Metal Catalysed Methanation and Steam Reforming

BY J. R. H. ROSS

1 Introduction

The methanation and steam reforming reactions are closely inter-related and, in general, catalysts used for one reaction will be usable, with some limitations, for the other. This similarity arises from the fact that both reactions occur under reducing conditions over metallic (most commonly, nickel) catalysts and, more importantly, that the types of reactive surface intermediate found during one reaction are also found during the other. Both reactions also suffer from the same constraints, for example, carbon deposition and susceptibility to sulphur poisoning, and hence similar approaches are adopted in both cases in attempts to overcome these constraints. The conditions under which the reactions are carried out depend to a large extent on the composition of the reactant mixture and, in the case of steam reforming, on the desired product distribution. The two reactions have another very different factor in common: there has been a considerable resurgence of commercial interesting variants of both processes. For example, although the methanation reaction has been known since the beginning of the century and it has been practiced commercially to remove traces of carbon monoxide prior to the synthesis reactor in ammonia plants, there has recently been considerable activity on the subject of the methanation of synthesis gas; this activity has arisen because of a resurgence of interest (if only transient) in coal gasification. Prior to the increase in oil prices that sparked these renewed efforts in coal gasification, there was also an increased interest in processes and catalysts for the production of synthetic natural gas (SNG) by the steam reforming of the then cheap naphtha fractions of crude oil. These developments have led to considerable research on the catalysts for these processes and also on the reactions themselves. For example, a total of 26 reviews were published on the subject of methanation in the first six months of 1982 and this puts the subject in the top fifteen 'research fronts' in the physical, chemical, and earth sciences. Under the index terms 'methanation' and 'methanation catalysts' alone, there were approximately 200 articles listed in the 1972–1976 cumulative index of Chemical Abstracts. Under the corresponding headings of the 1977–1981 index, there were about 650 references and there were 155 references in 1982 alone. Clearly, with such an enormous literature and with such an extensive coverage by reviews, it would be unreasonable to attempt to give a comprehensive description of all the work in the subject area embraced by the title of this review. Instead, an attempt will be made to draw a general picture of progress in steam reforming and methanation, with particular emphasis on the catalysts used. Most of the literature covered will be that from the last few years but, of necessity, some earlier work will also receive mention. The structure of the review will be such that a number of the processes themselves will be described in rather general terms in order to establish the requirements for the catalysts; some of the catalysts used for the processes themselves, particularly those based on nickel, will then be described, with particular emphasis on improvements in knowledge of the structure of these materials; finally, a brief description will be given-of some of the more relevant academic publications on the steam reforming and methanation reactions over these catalysts.

2 The Processes

Steam Reforming. – The steam reforming reaction may be described by the general equation:

CnH2n+2 + nH2O [right arrow] nCO + (2n + 1)H2 (1)

The CO formed may take part in two further reactions, the water-gas shift reaction:

CO + H2O [??] CO2 + H2 (2)

and the methanation reaction:

CO + 3H2 [??] CH4 + H2O (3)

Both of these reactions are exothermic and are favoured by reduction in temperature. Hence, while the products of the steam reforming reaction at higher temperatures (~800 °C) are CO and H2, lower temperatures are used to produce methane-rich gases; in this case, the overall reaction can be approximated by:

Cn + H2n+x + (n - 1)/2 H2O [right arrow] (3n + 1)/4 CH4 + (n -1)/4 CO2 (4)

The thermodynamics of these reactions have been discussed in some detail elsewhere.

High-temperature Steam Reforming. The high-temperature steam-reforming reaction is one of the most commonly occurring industrial processes. The major use of steam reforming is in ammonia plants, when the feedstock is most generally natural gas, but other feedstocks such as naphtha or LPG (liquefied petroleum gas) may be used if there is an economic advantage to be gained. The modern generation of ammonia plants have capacities of ~1000 tons per day of ammonia and utilize some 20 m of catalyst in the primary steam-reformer tubes. The service life of a primary steam-reforming catalyst is generally of the order of 2–3 years; however, the catalyst can still have adequate activity at this stage, the replacements being timed to coincide with routine maintenance of the plant. The secondary steam reformer contains a similar quantity of catalyst but here the duty is somewhat less and lives of around 5 years are normal. The secondary reformer in an ammonia plant brings about the complete conversion of the hydrocarbon feedstock by the injection of air to the process-steam prior to the reactor, the amount of air being adjusted to give the required amount of nitrogen for ammonia synthesis. However, although undoubtedly some of the reaction occurring in this bed is steam reforming according to equation (1), the main reaction can be thought of as that of the oxygen of the air with some of the product H2 and CO to form H2O and CO2. There are currently more than 100 plants in the world with capacities of the order of 1000 tons per day and it has been argued that in order to keep up with the fertiliser requirements for the production of food for an expanding world population, a new large-scale plant will need to be constructed each month.

