Refractories for Ammonia Production in Fertilizer Unit

Abstract

Apart from being used as a fuel, natural gas is used extensively for production of ammonia-based fertilizers. During the process of ammonia production natural gas is steam reformed for the production of Hydrogen and the same is converted into Ammonia, by Haber’s process, using nitrogen from air. Refractories are required for reformer lining since they are operated at high temperatures as well as in corrosive gas, primarily Carbon Monoxide and Hydrogen, environment. The refractories selected for reformer, thus, should resist the reformer operating temperature as well as the aforementioned gases. Owing to the presence of steam in the working environment magnesia and lime-based basic refractories cannot be used owing to their hydration tendency and thus, aluminosilicate refractories are the only choice. The effect of H2 and CO on aluminosilicate refractory is the primary focus of this paper. The main concerns are the reduction of siliceous components of the refractory by hydrogen and carbon deposition due to Carbon monoxide decomposition by Boudouard reaction. The effect of these gases on aluminosilicate refractories have been reviewed and based on the outcome suitable refractories have been recommended for ammonia production.

1. Introduction

Ammonia is one of the major raw materials for fertilizer industries. It is used for the production of nitro compounds, urea, ammonium sulfate, ammonium phosphate etc. The best and cheapest available source of nitrogen, for ammonia production, is air. Whereas hydrogen, required for ammonia production, is produced from various feedstocks but currently it is derived mostly from fossil fuels. Natural gas, Naphtha, Fuel oil, coal is used for the production of hydrogen. Natural gas is the most preferred option since it is the cheapest in terms of relative investment as well as relative specific energy requirements for ammonia production.

Any industrial process involving high temperature requires refractory. Since the production of Ammonia from natural gas, Ammonia-based fertilizer as well as its derivative involves high temperature, their production, thus, also requires refractory. The literal meaning of refractory is “Stubborn.” In the context of industrial processes, refractories connote the materials which are not markedly affected by their environment. In other words, refractories retain their original features as well as characteristics in aggressive industrial process conditions. During their usage, refractories are exposed to:

• Chemical attack, that is, interaction with the liquid as well as gases

• Abrading action

• Periodical heating and cooling cycle leading to mechanical stress, and/or

• Continuous high temperature, etc.

Against this backdrop the selected refractories, for any industrial process, should ensure that they do not undergo chemical degradation and also can withstand abrasive actions and mechanical as well as thermal stresses of the environment. In this context it also is prudent to mention that no refractory is everlasting. The primary objective is to select a refractory, for any industrial process, such that the impact of the aforementioned industrial process conditions is minimal and the refractory life is maximized. Performance of the refractory is one of the major determinants of the economic efficiency of virtually all industrial processes.

In this context it should also be mentioned that all the aforementioned refractory wearing parameters are not important or significant for all the industrial processes and as a consequence all the refractory properties are not relevant for a given process environment. For the requisite or best performance, the refractory wear contributors, for a given industrial process, need to be identified and refractories should be selected such that it can withstand the identified critical wear parameters. The relevant refractory properties, which would counter the critical wear contributors, need to be optimized. Identification of critical wear parameters, hence, is one of the most important steps for maximization of refractory life at the lowest cost. Apart from the refractory properties, operating parameters of the industrial units also play a key role in determining the refractory performance.

In this paper a brief production process of Ammonia, from natural gas, would be presented (Figure 1) and the emphasis would be on the refractory selection process for the ammonia production unit. The primary focus of this discussion would be to understand the operating condition at each step of ammonia production. Once the operating conditions are identified, rationale behind recommending refractories for the various units of the ammonia production process would be discussed. Needless to mention, the primary objective of this paper is to correlate the operating conditions with refractory properties, not the impact of process parameters on efficiency of Hydrogen and thus, ammonia production. For improving the process efficiency catalysts are used at all stages of hydrogen production as depicted in Figure 1. Catalysts and their impact on the Hydrogen / Ammonia production process is also not part of this paper. It is evident from the flow diagram (Figure 1) that at different stages of the process, part of the hydrogen yield is used as reactant. Prior to venturing into the hydrogen production process, basic information on refractory material would be shared and this would set the backdrop of the rationale behind the refractory recommendations for hydrogen/Ammonia production from natural gas.

Figure 1.

Process diagram for ammonia production.

1.1 Refractory

Chemically, refractories can be classified as Acidic, Basic and Neutral. The chosen refractory should be compatible with the chemical environment of the process equipment. For example, in case the environment of a process equipment, during operation, is chemically basic, a basic refractory needs to be selected for its lining. Based on the shape, refractory can be classified as shaped, which colloquially is known as bricks, and monolithic refractories. Unlike bricks, monolithics do not come with any specific shape. Shaping of the monolithic materials is done during installation as per the contour of the process equipment. A large number of refractory products fall in the monolithics category. Castable, which is akin to concrete with different material bases and cement, constitutes the largest volume of monolithics. Apart from castables there are numerous monolithic materials, which are installed mechanically, that is, their usage enhances the refractory installation rate, and thus, assists in reduction of process equipment downtime.

Monolithic refractories, in general, have the following advantages over shaped ones.

