LOW TEMPERATURE STABILIZATION OF LIQUID OILS

Abstract

The invention relates to a process for hydrotreating a liquid oil stream such as pyrolysis oil stream by, in continuous operation, reacting the liquid oil stream with hydrogen in the presence of a nickel-molybdenum (NiMo) based catalyst at a temperature of 20-240? C., a pressure of 100-200 barg and a liquid hourly space velocity (LHSV) of 0.1-1.1 h.sup.?1, and a hydrogen to liquid oil ratio, defined as the volume ratio of hydrogen to the flow of the liquid oil stream, of 1000-6000 NL/L thereby forming a stabilized liquid oil stream.

Claims

1. A process for hydrotreating a liquid oil stream by, in a continuous operation in a fixed bed reactor, reacting the liquid oil stream with hydrogen in the presence of a nickel-molybdenum (NiMo) based catalyst at a temperature of 20-240? C., a pressure of 100-200 barg, a liquid hourly space velocity (LHSV) of 0.1-1.1 h.sup.?1, and a hydrogen to liquid oil ratio, defined as the volume ratio of hydrogen to the flow of the liquid oil stream, of 1000-6000 NL/L thereby forming a stabilized liquid oil stream.

2. Process according to claim 1, wherein the liquid oil stream contains at least 20 wt % oxygen (O), such as at least 30 wt % 0, or at least 45 wt % 0.

3. Process according to claim 1 wherein the ratio of the carbonyl number as measured by ASTM E 3146 in mol/kg of the liquid oil stream with respect to the stabilized liquid oil stream, i.e. carbonyl number ratio, is 1.7 or higher.

4. Process according to claim 3, wherein the carbonyl number of the stabilized liquid oil stream is below 3.0 mol/kg, as measured by ASTM E 3146.

5. Process according to claim 14, wherein the liquid oil stream is a pyrolysis oil stream or a hydrothermal liquefaction oil (HTL oil) stream.

6. Process according to claim 1, wherein the temperature is in the range 100-225? C.; the pressure is 125-175 barg, and LHSV is 0.2-1.0 h.sup.?1.

7. Process according to claim 1, wherein the NiMo based catalyst is a supported catalyst having a Ni content of 3-5 wt %, and Mo content of 15-25 wt % based on the total weight of the catalyst.

8. Process according to claim 7, wherein the support is selected from alumina, silica, titania and combinations thereof.

9. Process according to claim 1, wherein the Ni Mo based catalyst is in sulfided form, i.e. NiMoS.

10. Process according to claim 1, further comprising a prior step of thermal decomposition of a solid renewable feedstock, for producing said liquid oil stream.

11. Process according to claim 10, wherein the thermal decomposition step is: pyrolysis, thereby producing a pyrolysis oil stream; or hydrothermal liquefaction, thereby producing a HTL oil stream.

12. Process according to claim 11, wherein the pyrolysis is fast pyrolysis, said fast pyrolysis being conducted without the presence of a catalyst and hydrogen.

13. Process according to claim 10, wherein the solid renewable feedstock is: a lignocellulosic biomass including: wood products, forestry waste, and agricultural residue; and/or municipal waste, where the municipal waste is defined as a feedstock containing materials of items discarded by the public, such as mixed municipal waste given the waste code 200301 in the European Waste Catalog.

14. Process according to claim 1, further comprising passing the stabilized liquid oil stream through a hydrodeoxygenation (HDO) step, wherein the HDO is conducted at a higher temperature and equal or lower pressure i.e. higher temperature and equal or lower pressure than the prior step for forming said stabilized liquid oil stream.

15. Process according to claim 14, further comprising passing the stabilized liquid oil stream through one or more metal guards active in hydrometallation (HDM) and/or hydrodeoxygenation (HDO), prior to said HDO step.

16. Process according to claim 1, wherein the hydrogen to liquid oil ratio is 2000-5000 NL/L.

17. Process according to claim 6, wherein the temperature is 150-200? C.

18. Process according to claim 6, wherein the LHSV is 0.2-0.6 h.sup.?1.

19. Process according to claim 7, wherein the NiMo based catalyst has a P content of 1-3 wt %, based on the total weight of the catalyst.

