogen, although combustion properties are poor compared to hydrogen and methane, with poor ignition and flame propagating behavior [
94]. The range of flammability limits is narrower than that of methane or other hydrocarbons, and the flame speed is just 7 cm/s, whereas this value is in the range of 35-100 for conventional fuels and biofuels, reaching values as high as 291 for hydrogen [
95]. Ammonia’s lower energy content per unit mass impacts combustion characteristics by releasing a lower amount of energy in the reaction zone and thereby reaching lower flame temperatures. This affects reactivity because a lower concentration of radicals is present in the flame front [
96].
Despite these previous characteristics, ammonia has been proposed as a complementing fuel in ICEs to decrease the use of fossil fuels since the absence of carbon in its molecular structures avoids releasing greenhouse gas emissions. The poor combustion performance can be compensated by introducing modifications in engines to attain either neat ammonia combustion or its use as a dual fuel. The transition to a decarbonized economy may be faster if an important reduction in fossil fuel consumption is achieved by using suitable and available substitutes that have a proven track record of large-scale production and handling. ICEs commonly used in the power generation and transport sector can be classified as compression ignition, spark ignition, and gas turbines. The different operating characteristics of these engines translate into special features and combustion requirements. Thus, different techniques are necessary to adapt engines and attain a suitable performance when using mixtures of ammonia and conventional fuels.
Compression ignition engines require a fuel with a high cetane number to allow the combustion chamber to reach a high compression stage, attaining the mixture’s autoignition by the associated temperature increase. In this system direct injection of fuel is carried out once the air has been fully admitted into the cylinder and once the piston is closed to reach the top dead center position. On the contrary, spark ignition engines require a spark plug to initiate fuel-air mixture ignition. Traditionally, in these later systems, the admission of fuel and air takes place in a premixed mode previous to the cylinder’s intake ports. Modifications in fuel injections have been implemented in current modern engines, with many counting with direct or stratified fuel injection. These modifications were introduced to improve fuel use efficiency and reduce combustion temperature to lower NOx emissions. Two examples are homogenous charge compression ignition (HCCI), in which the mixture enters the cylinder achieving a full homogenous state prior to autoignition, and premixed charge combustion ignition (PCCI), where a partial premix is created, allowing for fuel stratification before combustion. Thus, this concept may be classified as being between conventional combustion and HCCI [
97].
The amount of fuel-air added to a combustion system can be expressed in terms of the equivalence ratio (φ). This parameter gives an idea of the combustion conditions, referencing the stoichiometric point (ratio of stoichiometric air and the amount of air added). Thus, lean mixtures are defined by values of φ lower than unity due to the addition of excess air. Modifying the amount of air added to combustion allows for controlling temperature and the type of emissions produced. The recirculation of combustion gases also reduces combustion temperature by returning to the reaction zone inert species, creating a quenching effect and acting as NOx control strategy. Exhaust gas recirculation (EGR) is a common feature of diesel engines and, recently, of new spark ignition engines [
98,
99]. The different operating concepts in ICEs will need to be adapted based on the combustion characteristics of ammonia. This gas has a high ignition energy (8 mJ for ammonia, whereas propane only requires 0.5 mJ, [
100] and low flame speed. Therefore, attaining stable performance under different engine operating regimes will be a challenging work. Several experimental research works deal with the use of ammonia-fuel mixtures of conventional (diesel, gasoline) and alternative fuels (ethanol, biodiesel) [
101,
102,
103].
The release of pollutants is a characteristic of ammonia combustion, which is a cause of great concern; a higher release of nitrogen oxides and unburnt ammonia in the exhaust may be expected [
104]. Combustion of conventional fuels has been extensively studied to reduce the production of NOx emissions. Changes in operating conditions in the combustion chamber have been introduced to lower temperatures and avoid thermal NOx emissions derived from the Zeldovich Mechanism [
105,
106]. Prompt NOx emissions (Fenimore mechanism) are associated with the presence of CH radicals following a different route, where HCN and CN are formed [
107] and additional emissions may be expected from the nitrogen fuel content. Burning under lean fuel conditions, combustion staging, or including a recirculation stream of exhaust gases to create reducing atmospheres are just some examples of the alternatives available for avoiding NOx formation. However, these strategies need additional adjustment and operational modifications when considering the addition of ammonia as a dual fuel.