• Monolithic refractories do not require shaping or firing, and thus, can be produced in a very short notice

• Any shape can be given to monolithics during their installation, i. e. high shaping flexibility

• Since no firing is involved for monolithic production, energy requirement for their production, hence, is lower

• Since the installation of monolithics can be mechanized, their installation rate is significantly faster

• Mechanized installation of monolithics eliminates human error

• With the developments on the material front, monolithic refractories virtually have the same chemistry as bricks. This implies monolithics can be used in all industrial applications with the same operational efficiency as well as performance of bricks

Morphological features of refractories are schematically illustrated in Figure 2. All classes of refractories consist of granular material of various sizes, which are termed as aggregates and they are bonded together by very small sized material, which is designated as matrix. Refractories can be conceptualized as the matrix being the continuous phase, where the aggregates are embedded. In majority of the cases, aggregate and matrix chemistry are very different from each other. This makes refractories a heterogeneous material and thus, more complex. Apart from aggregate and matrix, pores are an inherent constituent of the refractories. Refractory porosity and its size distribution can be varied by controlling the proportion of aggregate of different sizes used in the refractory formulation.

Figure 2.

A schematic representation of refractory showing the heterogeneity of the product.

Figure 3 illustrates different kinds of pores present in the refractories. Channel pores are the ones which are open from both the ends, though the connecting path may be tortuous. Open pores are the ones which are open from one end but closed from the other. Closed pore, as the name suggests, is closed from all around, i. e. it is not accessible by any gas or liquid in contact with the refractory.

Figure 3.

Schematic representation of types of pores in a refractory.

Pore concentration as well as its size distribution in the refractories virtually governs its mechanical as well as thermal properties. For example, channel pores are the only ones which contribute to gas permeation. As the combined concentration of channel and open pores grows higher, the vulnerability of the refractories to chemical attack increases. Total concentration of all 3 types of pores determine the thermal conductivity as well as the strength. Pore concentration of refractories also determines their abrasion resistance, elasticity etc.

Selected refractories for any industrial application should conform to the requisite pore structure, its concentration as well as size. For example, for low thermal conductivity, pore concentration of the refractories should be high. On the contrary, for high strength, good resistance to abrasion as well as chemical attack, the refractory porosity should be low. So, not only the chemical compatibility of the refractories with the operating environment ensures the desired performance but its morphological features also play a significant role.

It is, thus, prudent that against the backdrop of the operating conditions of each of the units for Ammonia production, refractory recommendation is made. In the next section, hence, the Ammonia production process as well as the influence of the process parameters on the process efficiency would be discussed in brief. This will set the tone as well as backdrop of refractory selection for the fertilizer industry, particularly for ammonia production units.

2. Natural gas conversion into ammonia

No dearth of information is available in the literature for commercial ammonia production by Haber - Bosch process [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]. Among fossil fuels, Natural gas is the most preferred option for Hydrogen generation for Ammonia production. It is well known that the major constituent of Natural Gas is Methane. Apart from Methane, natural gas also contains hydrocarbons of higher molecular weight. Higher hydrocarbons are usually converted into methane by hydrocracking (Eq. 1) prior to further processing.

$${\mathrm{C}}_{\mathrm{n}}{\mathrm{H}}_{\mathrm{m}}+\left\{\left(\text{4n}-\mathrm{m}\right)/2\right\}{\mathrm{H}}_2={\mathrm{nCH}}_4$$E1

The hydrocracking reactions are endothermic. The reactions are carried out at (65–140 bar) and (400–800°C), in the presence of hydrogen.

The subsequent step for hydrogen generation is Sulfur removal from natural gas. Sulfur bearing compounds in natural gas need to be removed since it deactivates the catalysts used in the Ammonia production process. Sulfur, which is present in the natural gas as Thiol – Sulfur (RSH), is removed by catalytic hydrogenation and in the process hydrogen sulfide is generated (Eq. 2).

$${\mathrm{H}}_2+\mathrm{RSH}\rightleftarrows \mathrm{RH}+{\mathrm{H}}_2\mathrm{S}\left(\mathrm{gas}\right)$$E2

Hydrodesulfurization reaction occurs at 300 to 400°C and 30 to 130 bar absolute pressure. H₂S generated by this process is passed through beds of zinc oxide yielding zinc sulfide (Eq. 3).

$${\mathrm{H}}_2\mathrm{S}+\mathrm{ZnO}\rightleftarrows \mathrm{ZnS}+{\mathrm{H}}_2\mathrm{O}$$E3

Desulphurized methane is treated with high-temperature steam (700–1000°C) at 3–25 bar pressure, in the presence of a Nickel catalyst in Primary Reformer. The following reactions (Eqs. 4 and 5) occur during interaction of Methane with steam.

$${\mathrm{CH}}_4+{\mathrm{H}}_2\mathrm{O}\rightleftarrows \mathrm{CO}+{\text{3H}}_2\mathrm{\Delta H}=+206.1\mathrm{kJ}/\mathrm{mol}$$E4

$${\mathrm{CH}}_4+{\text{2H}}_2\mathrm{O}\rightleftarrows {\mathrm{CO}}_2+{\text{4H}}_2\mathrm{\Delta H}=+165.0\mathrm{kJ}/\mathrm{mol}$$E5

