Such a process is for example described in NL-A-8201479. Here the production of melamine is represented by the conversion of urea and/or thermal decomposition products thereof, the reaction mixture coming from a melamine reactor being cooled with the aid of water or an aqueous solution, and solid melamine being separated from the melamine solution or suspension thus formed. The reaction is carried out in a reactor containing a fluidised bed of catalyst particles, which bed is kept in a fluidising condition with the aid of gaseous ammonia fed to the reactor, which is distributed throughout the contents of the reactor with the aid of a gas distributor plate or fluidisation plate.

By-products that adversely affect the quality of the product and the fluidisation behaviour in the reactor are formed in the production of melamine. The flow rate of the gaseous ammonia supplied is important for obtaining good fluidisation behaviour.

This gas flow rate is necessary for efficiently distributing the relatively cold urea stream in the reactor, so that there will be no areas where cold urea may accumulate (cold spots) around the internal heat exchange elements. Cold spots may give rise to the formation of by-products which may adhere to the

catalyst. These by-products can optionally be removed.

In the Stamicarbon melamine process, for example, the by-products are partly removed in a filtration unit. In the event of a comparatively high concentration of by-products the burden on the filtration unit will be greater and this may limit the production capacity. The amount of by-products increases substantially at higher melamine vapour pressures in the reactor. This is one reason for choosing a low vapour pressure. In one example (US 4,156,080) the process in the melamine reactor is operated at a melamine vapour pressure that is lower than 0.019 MPa.

The production capacity is however increased by suppressing the formation of by-products.

If only a small amount of by-products is formed, it is also possible to keep the catalyst fluidised without any problems. All this makes it possible to operate the reactor at higher melamine vapour pressures than are mentioned in the state of the art. The advantage of this is that the production capacity increases.

It has surprisingly been found that the production of by-products in the reactor can on an industrial scale be reduced by choosing a certain combination of pressure and melamine vapour pressure in the reactor, together with a certain minimum impulse supply of the ammonia through the fluidisation plate to the reactor. This makes it possible to operate the reactor at a melamine vapour pressure of more than 0.019 MPa.

The applicant has found that in the range of 0.7-2.5 MPa absolute a number of conditions in the reactor can be chosen so as to reduce the amount of by-

products. The maximum melamine vapour pressure will then depend on the reactor pressure and reactor temperature, while certain fluidisation conditions must also be met.

The melamine vapour pressure must be less than: (5.3+9*ln (1.5*Preactor)) *T*lo-5 MPa, and preferably less than: (5.3+7*ln (1.5*Preactor)) *T*105 MPa, where Preactor stands for the pressure in the reactor's synthesis zone in MPa and T for the reactor temperature in °C.

There is no lower limit for the melamine vapour pressure, but for economic reasons it will not be chosen too small.

The melamine vapour pressure should however preferably be 0.02 MPa or more.

The melamine vapour pressure Pmelamine is calculated on the basis of the molar stream Q leaving the reactor using the following formula: <BR> <BR> <BR> <BR> <BR> <BR> Pmelamine=Preactor*Qmelamine/(QNH3+Qc02+Qmelamine+QnNCO) [MPa] where Qi = molar stream of component i (mol/s) Preactor = pressure in the reactor in MPa

QHNCO is calculated using the following formulas: Quera-QH20whereQ'urea= molarQurea= urea stream in the feed QH20 = molar water stream in the feed QHNCO-LC 'Urea-6*Qmelamine in mol/s.

In actual fact Q'urea is the urea feed stream based on water-free urea and QHNCO is the molar stream of the urea not converted into melamine (the greater part of which is present as HNCO) based on water-free urea.

In addition, the flow rate of the gas in part, at least, of the fluidised bed must be greater than the value according to the following formula: Vgas must be greater than 0.008* (rhop*g*dp2/etag) (in m/s) where rhop = density of catalyst particle in kg/m3 dp = average (Sauter) diameter of the catalyst in m etag = dynamic viscosity of the gas in Pa. s 9.8m/s2g= Preferably the velocity of the gas in the fluidised bed must be greater than 0.01* (rhop*g*dp2/etag) m/s and in particular greater than 0.012 * (rhop*g*dp2/etag) m/s.