The presence of nitrogen in the ammonia molecule would make anyone think of extremely high related fuel NOx emissions associated with direct ammonia oxidation. However, the reaction mechanism is complex, and the presence of different intermediary species allows for a window where the production of this undesirable contaminant is mitigated. The overall combustion reaction is the following:
The mechanism of ammonia combustion comprises the presence of HNO and NH as intermediary species participating in NO formation [
108]. Nitrogen-related contaminants are high [
109] but not as much as one may have inferred from its molecular composition. Shu et al. [
110] studied the mechanism of ammonia combustion, reporting that the presence of NH
2 intermediary species plays a fundamental role in auto-ignition, and its reaction with NO is crucial for the reaction outcome, either increasing or reducing the presence of this pollutant in exhaust gases. Other pathways responsible for NOx formation are those involving HNO and the Zeldovich pathway [
111].
Cai et al. [
109] reviewed different studies regarding ammonia combustion mechanisms and strategies to mitigate its formation. Operating under fuel-rich conditions, increasing pressure, staging combustion, and steam addition are some of the relevant strategies that have demonstrated success in controlling NOx presence [
112,
113,
114,
115].
Lean ammonia conditions favor the presence of O/H radicals, which are precursors to the formation of NO contaminants. Thus, lean ammonia flames show peaking NO levels at φ of 0.9, with higher values than those obtained from methane/air flames, but showing a minimum at slightly rich conditions (φ = 1) [
116]. The presence of H, OH and HNO radicals is crucial in the formation of NOx by the following reactions:
Therefore, adding H
2 to ammonia combustion under lean conditions increases the availability of these radicals in the flame front. On the contrary, increasing the φ over unity values, significantly reduces NOx formation because oxygen availability decreases as well as temperature [
117].
NO
2 and N
2O are also part of NOx emissions. NO
2 is responsible for smog formation, whereas N
2O is a compound that causes great concern because of its high warming potential (about 300 times greater than that of CO
2) [
118]. Therefore, benefits associated with CO
2 emission reduction may be in part mitigated by the release of N
2O or by the adverse effects associated with acid rain. The proportion of N
2O in the global NOx emission is believed to be closely related with incomplete combustion in ICEs, decreasing the yield of N
2O formation with temperature in the 850 – 1350 K range [
119,
120].
4.1. Gas Turbines
Generating a flame and attaining stable conditions in a combustion chamber can be achieved by the use of different strategies. However, attaining stable performance under different operating conditions with low levels of contaminants was a great challenge, which was only possible after several years of research and experience. In a very simplistic way, flames can be catalogued as premixed or diffusion types. In the first case, mixing of the fuel and oxidizer is performed prior to reaching the flame front, the combustion temperature reached corresponds to the φ supplied. The second case refers to the separate supply of fuel and air. Thus, diffusion flames are said to always reach higher combustion temperatures since there is a preference for the flame front to be located at the stoichiometric mixing condition point. Since the diffusion of the fuel and oxidizer to the reaction zone is required, these flames are diffusion limited, explaining thus its name.
The previous categorization is far from explaining all complexities taking place during the combustion. Burning fuels with excess air (lean conditions) reduces flame temperature and thus NOx emissions, but it also affects flame speed and may create acoustic interactions that adversely affect flame stability. In the case of burning fuels with air deficiency (rich conditions), flame temperature is also lowered, but unburnt hydrocarbons and CO emissions are a concern. Combustion science must deal with reaction chemistry, flow interactions, heat transport phenomena, and acoustic waves, thus explaining the difficulties in attaining stable performance under all possible engine operating regimes.
The use of swirling flow in combustion chambers increases flame stability by creating inner and outer recirculation zones that transport hot products and active radicals back to the reaction region, mixing with the fresh, unburnt mixture, increasing residence time and allowing for the fuel to be consumed [
121]. The presence of these zones allows for operation with high inlet flow velocities, although the flame speed of the fuel used may be low. A swirl flow creates a central vortex zone and low-pressure gradient, giving rise to a reverse flow [
122] that allows flame anchorage.
The strategies widely applied in gas turbines for attaining acceptable performance regarding combustion efficiency and emission control are dry low emission (DLE), rich-burn, quick-quench and lean-burn (RQL), and moderate or intense low oxygen dilution (MILD). The first refers to lean burn conditions where low adiabatic flame temperatures avoid NO formation under the thermal mechanism. Since the reduction in NOx emission is achieved without using water or steam as a heat sink, thus the denomination of dry [
123]. The RQL concept is based on the separation of combustion in two clearly distinct zones where rich conditions are initially favored, creating a pool of radicals that are subsequently oxidized in a secondary zone where excess air is sufficiently added, attaining a fast temperature drop and avoiding thermal NOx formation. MILD combustion refers to the recirculation of hot combustion gases, causing an oxygen dilution effect in the reaction zone and thus lowering peak temperatures [
124].
