PROCESS FOR PREPARING AN OLEFIN OXIDE

Abstract

The present invention relates to a process for preparing an olefin oxide from a reaction mixture stream in an epoxidation reactor R, wherein R contains z active reaction tubes T(i) arranged in parallel, z?2, i=1 . . . z, wherein each T(i) comprises a reaction zone Z(i) comprising a heterogeneous epoxidation catalyst, said reaction mixture stream comprising x components C(j), x?3, j=1 . . . x, the process comprising (i) providing m educt streams E(k), m?1, k=1 . . . m, wherein each E(k) exhibits a mass flow rate F.sub.E(k) and comprises y components C(j), y=1 . . . x, wherein a given component C(j) is contained in at least one E(k); (ii) dividing each E(k) into n educt substreams S(k,i), n?z, each S(k,i) exhibiting a mass flow rate F.sub.s(k,i), wherein to at least one E(k), the inequality (1) applies: Formulas (1), (2), (3), (iii) providing n reaction mixtures streams M(i) comprising the x components C(j), said providing comprising, for each i, either combining and admixing the n educt substreams S(k,i) obtaining the n reaction mixtures M(i) if m>1, or passing on the n educt substreams S(k,i) as the n reaction mixtures M(i) if m=1; (iv) feeding each M(i) obtained according to (iii) into Z(i) and contacting each M(i) in Z(i) with the epoxidation catalyst under epoxidation reaction conditions; wherein the x components C(j) comprise hydrogen peroxide, an organic solvent, and the olefin. The present invention further relates to an olefin oxide obtained or obtainable from said process.

Claims

1.-15. (canceled)

16. A process for preparing an olefin oxide from a reaction mixture stream in an epoxidation reactor R, wherein R contains z active reaction tubes T(i) arranged in parallel, z?100, i=1 . . . z, wherein each T(i) comprises a reaction zone Z(i) comprising a heterogeneous epoxidation catalyst, said reaction mixture stream comprising x components C(j), x?3, j=1 . . . x, the process comprising: (i) providing m educt streams E(k), m?1, k=1 . . . m, wherein each E(k) exhibits a mass flow rate F.sub.E(k) and comprises y components C(j), y=1 . . . x, wherein a given component C(j) is contained in at least one E(k); (ii) dividing each E(k) into n educt substreams S(k,i), n?z, each S(k,i) exhibiting a mass flow rate F.sub.S(k,i), wherein to at least one E(k), the inequality (1) applies: $\begin{array}{cc}{?}^n\left(k\right)=\frac{?\left(F_E\left(k\right)\right)}{F_E^n\left(k\right)}?0.4& \left(1\right)\end{array}$ $\begin{array}{cc}?\left(F_E\left(k\right)\right)=\sqrt{\frac{{.Math.}_{i=1}^n{\left(F_S\left(k,i\right)-F_E^n\left(k\right)\right)}^2}{n}}& \left(2\right)\end{array}$ $\begin{array}{cc}{F}_E^n=\frac{F_E\left(k\right)}{n}& \left(3\right)\end{array}$ (iii) providing n reaction mixtures streams M(i) comprising the x components C(j), said providing comprising, for each i, either combining and admixing the n educt substreams S(k,i) obtaining the n reaction mixtures M(i) if m>1, or passing on the n educt substreams S(k,i) as the n reaction mixtures M(i) if m=1; (iv) feeding each M(i) obtained according to (iii) into Z(i) and contacting each M(i) in Z(i) with the epoxidation catalyst under epoxidation reaction conditions; wherein the x components C(j) comprise hydrogen peroxide, an organic solvent, and the olefin.

17. The process of claim 16, wherein the inequality (1) applies to each E(k).

18. The process of claim 16, wherein m>1 and E(1) comprises two components C(1) and C(2) and is essentially free of C(3), and E(2) comprises one component C(3) and is essentially free of C(1) and C(2).