In addition, the velocity of the gas in part, at least, of the fluidised bed must be less than the value according to the following formula: Vgas must be less than 0.65* (rhop*g*dp2/etag) m/s

where: rhop = density of catalyst particle in kg/m3 dp = average (Sauter) diameter of the catalyst in m etag = dynamic viscosity of the gas in Pa. s g = 9.8 m/s2 Preferably the velocity of the gas in the fluidised bed must be less than 0.60* (rhop*g*dp2/etag) m/s and in particular less than 0.55* (rhop*g*dp2/etag) m/s. Vgas is here defined as the gas flow rate (in m3/s) divided by the minimum free area (in m2) for the free flow of the gas through the fluidised bed. This area will usually be smaller than the total cross-section of the column due to the presence of heating elements in the reactor. The velocity of the gas in the fluidised bed may not be too great either, because then the density of the bed will decrease, causing the interaction between the catalyst particles and the amount of urea supplied to decrease. The reactor's yield may consequently decrease.

The d50 of the catalyst is preferably between 40 and 350 micrometres and in particular between 40 and 200 micrometres. The density of the catalyst particles is determined or calculated for the conditions at room temperature (without the components present in the reactor) and will generally lie between 1100 and 2200 kg/m3.

Preferably the impulse supplied via the fluidisation plate (gas density * (velocity of the gas through the hole in the fluidisation plate) 2 = rhOg*V2hole) must be greater than 10 kg/ (m. s2) and in particular greater than 15 kg/ (m. s2). The velocity of the gas, Vhole is based on the free-flow area at the

point at which the gas enters the reactor. This means that no cold spots can be formed around the fluidisation plate in the reactor. This aspect is all the more important if the temperature of the supplied gas is not at the reactor temperature.

The more closely the value of the maximum vapour pressure is approached, the greater the amount of by-products will be. The extent to which by-products will be formed as the upper limit is approached will depend on the process parameters and the design of the reactor, e. g. type of catalyst, catalyst properties, amount of urea supplied and distribution of the temperature in the reactor.

The catalytic conversion of urea into melamine can be represented using the following equation: 6 CO (NH2) CsNgHg + 6 NH3 + 3 CO2 An example of a known process for the production of melamine from urea via a catalytic process is the Stamicarbon gas-phase melamine process as for example described in Nitrogen No. 139, Sept./Oct. 1982, pp 32-39. In this process liquid urea having a temperature of 130-150°C is supplied to a melamine reactor. This melamine reactor contains a fluidised bed of catalyst particles, which bed is kept in a fluidising condition by supplying gaseous ammonia.

Any of the known catalysts can be used as the catalyst, for example aluminium oxide, silica alumina, silicon oxide, titanium oxide, zirconium oxide, boron phosphate or a mixture of two or more of these catalysts. The expression catalysts'is here understood

to mean any material that promotes the conversion of urea into melamine under the employed reaction conditions.

The temperature at which the melamine is formed from urea is generally more than 325°C. In general, this temperature will not lie above 500°C, more in particular, temperatures of between 370 and 450°C are preferable. This temperature is for example maintained via coils in the reactor through which a melted salt is pumped. This salt is brought to a temperature of 420-480°C in an oven and circulated through the coils in the reactor. The pressure used during the synthesis in the presence of a catalyst is between 0.7 and 2.5 MPa. The amount of gaseous ammonia that is supplied to the reactor as a fluidisation gas via the bottom distributor plate is 0.7-4.5 mol per mol of urea.

An example of an embodiment for the production of melamine from urea is that a gas mixture containing melamine, ammonia and carbon dioxide is transferred from the melamine reactor to the melamine quench columns. Here the gas mixture leaving the reactor is cooled with the aid of an aqueous solution formed in the downstream-processing section of the melamine plant.

This leads to the formation of a suspension of solid melamine in liquid or a solution of melamine in liquid.

The gas stream leaving the quench columns, which consists substantially of ammonia, carbon dioxide and water vapour, is transferred via a heat exchanger to an adsorption column where a stream of virtually pure ammonia and an ammonium carbamate solution is formed.