The requirements of the primary reforming catalyst are generally thought to be greater than those of the secondary reformer. The predominant reaction is that given by equation (1), with n = 1 or higher, depending on the availability of fuel. For any value of n, the reaction is highly endothermic and so considerable heat has to be supplied to the reactor; this is generally achieved by burning a proportion of the feedstock, the flame being played directly on the exterior of the reactor tubes. However, in the steam-reforming reactor of the so-called Adam and Eve system (to be discussed further below), the heating is achieved by using a flow of preheated helium (< 950 °C). In order to achieve the desired conversions, the exit temperature of the catalyst bed is generally of the order of 820 °C. The inlet temperature achieved will depend to a large extent on the activity of the catalyst. If the catalyst is relatively active and the majority of the conversion occurs near the beginning of the bed, then the inlet temperature may drop to values of the order of 450 °C, as shown schematically in Figure 1. The function of the remainder of the bed is then largely to shift the product distribution towards that corresponding to the exit temperature, i.e., with reactions (2) and (3) as far as possible to the left-hand side. In conventional hydrogen plants, it is common practice to direct most of the heat at the beginning of the reactor tubes to encourage as large a conversion as possible at that point. The effectiveness of the catalyst in the primary reformer is often expressed by the approach to equilibrium of the exit gas. This quantity is computed by working out the temperature required to give an equilibrium gas mixture corresponding to the exit-gas composition and comparing this with the measured bed temperature at the exit; an approach of 0°C corresponds to complete equilibration of the gas mixture while an approach of greater than 10 °C will indicate that the catalyst is not as effective as it should be. Operating on methane as feedstock, an active catalyst can give an exit gas containing of the order of 0.1% CH at a bed exit temperature of 850 °C, but higher proportions are common. As the catalyst ages, for example by sintering, the temperature profile will gradually move down the bed, as is shown schematically in Figure 1, and the approach to equilibrium will deteriorate.

In ammonia plants, the secondary reformer is included to decrease further the proportion of methane in the final gas and also to introduce the required amount of nitrogen for ammonia synthesis. The bed temperature is maintained at ~ 1000 °C and this is achieved by adding air to the gas stream, the oxygen of the air reacting with the hydrogen of the gas stream to form water. The reactor consists of a packed bed and no additional heating is required. The exit gas contains less than 0.1% CH4. The catalyst for this reactor does not require to have very high activity but it must be stable under these reaction conditions.

A variant of the continuous steam reforming process for hydrogen production is the cyclic reforming process which is used largely for the production of towns' gas by the steam reforming of naphtha feedstocks. In these plants, the catalyst is maintained in a wide, relatively shallow bed which is heated by a flame fueled by the feedstock being used. When the upper part of the bed has reached a temperature of about 725 °C, the reactor is purged with steam and then steam reforming is begun, reaction being continued until the temperature drops considerably. The system is then purged once more and the bed is again heated with a flame. During steam reforming, carbon is deposited on the catalyst and this is burnt off again, exothermically, during the heating phase. In typical plants, cycle times are of the order of 4 min and steam reforming occurs for approximately half of that time. A typical exit gas contains 56% H, 15% CO, 6% CO2, 19% CH4, 4% N2, and a trace of oxygen. The catalyst for these purposes must be mechanically very stable and be able to resist the stresses caused by carbon deposition and by rapid cycles in bed temperature. As a result, the catalysts are often supported on refractory oxides such as α-Al2O3 (see later section dealing with the catalysts for these processes). Recent modifications of the cyclic reforming process include air injection during the steam reforming process; this apparently gives an improvement in the efficiency of the process. The process has also been used to steam reform methane to towns' gas in situations where conversion of gas mains and appliances is not economical.

Low-temperature Steam Reforming. The steam reforming of naphthas at lower temperatures is used to produce methane for use as substitute natural gas (SNG) particularly in situations where there is a shortage of natural gas or for supplementing supplies at peak-load periods. By operating at temperatures of the order of 450 °C, the methanation reaction is favoured and the all-over process can be represented by equation (3). The early developments in this area were carried out by the British Gas Corporation, whose Catalytic Rich Gas (CRG) process is in wide-spread use. The latest variant of the process, entitled the CRG Hydrogasification Process, has been described in some detail by Gray. In this, several CRG reactors are used. After the first, further naphtha is added to the product gas and gasified in another CRG reactor operated at lower temperature; the use of the unreacted steam from the first reactor to convert more naphtha improves the overall efficiency of the process. Another variant of the CRG process recirculates some of the product gas from the first CRG reactor back to its inlet. This is claimed to reduce the speed of catalyst deactivation and to enable heavier feedstocks to be gasified successfully.