19. The process of claim 18, wherein C(1) is hydrogen peroxide, C(2) is organic solvent, and C(3) is the olefin.

20. The process of claim 16, wherein x?4 and the x components C(j) further comprise water.

21. The process of claim 20, wherein m>1 and E(1) comprises three components C(1), C(2) and C(4) and is essentially free of C(3), and E(2) comprises one component C(3) and is essentially free of C(1), C(2) and C(4).

22. The process of claim 16, wherein ?.sup.n(k) is in the range of from 0 to 0.4.

23. The process of claim 16, wherein m is 1, 2, or 3.

24. The process of claim 16, wherein m=1, the process comprising: (i) providing an educt stream E exhibiting a mass flow rate F.sub.E and comprising y components C(j), y=1 . . . x; (ii) dividing E into n educt substreams S(i), n?z, each S(i) exhibiting a mass flow rate F.sub.S(i), wherein the inequality (1) applies: $\begin{array}{cc}{?}^n=\frac{?\left(F_E\right)}{F_E^n}?0.4& \left(1\right)\end{array}$ $\begin{array}{cc}?\left(F_E\right)=\sqrt{\frac{{.Math.}_{i=1}^n{\left(F_S\left(i\right)-F_E^n\right)}^2}{n}}& \left(2\right)\end{array}$ $\begin{array}{cc}{F}_E^n=\frac{F_E}{n}& \left(3\right)\end{array}$ (iii) feeding each M(i) into Z(i) and contacting each M(i) in Z(i) with the epoxidation catalyst under epoxidation reaction conditions.

25. The process of claim 24, wherein C(1) is hydrogen peroxide, C(2) is organic solvent, and C(3) is water; and/or wherein x?4 and C(1) is hydrogen peroxide, C(2) is organic solvent, C(3) is water, and C(4) is olefin.

26. The process of claim 24, wherein ?.sup.n is in the range of from 0 to 0.4.

27. The process of claim 16, wherein z is at least 1,000.

28. The process of claim 16, wherein 0.9z<n?z.

29. The process of claim 16, wherein the volume of the filling of a tube of the multitubular reactor with heterogeneous epoxidation catalyst deviates from the filled volume of each other tube by less than 10%.

30. An olefin oxide obtained from the process of claim 16.

Description

EXAMPLES

Simulations

[0109] All simulations were done with process simulation software Aspen Plus v.11. The components used in the process simulation and their characteristics respectively, were taken from the Dortmund Database. The kinetic model of Russo et al. was used without further modification (V. Russo, R. Tesser, E. Santacesaria, M. Di Serio, Ind. Eng. Chem. Res. 2013, 52, 1168-1178).

Reference Example 1: Set Up of Multitubular Reactor (Flooded Reactor, Upstream Mode)

[0110] A multitubular reactor (the reactor) was used with a bundle of 20,000 vertically arranged tubes made of stainless steel with a length of 2,000 mm of each tube and inner diameter of each tube of 28.5 mm. Through the tubes, the reaction mixture was passed from the bottom to the top, i.e. in upstream mode.

[0111] The heat transfer inside the tube was modelled according an axial flow model. The heat transfer outside of the tubes was assumed as non-limiting for the overall heat transport.

[0112] The pressure in the reactor was kept constant at 2.5 MPa.

[0113] The reactor was further equipped with a cooling jacket. As cooling medium, water was passed through the cooling jacket in upstream mode. The flow rate of the cooling medium was adjusted so that the temperature difference between the inlet temperature and the outlet temperature of the cooling medium was 2? C. at most. Typically, this temperature difference was only about 0.5? C.

[0114] All 20,000 tubes T(i) of the reactor each contained (in the reaction zone Z(i)) 620 g of strands of a heterogeneous titanium silicalite-1 (TS-1) catalyst, which was considered an ideal filling. The TS-1 catalyst had a bulk density in the range from 470 to 480 g/l. The titanium content of each TS-1 strand was 0.71 wt.-%, the Si content was 44 wt.-%, each based on the total weight of the TS-1 strand. The pore volume of the strands, determined via Hg porosimetry according to DIN 66133:1993-06, was 73 ml/g. The strands had a diameter of 1.5 mm and the length was in the range from 3 to 5 mm. The tubes had a inner diameter of 40 mm.