20. Process according to claim 8, wherein the supported catalyst comprises a molecular sieve having topology MFI, BEA or FAU.

Description

EXAMPLES

[0080] Example 1 which is according to the present invention, involves reaction of a pyrolysis model feed consisting of 81.8 mol % 1-propanol, 7.32 mol % furfural, 10.8 mol % 1-octanol, and 0.055 mol % dimethyl disulfide, reacted in the presence of a commercially available catalyst comprising of 3.5 wt % Ni, 19.4 wt % Mo, 2.0 wt % P. The catalyst was sulfided prior to the experiment. The test conditions and conversions are shown in Table 1, the total experimental time with the model feed was 121 hours, during which no sign of plugging was observed. The conversion of furfural was 35% at 100? C., but was 100% at 150, 175, and 200? C. The conversion of 2-propanol was 11% at 100? C. and increased with increasing temperature and was 93% at 200? C. The conversion of 1-octanol was 0% at 100? C., but increased with increasing temperature and was 84% at 200? C. This clearly shows that the NiMo based catalyst is active for stabilization in the temperature range for this pyrolysis model feed at 150 to 200? C.

[0081] Based on the results from gas chromatography-mass spectrometry of the liquid product (GC-MS) it was assumed that the reaction pathway shown below was the main reaction pathway for furfural:

##STR00001##

[0082] Surprisingly, furan-2-yl-methanol was not detected with GC-MS, thus it can be assumed that the hydrodeoxygenation of furan-2-yl-methanol to 2-methyl-furan was fast. In the literature, e.g. ACS Sustain. Chem. Eng. 2016, 4 (10), 5533-5545, it is often found that furan-2-yl-methanol is hydrogenated to tetrahydric-furan-2-yl-methanol before the alcohol is removed. While tetrahydric-furan-2-yl-methanol was not detected with GC-MS, petane-1,4-diol was detected, thus indicating that it is plausible that tetrahydric-furan-2-yl was formed:

##STR00002##

TABLE-US-00001 TABLE 1 Test conditions and conversion Example 1 Temperature (? C.) 100 150 175 200 H.sub.2/Oil (NL/L) 1134 1124 1133 1130 Pressure (barg) 152 152 152 152 LHSV (h.sup.?1) 0.88 0.89 0.88 0.88 Conversion 2-propanol (%) 11 14 34 93 Furfural (%) 35 100 100 100 1-octanol (%) 0 4 31 84

[0083] Example 2, which is comparative, involves reaction of a pyrolysis model feed consisting of 40.6 mol % furfural, 30.9 mol % toluene, 28.4 mol % n-heptane, and 0.2 mol % dimethyl disulfide, reacted in the presence of commercially active and typical hydrotreating catalyst, comprising of 1.7 wt % Ni and 6.1 wt % Mo. The catalyst was sulfided prior to the experiment. The test conditions and furfural conversion are shown in Table 2. The furfural conversion was 100% at run hour 23 but decreased to 42% at run hour 41. At the time the pressure drop over the reactor increased from 5 to 16 bar, thus showing that plugging occurred. Comparing Example 1 with Example 2 clearly shows the benefit of using the NiMo based catalyst according to the present invention which is combined with low temperature and high hydrogen pressure, compared to using a typical NiMo catalyst for hydrotreating at high temperature and moderate hydrogen pressure. In Example 1 according to the present invention, the run hour or total time on stream was 121 h with no sign of plugging.

TABLE-US-00002 TABLE 2 Test conditions and conversion Example 2 Run hour (h) 23 41 Temperature (? C.) 325 325 H.sub.2/oil (NL/L) 644 644 Outlet pressure (barg) 72 71 Pressure drop (bar) 5 16 WHSV (h.sup.?1) 17 17 Furfural conversion (%) 100 42

[0084] Example 3, which is according to the present invention, involves reaction of two pyrolysis model feeds having the molar composition shown in Table 3, reacted in the presence of commercially active material as in Example 1 comprising of 3.5 wt % Ni, 19.4 wt % Mo, 2.0 wt % P. The catalyst was sulfided prior to the experiment. The conversion and the product distribution in the produced organic liquid are shown in