All the reforming reactions are endothermic. The primary objective of the steam reforming process is to maximize Hydrogen yield. CO and CO₂ are the byproducts of the steam reforming process. Apart from the steam reforming reactions, Dry Reforming and Water Gas Shift (WGS) reactions (Eqs. 6 and 7) also occur by virtue of interactions between the steam reforming reaction products, viz. CO and CO₂, steam and Methane.

$${\mathrm{CH}}_4+{\mathrm{CO}}_2\rightleftarrows \text{2CO}+{\text{2H}}_2\mathrm{\Delta H}=247.3\mathrm{kJ}/\mathrm{mol}$$E6

$$\mathrm{CO}+{\mathrm{H}}_2\mathrm{O}\rightleftarrows {\mathrm{CO}}_2+{\mathrm{H}}_2\mathrm{\Delta H}=-41.15\mathrm{kJ}/\mathrm{mol}$$E7

WGS is called so since by virtue of this reaction the ratio of Water Gas constituents, viz. CO, and H₂, are altered. In this specific case, the reaction is carried out in such a way that the reaction proceeds in favor of Hydrogen generation. The yield of Primary reformers typically contains 60% Hydrogen. All the reactions in Primary Reformer are reversible in nature. Owing to this reason pressure, temperature, and ratio of the reactants determine the extent of Hydrogen generation. Figures 4 and 5 illustrate the impact of Steam - Methane ratio, temperature, and pressure on the reforming process. The observations are in the expected line of the reversible reactions and thus, the conversion of natural gas into CO and H₂ increases with:

• Increase in temperature

• Increase in steam to methane ratio, and

• Decrease in pressure

Figure 4.

Effect of pressure on the steam reforming process in primary reformer for the operating temperature of 800°C [6].

Figure 5.

Effect of temperature on the steam reforming process in primary reformer for operating pressure of 30 Bar [6].

In this context it is prudent to mention that Nickel catalyst, which is used for steam reforming reaction, also activates Boudouard reaction (Eq. 8) [6]. Since this reaction is reversible in nature, higher pressure favors the reaction to proceed to the right, that is, higher pressure favors CO decomposition.

$$\text{2CO}\rightleftarrows \mathrm{C}+{\mathrm{CO}}_2$$E8

By virtue of this reaction Carbon deposition tendency is fairly severe. This reaction has a significant impact on the life of refractories as well as its performance, which would be discussed in the refractory selection section.

The yield of theprimary reformer, that is, Hydrogen (H₂), unreacted CH₄, unreacted water (steam), CO, CO₂, Nitrogen, Argon, etc. are fed in the secondary reformer, via Transfer Line, for further processing to increase the Hydrogen yield. Air, apart from the yield of the Primary reformer, is a feed for the secondary reformer. The primary objective of the Secondary reformer is to complete Steam Reforming as well as WGS reactions to produce further quantities of Hydrogen and adjust the hydrogen and nitrogen ratio for Ammonia production.

Apart from the continuation of steam reforming, dry reforming, and WGS reactions (Reactions 4–7), the following reactions also take place in Secondary Reformers.

$${\text{2CH}}_4+{\mathrm{O}}_2=\text{2CO}+{\text{4H}}_2$$E9

$${\text{2CH}}_4+{\text{3O}}_2=\text{2CO}+{\text{4H}}_2\mathrm{O}$$E10

$${\mathrm{CH}}_4+{\text{2O}}_2={\mathrm{CO}}_2+{\text{2H}}_2\mathrm{O}$$E11

In addition to these reactions combustion of Hydrogen as well as Carbon Monoxide also occurs. The exothermic reactions raise the temperature of gases in the secondary reformer to ∼1000°C and the conversion to hydrogen achieved is of the order of 99% and the rest is a mix of CO, CO₂, CH_4, and water in chemical equilibrium.

The yield of the secondary reformer is further treated in the WGS reactor to increase the Hydrogen concentration (Eq. 7). Owing to the exothermic nature of the WGS reaction, lower temperature favors the formation of Hydrogen (Figure 6). The yield of secondary reformers is processed in 2 stages, viz. High-temperature WGS reactor at 300–450°C and Low - temperature WGS reactor at 200–250°C.

Figure 6.

Effect of temperature on water gas shift (WGS) reaction equilibrium [9].

All oxygen-containing substances, including water, are poisons for the ammonia synthesis catalyst and thus, need to be removed. The gas mix from the WGS reactor, hence, is further processed, for removal of CO and CO₂ by Pressure Swing Adsorption (PSA) or Cryogenic Distillation (CD). After the bulk removal of CO by WGS reaction and CO₂ removal by PSA or CD, the typical synthesis gas still contains 0.2–0.5 vol % CO and 0.005–0.2 vol % CO₂.