[0115] The reaction feed, .e. the educt stream E, consisting of methanol (69.0 wt.-%), propene (11.9 wt.-%), water (11.7 wt.-%) and hydrogen peroxide (7.4 wt.-%) consisted of one single liquid phase and was fed to the multitubular reactor with a mass flow rate F.sub.E=323 t/h at room temperature (25? C.), divided into 20,000 substreams S(i), each substream S(i) exhibiting a mass flow rate F.sub.S(i)=16.15 kg/h, wherein one substream S(i) was fed (as M(i)) to each of the 20,000 tubes. Also, the liquid reaction mixture in the reactor consisted of one single phase. The tubes were assumed to be cooled by an ideal cooling medium of constant temperature and the heat transfer coefficient from the tubes to the cooling medium was assumed to be high enough to not limit the heat transfer. The temperature of the cooling medium was chosen in such a way that the overall conversion of hydrogen peroxide at the exit of the multitubular reactor was exactly 90%

Example 1: Epoxidation of PropeneDistribution of Pressure Loss in the Tubes According Normal Distribution, ?.SUP.n .in the Range from Zero to 0.04

[0116] A reaction of propene with hydrogen peroxide in the presence of methanol and water over a TS-1 catalyst resulting in the desired main product propylene oxide and one or more unwanted by-product(s), which are selected from 1-methoxypropan-2-ol, 2-methoxypropan-1-ol and propylene glycol dimethyl ether, according to Reference Example 1 was modelled.

[0117] As reference, an idealized reactor with ?.sup.n=0 (that is, all tubes have exactly identical pressure loss) was taken, wherein the sum of the weights of the unwanted by-products 1-methoxypropan-2-ol, 2-methoxypropan-1-ol and propylene glycol dimethyl ether was set as 100%.

[0118] The modelling for a value of ?.sup.n=0.026 resulted in an increase of the unwanted by-products of +0.5% compared to the reference.

Comparative Example 1: Epoxidation of PropeneDistribution of Pressure Loss in the Tubes According Normal Distribution, ?.SUP.n.>0.04

[0119] The modelling of Example 1 was repeated, also based on the assumption of a distribution of pressure loss in the tubes according to a normal distribution. Contrary to Example 1, the modelling was based on a value of ?.sup.n=0.066. This resulted in an increase of the unwanted by-products of +5.5% compared to the reference.

Comparative Example 2: Epoxidation of PropeneDistribution of Pressure Loss in Ideally Filled Tubes According Normal Distribution, ?.SUP.n.=0

[0120] The modelling of Example 1 was repeated. Contrary to Example 1, only 19,999 tubes of the reactor were filled according to Reference Example 1; one (1) tube was not filled with TS-1 catalyst, i.e. remained empty.

[0121] For the 19,999 filled tubes of the reactor, a distribution of pressure loss in the tubes was assumed according to a normal distribution. Contrary to Example 1, the modelling was based on a value of ?.sup.n=0 for these 19,999 tubes. The use of one empty tube resulted in an increase of the unwanted by-products of +30.6% compared to the reference.

Comparative Example 3: Epoxidation of PropeneDistribution of Pressure Loss in Ideally Filled Tubes According Normal Distribution, ?.SUP.n.=0.026

[0122] The modelling of Example 1 was repeated. Contrary to Example 1, only 19,800 tubes of the reactor were filled according to Reference Example 1 with TS-1 catalyst; 200 tubes were filled with a catalytically inactive material.

[0123] For all 20,000 filled tubes of the reactor, a distribution of pressure loss in the tubes was assumed according to a normal distribution. Contrary to Example 1, the modelling was based on a value of ?.sup.n=0.026 for all tubes. The use of 1% of the tubes (200 of 20,000 tubes) resulted in an increase of the unwanted by-products of +6.1% compared to the reference.