Ammonia/air flames are less stable than those of methane under similar conditions, with lower laminar burning velocities, heat release rate and lower flame temperature [
116]. However, combustion of this gas has been possible with success by attaining appropriate hydrodynamic conditions in the combustion chamber. Many industrial applications for energy production use methane in gas turbines. Therefore, there is a great interest in retrofitting current engines to allow for ammonia use, either as a mixture of methane/ammonia or as a single fuel, keeping always in mind that NOx regulation should be fully complied. Iki et al. [
125] studied the performance of a micro-turbine using methane and mixtures with ammonia and neat ammonia as fuel. Changes in air supply were needed to provide air at higher temperatures and lower flow rates. Staging combustion was also necessary to create two zones (rich-lean) and avoid an excessive production of NOx. Okafor et al. [
126] tested single and two-stage (rich-lean) combustion using methane/ammonia mixtures, reporting high NO emissions in the first case but much lower ones in the latter due to the enhanced fuel consumption in the primary zone (where rich conditions are set). The methane/ammonia mixture was characterized by higher flame speed compared to that of neat ammonia flames, a feature that ensured lower NOx production in the secondary zone.
Gubbi et al. [
55] performed a numerical study, evaluating the NO emission expected from ammonia combustion by changing combustor pressure and residence time. These authors indicated that current NOx regulation limits can be complied with, by changing operating parameters and allowing higher residence time in the combustor. Increasing pressure lowers NOx emissions when using ammonia as fuel since it reduces the availability of O and H radicals [
116], and the effect of pressure is just opposite to the behavior obtained when burning conventional fuels [
127]. This study opens a door for opportunities regarding the use of ammonia for reducing CO
2 emissions since it is expected that the problem of NO formation will be resolved. However, the increase in pressure also causes a decrease in the laminar flame speed, and the presence of ammonia causes a decrease in temperature and net heat release rate [
128], thus requiring modifications intended to attain flame stabilization under such peculiar fuel characteristics.
Rocha et al. [
129] also performed a numerical evaluation of ammonia combustion and its emission characteristics by considering the three typical gas turbine operating concepts for keeping low NOx emissions, that is: DLE, RQL and MILD. Their results indicated that the DLE concept was inappropriate when dealing with ammonia as fuel, whereas RQL and MILD combustion achieved the lowest emissions. The presence of a post-combustion zone at high temperatures enhanced the consumption of NO previously formed by the reaction NH
2 + NO → H
2O + N
2 [
130].
The poor combustion characteristics of ammonia were the main reason for considering the use of more reactive fuels as additives either for starting up the system until steady combustion is attained or using ammonia in dual systems to warrant stable performance at all loading conditions. H
2 has been added as an ignition enhancer to increase the pool of radicals, flame speed, and stability [
131,
132], but the supplementation of this gas brings along problems associated with storage and handling, along with the use of separate deposits for ammonia and hydrogen. The autothermal cracking of ammonia to release a fraction of hydrogen contained in the ammonia molecule, or the use of catalysts to aid in the thermal decomposition of ammonia, has been proposed as a solution for avoiding the presence of an additional high-pressure container and reducing safety risks.
Regardless of difficulties encountered when attempting to use ammonia as an alternative fuel, several experiences have demonstrated that technical advances can overcome combustion inefficiencies observed and reduce the emission of pollutants. Therefore, there is a great interest in considering ammonia as a suitable substitute for conventional fuels and an enabler of the hydrogen economy thanks to its feasible use in power generation and transport sector (aircraft propulsion and road and marine transport) [
133].
Table 1 shows different experimental studies carried out to evaluate the combustion characteristics of ammonia and its performance in gas turbines as single fuel and as dual systems. In these experiences, it can be observed that the addition of ammonia under certain operating conditions causes an increase in NO emissions, specifically when working under premixed conditions [
134,
135] or using a swirl burner [
135,
136]. Verkamp et al. [
100] proposed partial dissociation of ammonia as a method for flame stabilization, achieving stable operation by working only with ammonia as fuel. It is also possible to keep NO emissions under control. Khateeb et al. [
136] succeeded in keeping NO emissions under control in a swirl burner under rich operating conditions, thus reducing the risk of backfiring. These experiences show that using ammonia as a fuel in turbines may be possible in the future with certain improvements in the current state of the art.