Methanation (Eqs. 12 and 13) is the simplest method to reduce the concentrations of the carbon oxides well below 10 ppm and is widely used in ammonia production units. There are two main purposes for methanation, viz. to purify synthesis gas, i. e. remove traces of carbon oxides, and to manufacture methane. CO and CO₂ methanation is carried out in the temperature range of 200–600°C and 350–450°C, respectively.

$$\mathrm{CO}+{\text{3H}}_2\rightleftarrows {\mathrm{CH}}_4+{\mathrm{H}}_2\mathrm{O}$$E12

$${\mathrm{CO}}_2+{\text{4H}}_2\rightleftarrows {\mathrm{CH}}_4+{\text{2H}}_2\mathrm{O}$$E13

The hydrogen - nitrogen mix, obtained after methanation, is used for Ammonia production by the Haber - Bosch process.

3. Refractories of ammonia production

It is apparent from the previous section that the environment of all the units for ammonia production contains steam, since it is one of the reactants used in ammonia production. MgO or CaO based, that is, basic, refractories, thus, would be unsuitable since they are prone to hydration. The hydration of both CaO and MgO yields hydroxides. The formation of these oxides is accompanied by large volume expansion. As a result, the soundness of the basic refractories in the ammonia production unit would be destroyed and thus, are unsuitable.

Against this backdrop, it can be concluded that chemically only aluminosilicate refractories would be suitable. Aluminosilicate refractories, as the name suggests, are based on Alumina (Al₂O₃) and Silica (SiO₂). One compound, viz. Mullite (3Al₂O₃.2SiO₂), which contains 72% Al₂O₃, is the only thermally stable binary phase in the aluminosilicate system. With the increase of alumina content of aluminosilicate refractories, their refractoriness, i. e. temperature withstanding capability, increases. Apart from alumina and silica, aluminosilicate refractories also contain certain impurities, which are inherently present in the natural aluminosilicate raw materials used for their production. Primary impurities in aluminosilicate refractories are iron oxide, titanium dioxide, alkali, and alkaline earth oxides and these minor constituents play a significant role in deciding the ultimate refractory performance in any given environment and hence, the impact of these impurities also should be evaluated.

3.1 Impact of hydrogen on aluminosilicate refractories

The effect of Hydrogen gas on fused silica and a wide range of aluminosilicate refractories has been studied in depth [15, 16, 17, 18, 19, 20]. The primary effect is reduction of silica by hydrogen gas as per Eq. 14 [16].

$${\mathrm{SiO}}_2\left(\mathrm{s}\right)+{\mathrm{H}}_2\left(\mathrm{g}\right)\begin{array}{l}\stackrel{\text{Higher Temperature}}{\to }\\ \underset{\text{Lower Temperature}}{\leftarrow }\end{array}\mathrm{SiO}\left(\mathrm{g}\right)+{\mathrm{H}}_2\mathrm{O}\left(\mathrm{g}\right)$$E14

The gaseous products generated by virtue of reaction 14 are carried off by the process stream. Downstream, when the temperature is conducive for solidification of SiO, it condenses and gets deposited as SiO₂ and Si mix causing heat-exchanger fouling and product contamination. Figure 7 illustrates the effect of time as well as temperature on the reduction of fused silica by Hydrogen gas. As expected, with the increase in duration of fused silica and hydrogen interaction the loss of silica increases. The effect is similar when the interaction temperature increases.

Figure 7.

Effect of time and temperature on fused silica reduction [15].

The effect of time and temperature in the hydrogen environment on aluminosilicate refractories of different silica concentration follow the similar trend. Figure 8 illustrates that for a given level of silica in aluminosilicate refractories, increase of temperature causes higher loss of silica. It is also seen that for a given temperature, silica loss increases with the increase of silica concentration in aluminosilicate refractories.

Figure 8.

Effect of temperature on reduction of refractories with different silica content after reduction for 33 hrs in 100% H₂ atmosphere [16].

Figure 9 illustrates the effect of time on the silica loss of aluminosilicate refractories with different silica content. As contemplated, the increase of refractory - hydrogen interaction duration causes greater silica loss. It also is in the expected line that for a given refractory - hydrogen interaction duration, silica loss increases with its silica concentration. But it is observed that upto 10% silica concentration in the refractory, the silica loss by SiO₂ reduction is marginal.

Figure 9.

Effect of time on reduction of refractories with different silica content for reduction by hydrogen at 2600°F in 100% H₂ atmosphere [17].

Figure 10 reports the impact of pressure as well as silica concentration of refractories in the hydrogen environment at 2400°F. Loss of silica, for all levels of pressure, increases with increase of silica concentration. It is fairly evident from the results that for a given silica concentration, the silica loss decreases with the increase of hydrogen gas pressure. In other words, higher operating pressure would protect the siliceous part of aluminosilicate refractories from the Hydrogen present in reformers. The operating pressure of primary as well as secondary reformers is >30 bar and thus, the silica loss from the aluminosilicate refractories is expected to be low.

Figure 10.

Effect of hydrogen pressure on reduction of refractories with different silica content. These data are for 2400°F and hydrogen flow of 4.6 liters/minute [17].