Comparative Example 4: Epoxidation of PropeneDistribution of Pressure Loss in Ideally Filled Tubes According Triangular Symmetric Distribution, ?.SUP.n.=0.062

[0124] The modelling of Example 1 was repeated. Contrary to Example 1, the modelling was based on a triangular symmetric distribution of pressure loss. Contrary to Example 1, the modelling was based on a value of ?.sup.n=0.062 for all tubes. This resulted in an increase of the unwanted by-products of +10.8% compared to the reference.

[0125] Example 1 and Comparative Examples 1 to 4 were also simulated based on a first alternative multitubular reactor with 20,000 tubes operated in flooded, downstream mode) and also based on a second alternative multitubular reactor with 20,000 tubes operated in trickle bed mode with fed from the top. Comparable results as in Example 1 and Comparative Examples 1 to 4 were achieved for the first and second alternative multitubular reactors, which means that the amount of the unwanted by-products was in each case the same value as in the respective Example or Comparative Example ?0.05%.

[0126] Summary

[0127] The results from Example 1 and from Comparative Examples 1 to 4 are summarized below in table 1, wherein the increase of the overall amount of the unwanted by-products 1-methoxypropan-2-ol, 2-methoxypropan-1-ol and propylene glycol dimethyl ether is reported as %-increase relative to the ideal case, where the overall amount of the unwanted by-products was defined as 100%.

TABLE-US-00001 TABLE 1 Results from epoxidation of propene with hydrogen peroxide according to Example 1 and Comparative Examples 1 to 4. % increase of Example ?.sup.n by-products Remarks Example 1 0.026 +0.5% .sup.a) All 20,000 tubes ideally filled with TS-1 catalyst, normal distribution of pressure loss/flow rates Comparative 0.066 +5.5% .sup.a) All 20,000 tubes ideally filled with TS-1 catalyst, Example 1 normal distribution of pressure loss/flow rates Comparative 0 +30.6% .sup.a) 19,999 tubes ideally filled with TS-1 catalyst + 1 Example 2 empty tube, normal distribution of pressure loss/flow rates Comparative 0.026 +6.1% .sup.a) Normal distribution of pressure loss, but 1% of Example 3 the 20,000 tubes (in this case 200 tubes) are filled with catalytically inactive material Comparative 0.062 +10.8% .sup.b) All 20,000 tubes ideally filled with TS-1 catalyst, Example 4 triangular symmetric distribution of pressure loss/flow rates .sup.a) Compared to the case where ?.sup.n is zero with all tubes ideally filled with TS-1 catalyst and based on a normal distribution of pressure loss/flow rates. .sup.b) Compared to the case where ?.sup.n is zero with all tubes ideally filled with TS-1 catalyst and based on a triangular symmetric distribution of pressure loss/flow rates.

[0128] It was found that, especially when each tube is filled with heterogeneous epoxidation catalyst (TS-1 catalyst), maintaining an below or equal to 0.4 allowed to minimize the amount of unwanted by-products.

CITED LITERATURE

[0129] W. Ruppel in Ullmann's Encyclopedia of Industrial Chemistry, Catalytic Fixed bed Reactors, Wiley-VCH Verlag GmbH & Co. KGaA, Jul. 15, 2012 [0130] Trickle Bed Reactors: Reactor Engineering and Applications, Vivek V. Ranade, Raghunath Chaudhari, Prashant R. Gunjal, Elsevier, Mar. 18, 2011 [0131] IIPressure Drop in Packed Tubes, Industrial and Engineering Chemistry, pages 913 to 919, August 1931 [0132] Ullmann's Encyclopedia of Industrial Chemistry, 5.sup.th edition, volume A 13 (1989) pages 443-466 [0133] EP1122249A1 [0134] WO 2015/049327 A1 [0135] V. Russo, R. Tesser, E. Santacesaria, M. Di Serio, Ind. Eng. Chem. Res. 2013, 52, 1168-1178