4.2. Spark Ignition Engines
ICEs are widely used in the transport sector and power generation. Spark ignition engines are the preferred option in passenger cars, although the higher efficiency of diesel engines and advancements performed in the last decade regarding particle emission control and NOx production. The diesel scandal damaged public perception towards diesel passenger cars, and political pressures set a regulation preference for gasoline and electric vehicles. Even though diesel vehicles have an equivalent environmental classification in Spain as gasoline and lower CO
2 emission per kilometer, the market experienced a severe drop in sales of this type of tourisms [
147].
Spark ignition (SI) engines can be easily adapted to use gaseous fuels such as natural gas and liquefied petroleum gases (LPG), currently becoming a practical solution for many consumers due to their lower fuel consumption and price. SI engines are also widely used for valorizing biogas since these engines produce electricity and heat recovery concomitantly, which is essential for keeping digestion temperature. When operating under lean burn conditions, NOx emissions are restricted to 400 – 500 ppm [
148,
149]. A further reduction in NOx levels, then requires the use of selective catalytic reactors (SCR). Main commercial combined heat and power (CHP) engines include Jenbacher®, MWM®, Caterpillar®, MTU®, Waukesha® and MAN® [
150,
151,
152,
153,
154,
155] and many of these engines are currently adapted for also using hydrogen as fuel, just as in the case of gas turbines.
Ammonia has been considered a fuel worth research by industrial manufacturers regarding its use in gas turbines and compression ignition engines, but this is not the case for spark ignition engines, where the main research is found in academic sites. Early research performed by Cornelius et al. [
156] and Starkman et al. [
157] already provided a clear insight into the modifications required when using this gas in SI engines, such as the need for increasing compression ratio and using supercharges due to the lower energy content of this molecule, the lower air to fuel ratio to attain stoichiometric conditions and the need of partial ammonia dissociation to release hydrogen (or direct hydrogen addition), which is critical to attain stable performance under differing engine loading. Hydrogen enrichment or increased intake pressure may serve as suitable measurements to attain operation with satisfactory combustion efficiency [
49,
158].
Natural gas engines are more suitable for retrofitting to ammonia-hydrogen operation due to their higher compression ratio and the lower dependence on fuel laminar flame speed of their combustion chamber [
159]. However, there is a limit to the addition of hydrogen in the mixture based on the risk of backfire and wall heat losses [
160]. Low levels of hydrogen addition allow stability improvement, acting as an ignition promoter and increasing laminar flame speed, but this single factor does not fully explain the improvement in operation achieved [
161]. The high octane number of ammonia (130) makes the engine operate under higher compression ratios than the one used for gasoline [
162], and the different burning characteristics of hydrogen and ammonia allow for attaining a mixture proportion capable of reaching a similar performance regarding work output as that obtained with gasoline [
163]. This approach's main disadvantage is the need for separate fuel storing tanks, making the system too complex for on-board applications. However, in the case of liquid and a gaseous supplementary fuels the technological configuration resembles that of the already tested technology for LPG and petrol passenger cars.
Experiences carried out by Grannell et al. [
164] using a Cooperative Fuel Research (CFR) engine showed that ammonia could be used up to 70-30% ammonia-gasoline mixtures, although under idle conditions, the engine needed to be operated with gasoline as single fuel. Ryu et al. [
165] indicated that gaseous injection of ammonia is preferred because of the high value of the latent heat of vaporization (1370 kJ/kg) compared with that of gasoline (348.7 kJ/kg), causing a significant reduction in in-cylinder temperature. In addition, the low flame speed of ammonia translates into an increment in combustion duration, thus requiring a greater advancement in spark timing than that applied for typical SI engines [
166].
An alternative to avoid using two separate storage fuel systems is to generate hydrogen from the ammonia decomposition. The cracking of ammonia has been studied by several authors [
167,
168,
169], proposing the use of thermal catalytic reactors that allow reducing reaction temperature. Catalysts commonly employed contain noble metals (Rh, Ru, Pd, Pt, among others), but also Fe and Ni have been tested in order to reduce costs and dope the catalyst with rare earth metals to increase activity or using bimetallic configurations [
170,
171]. The industrial application of this technology at large scale has been reviewed by Spatolisano et al. [
172], analyzing implications of technology deployment. Many technologies currently developed are still in the experimental stage; thus, information regarding long term performance, economics and specific environmental impacts are scarce [
173]. Testing at a pre-commercial stage is necessary for technical maturity before large-scale implementation.