3.2 Impact of carbon monoxide on aluminosilicate refractories

The effect of Carbon Monoxide (CO) on the refractories is attributed to its decomposition as per the reverse Boudouard/Bell reaction (Eq. 8) [21, 22, 23, 24, 25, 26, 27, 28]. In the context of refractories, it can be stated that CO gas diffuses into open as well as channel pores. By virtue of the reverse Boudouard reaction, carbon is deposited in the pores. Deposited carbon grows and generates stress within the refractories. When the stress exceeds the strength of the refractories, cracks form and the refractories get damaged. In the worst possible scenario, the refractory is destroyed. Figure 11 illustrates the stability of CO as a function of temperature. With the increase in temperature, the stability of CO increases. It is apparent from the figure that destruction of refractories, by deposition of carbon, will not occur when the temperature exceeds 1100°C and at <500°C, though CO decomposition is thermodynamically possible, the reaction kinetics is slow and hence, CO decomposition rate is low. Refractories, thus, are vulnerable to destruction by CO in the 500–700°C range.

Figure 11.

CO - CO₂ equilibrium as per FACTSage software [21].

Metallic iron is a known catalyst for CO decomposition [27, 28]. The CO decomposition process proceeds by reduction of iron oxides by CO, to metallic iron. The iron subsequently is carburized into Fe - carbide, which decomposes into Fe according to Eq. 15.

$${\mathrm{Fe}}_3\mathrm{C}=\text{3Fe}+\mathrm{C}$$E15

The Fe, formed via Iron Carbide decomposition, has high surface activity and hence, enhances the CO decomposition rate. The catalytic activity is believed to proceed via the formation of Iron Carbide (Fe₃C).

3.3 Impact of simultaneous presence of hydrogen and CO on aluminosilicate refractories

Apart from reverse Boudouard reaction, reverse water gas reaction also contributes to carbon deposition in refractories as per Eq. 16. This reaction, however, can proceed only below 680°C [29]. Beyond this temperature Carbon monoxide and Hydrogen are stable phases and hence, carbon deposition within the refractory is not expected. This prediction, however, is based on the assumption of standard states, i. e. the activity of the reactants and the products are 1. In real situations the reactant as well as product activities would be less than 1, which means there would be certain deviations in the temperature predicted above.

$$\mathrm{CO}+{\mathrm{H}}_2\rightleftarrows \mathrm{C}+{\mathrm{H}}_2\mathrm{O}$$E16

Simultaneous presence of H₂ and CO, which is the case for the reformers, enhances the CO decomposition rate when the temperature exceeds 577°C [23]. The presence of H₂, in the CO environment, not only alters the carbon deposition rate but also determines the temperature at which the maximum carbon deposition occurs. For example, for 0.8% and 19.9% H₂, maximum Carbon deposition occurs at 530 and 630°C, respectively. At lower temperature, however, the effect of hydrogen is marginal [28]. It, thus, is a precondition that refractories for the CO environment should be low in their iron oxide content.

3.4 Impact of steam on aluminosilicate refractories

Steam, in general, appears to be inter towards the oxide refractory materials. Steam, however, interacts with silica as per the Eq. 17, which is reversible in nature. It is evident from this equation that SiO₂ in the presence of steam is converted into Si(OH)₄.

$${\mathrm{SiO}}_{2\left(\mathrm{s}\right)}+{\text{2H}}_2{\mathrm{O}}_{\left(\mathrm{g}\right)}\rightleftarrows \mathrm{Si}{\left(\mathrm{OH}\right)}_{4\left(\mathrm{g}\right)}$$E17

Figure 12 illustrates the effect of temperature as well as time, on the weight loss of cristobalite, at 0.84-atmosphere steam pressure. The magnitude of silica loss, via Si(OH)₄ formation, increases with the increase of temperature up to 1350°C. Beyond 1350°C, however, the rate, as well as the magnitude of silica loss, reduces. This observation has been attributed to the reversal of Eq. 17. The opinion related to the temperature impact on Si(OH)₄ formation in aluminosilicate refractories, however, appears to be different from that for pure silica. It is also believed that at <980°C steam reacts with siliceous components of aluminosilicate refractories and yields Si(OH)₄. At >1000°C, however, the same reactants primarily yield SiO¹⁹.

Figure 12.

Effect of steam temperature on the loss of silica by 0.84-atmosphere pressure [17].

Figure 13 illustrates the impact of steam pressure on cristobalite weight loss by virtue of Eq. 17. It is obvious from the figure that with the increase of pressure, the silica loss increases. The silica loss is related to pressure by Eq. 18 [17].

Figure 13.

Effect of steam pressure on the volatilization of - 325 mesh 24 gm Cristabolite in 2 hours at 900°C. the steam flow rate was maintained at 2.73 cm/second [17].

$$\text{Weight Loss},\%=2.02\mathrm{x}{10}^{-3}{\mathrm{P}}^{1.34}$$E18

The index of pressure in Eq. 18 is >1 but <2. Indices of 1 and 2 indicate the reaction is controlled by transportation through the boundary layer and reaction at the interface, respectively. The mechanism of Si(OH)₄ formation, by virtue of Silica - steam interaction (Eq. 17), thus, is not very clearly defined.