The ammonia is recycled to the melamine reactor, where it is used as the fluidisation gas, and the ammonium

carbamate solution is for example transferred to an adjacent urea plant. The melamine solution or melamine slurry is transferred to a desorption column. In this desorption column part of the ammonia and carbon dioxide dissolved in the solution or suspension is desorbed with the supply of heat. The gas mixture that is formed in the process is returned to the quench columns. The melamine solution or melamine slurry can be directly processed further or crystallised. In the event of a slurry, said slurry can first be dissolved, after which insoluble contaminants are removed through filtration, optionally with the addition of a filtering aid. The filtered solution is supplied to the crystallisers. The crystallisation takes place at a pressure of between 0.02 and 0.1 MPa and a temperature of between 60 and 100°C.

The present invention is not limited to the above embodiment, but can also be used in other downstream processing variants.

The invention will now be elucidated with reference to the following examples.

Example I The melamine is produced in a cylindrical fluidised bed with an internal diameter of 1 metre and a height of 15 m. The catalyst is fluidised by introducing ammonia via a gas distributor plate and is heated by heat exchange tubes installed in the reactor through which a melted salt flows. Liquid urea is dosed to the reactor under the heat exchange tubes with the aid of a two-phase spray using ammonia as the atomising gas. The reactor contains 2800 kg of catalyst with an average (Sauter) diameter of 76 microns and a particle

density of 1320 kg/m3 and a bulk density of 970 kg/m3.

The reactor is operated at 390°C and 1.2 MPa total pressure. The urea is dosed at a rate of 1.4 tons/hour using 0.7 ton/hour ammonia via the two-phase sprays.

The flow rate of the ammonia supplied via the fluidisation plate is 0.7 ton/hour. The impulse supplied via the fluidisation plate is 720 kg/(m. s2).

The degree of conversion of water-free urea into melamine relative to the equilibrium is more than 98%.

The melamine vapour pressure in the effluent is 0. 035 MPa. The gas stream leaving the reactor is quenched with excess water. The concentration of melem relative to melamine is 0.04 wt. %.

Example II The melamine is produced in the same reactor using the same catalyst as in Example I. The reactor is operated at 390 C and 1.8 MPa total pressure. The urea dose is 1.4 tons/hour, with 1.0 ton/hour ammonia via the two-phase sprays. The flow rate of the ammonia supplied via the fluidisation plate is 1.0 ton/hour. The impulse supplied via the fluidisation plate is 980 kg/ (m. s2). The degree of conversion of water-free urea into melamine relative to the equilibrium is more than 98%. The melamine vapour pressure in the effluent is 0.039 Mpa. The gas stream leaving the reactor is quenched with excess water. The concentration of melem relative to melamine is 0.05 wt%.

Example III The melamine is produced in the same reactor using the same catalyst as in Example I. The reactor is operated at 390 C and 1.2 MPa total pressure. The urea is dosed at a rate of 1.4 tons/hour, with 1.5 tons/hour ammonia via the two-phase sprays.

The flow rate of the ammonia supplied via the fluidisation plate is 0.5 ton/hour. The impulse supplied via the fluidisation plate is 370 kg/(m. s2).

The degree of conversion of water-free urea into melamine relative to the equilibrium is more than 98%.

The melamine vapour pressure in the effluent is 0.026 Mpa. The gas stream leaving the reactor is quenched with excess water. The concentration of melem relative to melamine is 0.03 wt. %.

Comparative Example A The same set-up and the same catalyst as in Example I are used to produce melamine from urea. The reactor is operated at 390°C and 1.2 MPa total pressure.

The urea is dosed at a rate of 1.4 tons/hour, with 0.7 ton/hour ammonia via the two-phase sprays. The flow rate of the ammonia supplied via the fluidisation plate is 0.3 ton/hour. The degree of conversion of water-free urea into melamine relative to the equilibrium is more than 98%. The melamine vapour pressure in the effluent is 0.044 MPa. The gas stream leaving the reactor is quenched with excess water. The concentration of melem relative to melamine is 0.09 wt. %. Under these conditions the fluidised bed can moreover not be operated for a long time on account of the deposition of melem in the reactor and on the catalyst. The melem concentration obtained is consequently not representative of the actual melem production and is probably much higher.