The ammonia cracking reaction is endothermic. Therefore, onboard applications may benefit from using exhaust combustion gases, which can provide the heat needed to release hydrogen, but at the cold start of the engine, external heaters are necessary. Comotti and Frigo [
174] proposed using a Ru-based catalyst with electric heaters in the internal section of the reactor, and the outer section used the exhaust gases as heat supply. Their system allowed the use of ammonia as a single fuel in SI engines operating successfully under different loading conditions. One crucial aspect to consider is that when hydrogen is produced from ammonia cracking, the extra amount of N
2 introduced into the cylinder should be taken into account to optimize operating conditions.
Table 2 shows some of the experiences carried out using ammonia as fuel in spark ignition engines. The use of ammonia as the only fuel in a spark ignition engine is practically unfeasible due to the difficulty of ignition, low flame stability, combustion efficiency, and power delivery, so dual fuel configuration seems the best option. Using hydrogen as a dual fuel in proportions between 10-15% allows for solving these problems [
160,
161]. The required hydrogen may be obtained from an additional storage tank or the partial dissociation of ammonia; in any case, it is necessary to use systems such as EGR to reduce ammonia and NO emissions [
175]. Other fuels have also been evaluated in dual configuration with ammonia. Tutak et al. [
176] used DME as a complement to ammonia, allowing stable operation to be achieved. Oh et al. [
177] reported that using DME along with methane, slightly decreased braking power.
4.3. Compression Ignition Engines
Compression ignition engines apply a higher compression ratio and thus achieve fuel ignition. Since the fuel is directly injected into the combustion chamber after compression, knock limitations are eliminated, fuel efficiency is high, and fuel consumption can be optimized to provide power. However, the direct injection of the fuel causes diffusion combustion, which translates into high localized temperatures, increasing the emission of NOx and particles. For this reason, it is said that CI engines, although highly efficient, have an emission problem [
182]. Fortunately, advances in engine design and emission abatement technologies have guaranteed that post-combustion gases are adequately treated. Exhaust gas recirculation decreases combustion temperature, thus reducing thermal NOx production. Lean NOx trap technology has been developed to trap and reduce NOx compounds under lean operating conditions. Selective catalytic reactors use a nitrogen compound to transform NOx into N
2, and particle filters are incorporated to remove soot [
183,
184]. Different technological developments have allowed diesel engines to keep playing a fundamental role in energy production and transport. The high energy density of diesel fuel (35.94 MJ/L – [
185]) makes it the ideal choice for heavy road transport, but it also makes difficult the task of finding a suitable alternative due to the low storage volume needed for on-board applications and the high efficiency reached by these type of engines.
After the initial work of Cornelius et al. [
156], Gray et al. [
186], and Starkman et al. [
187], the experiences of Pearsall and Garabedian [
188] focused on using ammonia in diesel engines. However, in this case, dual-fuel engines were the best alternative for attaining an acceptable output by controlling ignition thanks to the injection of diesel fuel. Reiter and Kong [
189] tested a dual-fuel (ammonia-diesel) engine and reported that NOx emissions did not increase, despite the presence of ammonia, as long as energy substitution attained by ammonia did not reach 60%. However, in a later study, the same authors [
190] lowered the ammonia energy supply in the mixture below 40%. Otherwise, NOx emissions would increase significantly due to fuel bound emissions. These authors also reported high ammonia slip (1000 – 3000 ppmv) in exhaust gases, which may be regarded as disappointing news for this technology.
The experimental and demonstration work have been extensive in the last decade due to the impact of greenhouse gas emissions and the feasibility of using a non-carbon gas for which industrial technology is already available. Reviews about the use of ammonia as fuel in compression ignition engines have been carried out by Dimitriou and Javaid, [
45], Kurien and Mittal [
22], and Voniati et al. [
191] where it is reported the need to install several after-treatment devices for emission control and this requirement also limits the application to marine, power generation, and heavy-duty applications. In addition, the formation of N
2O, which has a much greater global warming impact, is one of the main obstacles to surpass when considering the use of ammonia in ICEs [
192].