A number of other commercial processes have been described which are similar to the CRG processes discussed above. For example, Skov has claimed a process in which half the reactant stream is fed to the first reactor. The product of this reactor is combined with the remainder of the reactant stream and fed to a second reactor in which a methane content of greater than 95% is achieved. Similarly, Nikki has claimed a process in which, after steam reforming at 350 – 550 °C, a CH content of 98% is achieved by methanation at a temperature of 220 °C.

Methanation of Coal-derived Synthesis Gas. – The majority of the energy requirements of the world are supplied by fossil fuels, i.e., natural gas, oil, and coal. Which is the preferred feedstock at any time and in any geographical situation depends on a complex inter-relationship between political, economical, and environmental factors. At the present time, oil is still the preferred feedstock in most developed nations because of its price and because well-developed technology exists to utilize most of the factions of the oil. The lighter factions are used as chemical feedstocks and for petroleum and domestic heating purposes while the heavier fractions are used, e.g., in electricity generation. Whenever natural gas is available, it is used as an alternative to oil, both as a fuel and as a chemical feedstock. Coal is generally, however, used as a fuel and only in places where there is a lack of oil and gas is it used as a chemical feedstock. The gasification of coal and the Fischer –Tropsch process for hydrocarbon production were developed in Germany in the period prior to the second World War. This technology is now practiced in a number of plants operated by SASOL in South Africa. The gasification in these SASOL plants is carried out in Lurgi Gasifiers. In such as gasifier, a fixed bed of graded coal is exposed, under pressure, to a mixture of steam and oxygen, the ash produced being discharged by a rotating grate as an unfused granular solid. The Lurgi process was, at least until recently,' the only commercially proven process in the world suitable for the manufacture of SNG from coal. The greatest problem with such a gasifier is the requirements to supply sufficient excess steam to keep temperatures in the fuel bed below that at which the ash, which forms a substantial proportion of the coal, melts or 'clinkers' and causes problems in the grate of the gasifier. The addition of excess steam reduces the efficiency of the plant and also increases the cost of treatment of the effluent from the plant. However, if only enough steam is added to the gasifier to ensure complete reaction, bed temperatures around 2000 °C are produced and this is sufficient to melt the ash which can then be removed as a slag at the bottom of the gasifier. The so-called 'Slagging Gasifier' was initially developed by British Gas between 1955 and 1964, when the project was closed down. The work was recommenced in 1974 and has resulted in the British Gas 'High Carbon Monoxide' (HCM) process, which incorporates water-gas shift and methanation reactors and produces methane (SNG) with good efficiency. Some of the work was carried out in collaboration with Conoco and details of part of it have been published.

The future of coal gasification for SNG production is uncertain as it depends upon a number of economical and political factors as much as upon technical achievements. It is felt in some quarters that the solution to the current problem of 'Acid Rain' may lie not in the use of SO2 (and NOx) removal processes in the flue gases of oil- and coal-burning power stations but in the prior gasification of the coal or oil, desulphurization of the syngas produced, followed either by the combustion of the syngas or methanation to give SNG. The SNG can then either be distributed to consumers or itself used for electricity generation; the heat liberated in the methantion process would also make an important contribution to the energy balance (see also the following section). At the time of the energy crisis, a number of government and industrially sponsored coal-gasification projects were started but, with the stabilization of oil prices, a proportion of these appear to have been cancelled.

A number of papers have described the design of reactors for methantion of coal- and naphtha-derived synthesis gas; Frohning and Hammer have reviewed some of these and Cornils has described research in West Germany on methanation (and on the Fischer–Tropsch reaction) and the reactors used in this work. A number of different conformations have been described. In one arrangement, a high recycle ratio of product gas is passed back through the methanator to moderate the temperature rise. (A rough rule of thumb is that there will be an 80 °C temperature rise for the conversion of each 1% CO in the feed gas). An alternative arrangement, which seems to be more generally favoured' (see also the next section) is that using three consecutive reactors. The inlet temperature of the first may be ~ 400 °C and there may be a temperature rise of more than 300 °C, even with control of temperature by the injection of some steam; the inlet temperatures and temperature rises of the subsequent beds are somewhat lower. A Union Carbide patent describes a two-bed system in which half the feed is water-gas shifted and then combined with the remainder of the stream and fed to the methanation reactor, this feed-gas now having a CO content of 3–6 vol %. The Forster Wheeler Energy Corporation has described the use of a reactor which includes a twisted nickel ribbon as catalyst and which operates with an outlet temperature of 785 °C and gives a gas containing 54.5% CH4, 41.1% H2O, 0.4% CO, 3.5% CO2, and 0.5% N2, i.e., almost complete conversion. Pennline and his colleagues from the Pittsburg Energy Technology Center have reviewed the use of catalyst-sprayed tube-wall reactors for methanation, while a patent to the French Institute of Petroleum (IFP) describes the use of slurry reactors operating at 200–350 °C. Details of some of the pilot methanation plants in current operation were summarized in the review by Hohlein, Menger, and Range.