Figure 14 illustrates the impact of steam, on the weight loss of silica in aluminosilicate refractories, in the hydrogen environment. As has been observed earlier, the weight loss increases with the increase in silica content of the refractories. When steam is present in the environment, along with hydrogen, the rate of silica reduction is reduced significantly. Additionally, silica loss shows only marginal dependence on the silica content of the refractory, when steam is present in the hydrogen environment. In short, the impact of hydrogen on silica reduction is reduced significantly when steam and hydrogen are present simultaneously. This implies that the presence of steam in the working environment, which will be the case in both the reformers, would protect the aluminosilicate refractories better by reducing the reductant effect of hydrogen.

Figure 14.

Effect of steam on reduction of refractories with different silica content in a hydrogen atmosphere at 2400°F [18].

3.5 Impact of simultaneous presence of hydrogen and steam on aluminosilicate refractories

Figure 15 illustrates the impact of the simultaneous presence of Hydrogen and steam on silica loss of various aluminosilicate refractories. It is evident from the figure that the silica loss in the hydrogen atmosphere comes to a rest, for the aluminosilicate refractories, in the presence of steam. In other words, the reduction of silica by hydrogen ceases when Hydrogen and steam are present simultaneously. The corollary of the same is the adverse impact of steam, on aluminosilicate refractories, is annulled by the presence of hydrogen.

Figure 15.

Impact of steam on the reduction of aluminosilicate refractories at 2500°F by hydrogen [18].

3.6 Impact of gaseous environment on the refractory characteristics

Reduction of the siliceous component, by hydrogen, increases the refractory porosity and increased porosity adversely affects the refractory strength [15] Figure 16 illustrates the impact of silica loss on the strength of 52% alumina bricks. As expected, the increase in silica loss leads to a greater loss of strength. A loss of 10% silica causes approximately 50% strength reduction.

Figure 16.

Effect of silica loss on strength of refractories. The reduction was carried out in the hydrogen atmosphere [16].

As has been seen in the earlier section, the reduction of silica by hydrogen is inhibited by the presence of steam. The reduction of refractory strength, hence, is lesser in the presence of steam, compared to when Hydrogen is present by itself (Figure 17). It is evident from the illustration that the reduction of strength for an aluminosilicate brick is lower by approximately 20% when the hydrogen environment contains steam. In fact, during ammonia production steam is always present in the operating environment of all the units. The impact of hydrogen, as a reductant thus, is expected to be lower in all the units of the ammonia production process.

Figure 17.

Impact of 25 mm steam on strength reduction of 85% alumina brick in a hydrogen atmosphere at 2500°F [16].

Figure 18 illustrates the impact of alumina as well as iron content of the refractories in the CO environment. It is evident that CO by itself gets decomposed into C and CO₂ as per the reverse Boudouard (Bell) reaction (Eq. 8). This is reflected by the reduction of the strength of Fe - free 45 as well as 90% alumina containing refractories with an increase in interaction time with CO. It also is obvious from Figure 15 that Fe acts as a catalyst for the decomposition of CO. In the presence as well as the absence of iron in the refractories, the strength reduction rate of 50% alumina refractory is faster, compared to the one containing 90% alumina [22]. 50% Alumina refractory is destroyed approximately 5 times faster than 90% alumina products. Lower alumina refractories, thus, are more vulnerable to disintegration due to CO decomposition.

Figure 18.

Impact of 1 Bar carbon monoxide atmosphere on strength for refractories with different alumina and iron content [21].

Literature reports enhancement of CO disintegration in the presence of steam [26]. Figure 19 illustrates that the strength of White Tabular Alumina (WTA) based castable reduces in CO - Steam atmosphere. On the other hand, it is observed that the presence of Nitrogen, in the CO environment, enhances the strength of WTA-based castable, that is, the trend is reversed when Nitrogen is replaced by steam. Nitrogen is an inert gas and by itself, it does not react with aluminosilicate refractories. But when it is present together with CO gas, it alters the interaction process with the refractories and enhances its strength. In fact, Nitrogen is present in the operating environment in secondary reformers and downstream. The adverse effect of steam on castable strength, in the CO environment of the ammonia production unit, is nullified to some extent owing to the simultaneous presence of nitrogen.

Figure 19.

Effect of CO - steam on the strength of WTA based Castable at 32 bar pressure [26].

3.7 Ammonia plant operating condition

The operating conditions of ammonia production units can be summed up as follows:

• The gases present in the working environment of all the units are Hydrogen, CO, CO₂, Steam, Nitrogen, and Methane.

• Hydrogen concentration increases downstream, whereas the same for CO, CO₂, steam, and CH₄ decreases.

• Typical operating temperature and pressure of primary reformer is 980°C and 35 bar, respectively.

• Whereas the same for secondary reformers is 1350°C and 35 bar, respectively. Such high temperature in secondary reformers is observed only around the combustion region. The temperature of the other sections of secondary reformer typically is 1100°C

• Operating temperature and pressure of Transfer Line, which allows the passage of primary reformer yield to secondary reformer, are in the vicinity of those in the reformers

• The high temperature shift converter operates at 480°C, whereas the operating temperature of low temperature shift converter is 230°C

• Methanator usually operates at <350°C

• The typical operating temperature of ammonia production unit, thus, does not exceed 1100°C, except around the burner in secondary reformer, where the temperature is ∼1350°C

• Operating temperature of each unit in the Ammonia production plant is different. The operating temperature of a given unit, however, does not fluctuate. Additionally, the operation is continuous. These 2 aspects imply that no temperature fluctuation occurs in a given unit of ammonia production plant, that is, there is no thermal shock on the refractories.