[0085] Table 4. The total experimental time with the model feeds was 218 hours and no sign of plugging was observed during the experiment. The conversion of cyclopentanone was 100% and was not affected by the addition of 500 wt ppm N in the form of pyridine to the mode feed. However, the addition of pyridine changed the yield of products from cyclopentanone. Without pyridine in the feed the main cyclopentanone product was cyclopentane (92.6%), but adding pyridine decreased yield of cyclopentanone and the yield of cyclopentanol increased to 47.5%. The conversion of 1-octanol was 93.6% before pyridine was added to the feed, but adding pyridine to the feed decreased the conversion to 4.6%. The main product from 1-octanol was octane (99.9%), but small amounts (0.1%) of C8 olefines were also observed. Adding pyridine to the feed also changed the 1-octanol product distribution, thus no C8 olefins were observed after pyridine was added, and small amounts of 1.1-oxybis octane (5.1%) and octyl octanoate (2.3%) was observed. It is assumed that 1.1-oxybis octane and octyl octanoate will be subsequently converted to octane under normal HDO conditions (temperature above 300? C.).

[0086] Since alcohols are less prone to form coke than carbonyls, an important reaction in the stabilization of pyrolysis oil is the conversion of carbonyls into alcohols. The observed inhibition of hydrodeoxygenation of alcohols by pyridine is therefore not considered to have a negative impact on the stabilization step and the alcohols can be removed in the following hydrodeoxygenation reactor. Furthermore, the inhibition of the hydrodeoxygenation leads to a lower exotherm in the stabilization reactor, thus making it easier to control the temperature.

TABLE-US-00003 TABLE 3 Composition of feeds used in Example 3 Feed 1 Feed 2 Cyclopentanone (mol %) 7.53 7.52 1-octanol (mol %) 92.37 92.16 DMDS (mol %) 0.10 0.10 Pyridine (mol %) 0.00 0.22 N content (wt ppm) 0 500

TABLE-US-00004 TABLE 4 Conversion and yields (temperature: 190? C., pressure: 151 barg, H.sub.2/oil ratio: 1160 NL/L, LHSV: 0.86 h.sup.?1) Feed 1 Feed 2 Cyclopentanone conversion 100 100 1-octanol conversion 93.6 4.6 Yield of cyclopentanone products Cyclopentane (mol %) 95.2 52.1 Cyclopentanol (mol %) 0.0 47.5 Pentane (mol %) 0.3 0.3 C5 olefine (mol %) 4.5 0.2 Yeild of 1-octanol products Octane (mol %) 99.9 92.7 C8 olefine (mol %) 0.1 0.0 1.1-oxybis octane (mol %) 0.0 5.1 Octyl octanoate (mol %) 0.0 2.3

[0087] This experiment shows that 1-octanol is first dehydrated to octene, which is then hydrogenated to octane. Cyclopentanone is first hydrogenated to cyclopentanol, then dehydrated to cyclopentene, and then hydrogenated to cyclopentane. The dehydration is inhibited by pyridine (heterocyclic organic compound, C.sub.5H.sub.5N), thus showing that pyridine is adsorbed on the acid sites, however the hydrogenation is not inhibited by pyridine, thus indicating that the catalyst e.g. TK-6001 HySwell at the above process conditions, will be able to convert aldehydes and ketones to alcohols using similar condition reactions with a real pyrolysis oil, which normally both contains organic sulfur and nitrogen.

[0088] The reaction scheme below shows the dehydration of 1-octanol to octene and subsequent hydrogenation to octane:

##STR00003##

[0089] The reaction scheme below shows the hydrogenation of cyclopentanone (a cyclic ketone) to cyclopentanol, subsequent dehydration to cyclopentene and then hydrogenation to cyclopentane:

##STR00004##

[0090] The conversion is defined as:

[00001]$x=\left(1-\frac{F_{\mathrm{outlet}}}{F_{\mathrm{inlet}}}\right)?100\%$

[0091] where F.sub.outlet is the molar flow of the reactant out of the reactor and F.sub.inlet in the molar flow of the reactant into the reactor.