The variety of options currently being considered to reduce CO
2 emissions may add optimism and give hope to a future free of carbon fossil fuels, but it requires significant effort from researchers and industry. It also comes along with an adverse effect by creating uncertainty regarding the future of the dominant technology and the limited space available in fuel stations and cargo ports since this variety of options also implies having installations for electrification, LPG, hydrogen, ammonia and methane storage, which has been recently incorporated to the maritime transport sector, thus resulting in an excessive complexity of infrastructure. Space limitations can make it unfeasible for different technologies to coexist. As an example, Imhoff et al. [
193] studied the feasibility of implementing ammonia power trains in current vessels, indicating that redesign and reconfiguration efforts are needed but achievable; thus, ammonia-fuel ships may represent a feasible alternative. However, the maritime transport sector needs the certainty that an investment, which may be extremely high, will stay long enough to be worth it.
Table 3 lists different studies dealing with the use of ammonia as fuel in ICEs. The risks of using anhydrous ammonia and pressurization tanks may be avoided if other forms of ammonia are considered. Schönborn [
194] proposed the use of water ammonia solution in diesel engines. This way, ammonia can be stored in liquid form under ambient conditions, reducing the risk of vapor ammonia release. The author analyzed the behavior of a diesel engine when injecting a 25% ammonia solution as fuel using numerical simulation. The compression ratio needed to be increased to 27 to attain ignition of the ammonia solution since the high water content creates a dilution effect, making ignition possible only if higher energy was added. However, the dilution effect also allowed reducing NOx emissions. Therefore, in addition to facilitating storage, other benefits are associated with this approach, whereas improving ignition may be achievable by adding specific promoters such as hydrogen.
Hydrogen has also been widely studied as an ignition promoter when testing compression ignition engines, but also methanol, ammonia nitrite, and diesel fuel (n-heptane in simulation runs) [
199,
200,
201,
202]. Although modifications in fuel injection strategies are needed, ammonia-diesel dual fuel engines may reduce CO
2 emissions by 90% and NOx emissions by 70% compared with original diesel engines [
203]. Other organic solvents may also be proposed as a liquid phase for storing ammonia without causing detriments in combustion characteristics, as for the case of water, with methanol being considered a better candidate [
204]. Other alternatives proposed the use of spark-assisted ICEs to overcome poor ignition characteristics [
205] or changing the injection strategy by allowing the entrance of a small quantity of ammonia that will aid in main fuel ignition, thanks to the pre-combustion stage during the compression cycle, thus increasing temperature and pressure and facilitating ammonia ignition [
206]. Scharl and Sattelmayer [
207] studied the presence of diesel piloted injections and concluded that reliable ignition of ammonia was possible when adding a small fraction of energy as piloted injected fuel (3.2% of LHV based on loading conditions).
5. Conclusions
The aim of a net zero carbon emission by 2050 is a desirable goal, but this target needs to be carefully evaluated since implications associated with the decarbonization of the economy bring along an increase in costs, which may create market distortions. Several alternatives are currently under development for reducing the use of fossil fuels. However, many of these options result in an increase in process energy demand. Electrification and electricity production from renewable sources aid in achieving decarbonization goals. Impacts regarding mineral extraction, battery recycling and uncertainty in renewable energy production need careful evaluation. Given the immediacy of the targets, it seems more reasonable to concentrate efforts on well-known and worldwide applied technologies.
Ammonia as an alternative fuel has the main advantage of counting with great experience in large-scale production, storage and handling. Although several aspects regarding emissions need to be addressed, vast experience exists in technologies capable of mitigating combustion pollutants. Another matter that may be a cause of concern is the effect of massive ammonia use for other applications besides agriculture. The expected increase in price resulting from the higher costs in production chains caused by the adoption of green hydrogen in the manufacturing process, as well as the pressure induced by a growing ammonia demand, are factors that should not be overlooked. These factors can have a significant impact on human well-being and the food supply chain.
There is still a long way to go before attaining a feasible solution to decarbonizing the economy. The use of ammonia may be one of them, but in the short term, it seems more reasonable to focus on reducing CO2 emissions by reducing fuel consumption and optimizing current technologies rather than forcing a complete shift towards electrification of the transport sector, which requires adaptation to a completely different technology, raising costs without the certainty of how production cycles and critical materials will be managed. Electrification can help reduce pollution in some urban regions. However, other alternatives, such as hybridization and current highly efficient combustion engines, can be more effective in achieving a global reduction in emissions by already having global technological deployment.