• Processing units contain only gases, that is, they contain no solid or liquid. The abrasion experienced by the refractories, thus, is minimal. Abrasion, hence, is not a major issue for the refractories in ammonia production units.

• Except for the dome section of the secondary reformer, which supports the catalyst carriers, there is no direct load on the refractories in any of the units

3.8 Refractories for ammonia production unit

As per conventional wisdom, the ammonia production unit operating temperature is not a matter of concern since it is moderate and does not exceed 1350°C. Considering only the operating temperature, hence, it can be stated that even the lowest grade of aluminosilicate refractories, except for the secondary reformer dome, would suffice. Only the strength of the refractories should be high enough to withstand the operating pressure. But is this the case when we consider the gaseous environment in ammonia production units?

The discussion in the previous sections revealed that even the “inert” gases, like nitrogen and steam, influence the refractory properties and hence, performance. It also is evident that it is not only the high temperature that damages the refractories. Rather for certain working environments, for example, destruction of refractories by CO, low, rather than high, temperature causes greater damage to refractories. It is not only operating temperature, but the impact of the gaseous environment also needs to be considered for refractory selection.

The gases from the working environment can permeate through channels as well as open pores (Figure 3) to the interior section of the refractories. In the specific case of reformers, where the operating pressure is of the order of 35 bar, gas permeation is a very likely possibility. Additionally, refractories have finite thermal conductivities, i. e. the temperature along with the thickness of refractory lining decreases.

The silica reduction rate, by hydrogen, increases with the increase of temperature and becomes significant when the temperature exceeds 1100°C (Figure 7). This implies that silica reduction by hydrogen is expected to occur on the refractory hot face, not in the interior part. Refractory destruction by CO, on the other hand, will not occur on the refractory hot face since the reformer operating temperature is >1000°C. But at a certain distance from the hot face, along with the thickness of the refractory, where the temperature is in the 500 to 700°C range, the extent of CO decomposition and as a consequence refractory destruction would be high.

In short, the refractories remain vulnerable to destruction by CO even if the operating temperature of the unit is >1000°C. In the context of ammonia production units, thus, refractories are at risk of destruction by CO and the destruction process may commence at the interior part of the refractory, where the temperature is conducive for CO decomposition.

During the interaction of aluminosilicate refractories with hydrogen, silica is preferentially attacked and this leaves a porous alumina network [15]. Both Hydrogen, as well as CO, permeate through the residual porous alumina structure and thus, silica reduction by hydrogen as well as CO attack continue in the interior part of the refractory. The SEM analysis of the used refractories from an ammonia plant transfer line shows that the loss of silica occurs from the aggregate and not the matrix [19]. This observation contradicts the conventional belief that the matrix, being a finer part of the refractory formulation, has higher reactivity and thus, is more vulnerable to chemical attack.

Since primary as well as secondary reformers and transfer lines operate at high pressure, at ∼35 bar, it is prudent to see the impact of pressure on Hydrogen and CO attack of refractories. An increase in pressure reduces the rate of silica reduction by hydrogen (Figure 10). Higher operating pressure, thus, is favorable for the prevention of refractory degradation by silica reduction. Le chatelier principle, on the contrary, predicts that an increase of pressure would enhance the decomposition rate of CO into carbon (Eq. 8). Such contradicting situations make the refractory selection for ammonia plants more convoluted.

Based on the discussion of the previous section it is obvious that the combined effect of Steam - Nitrogen - CO and H₂ on the aluminosilicate refractory is fairly contradicting as well as complex since:

• Refractory degradation, in the CO environment, is enhanced in the presence of steam or hydrogen and also when they are present simultaneously. Steam on the other hand inhibits the reductant effect of hydrogen.

• Simultaneous presence of CO and Nitrogen, however, reverses the refractory degradation process enhanced by steam. Both steam, as well as nitrogen, are present in the secondary reformer environment

• Silica reduction, by hydrogen, increases with the increase of temperature. On the contrary, CO gas does not decompose into CO₂ and C beyond ∼1000°C since it is thermodynamically stable at this temperature.

• Increase in operating pressure reduces the reductant effect of Hydrogen. Whereas an increase in pressure increases the decomposition rate of CO

• Steam, if present by itself, causes silica loss by the way of Si(OH)₄ formation up to ∼1350°C. Beyond this temperature, however, Si(OH)₄ formation is suppressed and loss of silica is prevented. Up to 1350°C silica loss increases with the increase of steam pressure.

• In the presence of Hydrogen, the adverse effect of steam, on aluminosilicate refractories, is suppressed.