[0092] Yield for cyclopentanone's products is defined as:

[00002]$y_i=\frac{n_{C,i}?F_i}{x_{\mathrm{cyclopentanone}}/100?F_{\mathrm{cyclopentanone},\mathrm{inlet}}?n_{C,\mathrm{cyclopentanone}}}$

[0093] Here F.sub.i the molar flow of product i out of the reactor and n.sub.C,i is the number of carbon atoms in product i.

[0094] Similarly, the yield of 1-octanols products is calculated as:

[00003]$y_i=\frac{n_{C,i}?F_i}{x_{1-octa\mathrm{nol}}/100?F_{1-octa\mathrm{nol},\mathrm{inlet}}?n_{C,1-\mathrm{octanol}}}$

[0095] Example 4 shows the effect of carbonyl number on plugging of an actual pyrolysis oil. Table 5 shows the composition of the pyrolysis oil and Table 6 shows the test conditions and product composition. Stabilization reactor, reactor 1 (R1), was loaded with a stabilization catalyst as in Example 1 and thus according to the present invention. A downstream HDO reactor 2 (R2) was loaded with a) a medium-active catalyst having demetallization activity and moderate hydrodesulfurization activity, such as a commercial TK-743 catalyst, and b) a high-active hydrotreating catalyst, such as the high activity NiMo catalyst TK-611 HyBRIM?. No pressure drop was observed for any of the reactors during the first test conditions but decreasing the outlet temperature in R1 to 225? C. (condition 2) lead to plugging of R2 after additional 19 hours. The sample from the interstage (between R1 and R2), thus in the product stream from R1, showed that the carbonyl content had increased from 2.6 mol/kg in condition 1 to 3.9 mol/kg in condition 2, hence indicating that the increase in carbonyl content up to a certain threshold may be at least an indication for the coking of reactor, which lead to the undesired plugging. The increase in carbonyl number ratio or carbonyl content (carbonyl umber) may explain and could be the reason for the coking and thus plugging. The carbonyl number ratio (ratio of carbonyl number in feed to carbonyl number in product of R1) is suitably 1.7 or higher, such as here 1.9, and/or the carbonyl number is suitably decreased to below 3.0 in the stabilized liquid oil from the stabilization reactor in order to avoid plugging of the HDO reactor.

TABLE-US-00005 TABLE 5 Feed composition of pyrolysis oil Analysis Method Oxygen (wt %) Elemental analysis 46 Hydrogen (wt %) D 7171 7.79 Water (wt %) Karl fischer titration 21.3 Carbonyl number (mol/kg) ASTM E 3146 5.0 Acid number (mg KOH/g) ASTM D 664 79.7 Sulfur (wt ppm) ASTM D 5453 63 Sulfur* (wt ppm) ASTM D 5453 119 Nitrogen (wt ppm) ASTM D 4629 416 SG @ 60/60? F. ASTM D 4052 1.1997 *Sulfur content after doping

TABLE-US-00006 TABLE 6 Test conditions and product composition Analysis Method Condition 1 Condition 2 Temperature, i.e. inlet temperature R1 (? C.) 100 100 Outlet temperature R1 (? C.) 250 225 Temperature, i.e. inlet temperature R2 (? C.) 250 250 Outlet temperature R2 (? C.) 330 340 Pressure (barg) 160 160 H2/oil ratio (NL/L) 5075 5075 LHSV R1 (h.sup.?1) 0.11 0.11 LHSV R2 (h.sup.?1) 0.24 0.24 Carbonyl number product from R1 (mol/kg) ASTM E 3146 2.6 3.9 Carbonyl number ratio, feed to product in R1 1.9 1.3 Sulfur (wt ppm) ASTM D 5453 10 6 Nitrogen (wt ppm) ASTM D 4629 2.1 19 SG @ 60/60? F. ASTM D 4052 0.8710 0.8819 SimDist ASTM D 7500 IBP (? C.) 83 100 50 wt % (? C.) 297 310 95 wt % (? C.) 536 550 Time on stream (hours) 0-173 173-192* *Plugged after 192 hours due to plug in reactor 2.