Apart from the opposing conditions of refractory-gaseous environment interactions, marginal volatilization of silica from the refractory may poison the catalysts downstream. Refractories, hence, are selected so that even the slightest silica reduction in the process is mitigated. This is the reason why the lowest possible iron oxide as well as silica-containing refractories, are selected for primary as well as secondary reformers. This is because, in the absence of iron, the CO decomposition rate is reduced, whereas the reductant effect of hydrogen is prevented by the absence of silica in the refractories. Needless to mention that such high purity raw materials are not available in nature and thus, the refractories for primary and secondary reformers should be based on synthetic raw materials like WTA or White Fused Alumina (WFA), which are low in both iron oxide as well as silica.

Refractories, based on synthetic raw materials like WTA or WFA, are typically recommended only for locations where the operating temperature exceeds 1650°C. Primary and secondary reformers are examples, where although the operating temperature is low, still synthetic raw material-based refractories are recommended due to the gas composition in their operating environment as well as the possibility of catalyst poisoning. In other units of ammonia making plant, that is, other than primary and secondary reformers, the conditions are not as severe due to the operating temperature being lower than 500°C, which is the temperature of CO decomposition initiation, and hence, lower alumina aluminosilicate refractories with lesser purity would suffice.

4. Brick or monolithic for ammonia plant

Owing to the progress made on the material development front, there is no significant difference between brick and monolithic chemistry. The gap between brick and monolithic has been bridged primarily by reducing the flux concentration in monolithic formulations. For example, the fluxing component in castable and majority of aluminosilicate gunning formulations is CaO originating from the cement. With the progress in materials technology, aluminosilicate castables and gunning materials can be produced without any cement, that is, without CaO. And thus, in the majority of industrial applications, refractory bricks can be replaced by monolithic.

The aforementioned progress in materials technology can be exploited and monolithic, instead of bricks, based on WTA, can be used in ammonia plants. The primary advantage of monolithic, that is, mechanized installation, thus, can be capitalized for the refractory lining of ammonia plants. The fallout of mechanized installation of monolithic is the requirement of skilled bricklayers is eliminated and simultaneously, the refractory installation rate also is enhanced. Monolithics, depending on its type, the installation rate is 2.5 to 10 times higher than those for the bricks. The substitution of bricks by monolithics, with similar chemistry, would reduce the inventory cost, delivery time, dependence on human skill, and above all the installation time. All these factors put together would reduce the plant downtime when a brick is replaced by monolithic.

Figure 20 illustrates the installation rates of various classes of monolithics and their characteristics. Castables and gunning materials typically have 1.5 and 8% CaO, respectively. Apart from the high concentration of CaO, the strength characteristic of the gunning formulations is not very favorable. But the time required for gunning material installation is significantly lower than that for the castable. New generation gunning formulations, which also are known as wet gunning or shotcreting, have strength characteristics as well as CaO concentration similar to that of castable. On the contrary, the time required for shotcrete installation is comparable to that for gunning material. An additional advantage of shotcrete material is extremely low rebound loss, that is, lower dust generation during installation compared to that of gunning material. This makes shotcrete significantly more user as well as environment-friendly compared to gunning material. In short, shotcrete materials have the advantages of both castable as well as gunning materials, that is, faster installation like gunning material and installed material property as well as chemistry similar to that of castable.

Figure 20.

Features and installation aspects of castable, gunning, and Shotcreting materials.

In short, low iron - low silica-alumina based bricks and monolithics are recommended for primary as well as secondary reformers. Lower alumina products with lesser purity would suffice for the rest of the ammonia plant.

5. Conclusion

Going by conventional perception 45% alumina refractories would suffice for the fertilizer industry since the operating temperature does not exceed 1350°C. This, however, is not the case owing to the simultaneous presence of CO, H₂, CO₂, Steam, and N₂ in the environment of ammonia production units. Generally, steam and N₂ are treated as inert gases in regard to their interaction with refractory materials. But steam reduces the reduction rate of SiO₂ of Aluminosilicate refractories by Hydrogen gas. Steam, on the other hand, enhances the decomposition of CO as per the Boudouard reaction. The favorable effect of steam for silica loss by its reduction, hence, can not be exploited owing to the simultaneous presence of CO and Hydrogen in the ammonia production unit environment. The aluminosilicate refractories recommended for ammonia production units, mainly for primary and secondary reformers, should, thus, be low in iron oxide as well as silica content so that the adverse impact of both Hydrogen and CO is minimized. Such formulation can be either in the form of bricks or monolithic. Refractories with such stringent chemistry requirements can only be met by those based on synthetic raw materials like WTA or WFA.

The fertilizer industry is one good example that epitomizes that the alumina content of aluminosilicate refractories is not decided only by the operating temperature. Owing to the simultaneous presence of Hydrogen, CO, N_2, and steam in the operating environment, there is no option but to use aluminosilicate refractories with a low concentration of iron oxide as well as Silica through the operating temperature of the reformers barely exceeds 1100°C. Based on this analysis it is evident that operating temperature is not the only determinant of refractory quality for an industrial process but the gaseous environment of the unit also plays a significant role in the refractory selection process.

Currently, the reduction of greenhouse gas emissions is a major focus of all industrial processes. To achieve this goal, replacement of fossil fuels by hydrogen for iron production as well as other industrial processes are being targeted. The analyses presented in this paper also will provide direction for refractory selection for the industrial processes where hydrogen is being used or hydrogen is a yield, e.g. Gasification.