A Practical Approach to ppm Sulfur Diesel Production Yuichi Tanaka, Hideshi Iki, Kazuaki Hayasaka, and Shigeto Hatanaka Central Technical Research Laboratory Nippon Oil Corporation 8, Chidoricho, Naka-ku, Yokohama, Japan Abstract Recently, the degree of sulfur reduction for diesel oil has been increasing worldwide. In Japan, the specification for the sulfur content of diesel oil will be reduced from 5 ppm to 5 ppm in 24. At present, the necessity of even deeper desulfurization is being discussed in Europe and the United States as well as in Japan. By 22, the Nippon Oil Corporation (NOC) group had already achieved 5 ppm hydrodesulfurization operation in six refineries by means of catalyst development and the improvement of hydrodesulfurization units. We are now continuing to study the development of catalysts and the improvement of the units in order to achieve ppm sulfur diesel oil production in the near future. 1. Introduction In recent years, strict regulations against environmental pollution have been adopted, and a new movement has arisen to specify the sulfur content of diesel gas oil (table 1). Deeper hydrodesulfurization of diesel oil is said to be necessary because SOx in diesel exhaust gas is very poisonous for de-nox and de-pm catalysts. In Japan, the specification for the diesel oil sulfur content will be reduced from 5 ppm to 5 ppm by the end of 24. Japanese refineries are expecting to have to produce ppm sulfur diesel oil soon, because the Central Environment Council, a Japanese governmental organization, issued a report in July 23 recommending that Japanese oil companies supply ppm sulfur diesel oil in 27; the Council also stated that it hoped that this reduction could be achieved by 25. The deeper the hydrodesulfurization of diesel oil that is required, the larger the volume of catalyst that is necessary. A very large increase in investment for hydrodesulfurization units is inevitable if the catalyst activity is not improved and the reaction conditions are not optimized. This paper introduces our practical approach for the production of ppm diesel oil by means of improved catalysts and the optimization of the reaction conditions without excessive investment. -1-
Table 1 Specifications for the sulfur content of diesel gas oil Japan EU specification (ppm) 5 5 35 5 (1997 ) (24 ) (27, 25 expected) (2 ) (25 ) (29 ) USA 15 (26 ) 2. RESULTS AND DISCUSSION 2.1. Difficulty of ppm Hydrodesulfurization Research was necessary because of the difficulty of ppm sulfur diesel production. The hydrodesulfurization rate constant is described by the following equations: k = A exp(-ea / RT) = LHSV/(n-1) * (Sp n-1 Sf n-1 ) (equation 1) k: rate constant A: frequency factor Ea: activation energy R: gas constant T: reaction temperature n: reaction order Sp: product sulfur Sf: feedstock sulfur The frequency factor in equation 1 can be further factorized into several parameters: k =A(cat.)* (pph 2 )* (H 2 /Oil)* (H 2 S)* (Feed) *exp(-ea / RT) (equation 2) A(cat.): catalyst activity under the standard conditions (pph 2 ): effect of H 2 partial pressure (H 2 /Oil): effect of H 2 /oil ratio (H 2 S): effect of H 2 S concentration in combined feed gas (Feed): desulfurization reactivity of feedstock -2-
For the purpose of evaluating the independent parameters, a number of bench plant tests were carried out, and a hydrodesulfurization kinetic model simulator was constructed. Several results of bench plant hydrodesulfurization activity tests are shown in figures 1 to 6. 1.5 5ppm operation 5-5ppm operation.5 2 3 4 5 6 7 H 2 partial pressure, MPa Figure 1 Effect of H 2 partial pressure 1.4 1.2 5-5ppm operation 5ppm operation.8 15 2 25 3 35 4 H 2 /Oil ratio, nl/l Figure 2 Effect of H 2 /oil ratio 3. 2.5 2. 1.5.5 (Arabian Light Crude % ) 32 33 34 35 36 37 feedstock T9, deg-c difficult sulfur, massppm 32 33 34 35 36 37 feedstock T9, deg-c Figure 3 Relation between T9 and Figure 4 Difficult sulfur in feedstock (4,6-DMDBT and its derivatives) 6 5 4 3 2 3R+ aromatics, vol% 8 6 4 2 28 3 32 34 36 38 feedstock T9, deg-c Figure 5 3R+aromatics in feedstock 1.2.8.6.4.2 2 3 4 5 6 nitrogen in feedstock massppm Figure 6 Relation between nitrogen in feedstock and The under each condition is calculated by comparison of the rate constants from several sets of experimental data. The standard activities for each experiment (5 ppm / 5 5 ppm) are defined from the tests under each set of standard conditions. As shown in figure 1, it was determined that the effect of H 2 partial pressure -3-
on 5-5 ppm hydrodesulfurization operation is larger than 5 ppm. As shown in figure 2, the effect of the H 2 /oil ratio also becomes larger. It is generally said that the reactivity of feedstock is changed by feedstock distillation. The heavier the feedstock T9 is, the larger the observed rate constant becomes. As shown in figure 3, it is found that there is a good relation between the feedstock T9 and the reactivity of the feedstock. This occurs because there is a good relation between the content of difficult sulfur, represented by 4,6-dimethyldibenzothiophene(4,6-DMDBT), and that of larger aromatics (3R+) that poison the hydrodesulfurization catalysis, as shown in figure 4 and figure 5. The nitrogen content, which is poisonous to hydrodesulfurization reactivity as shown in figure 6, is related not only to the feedstock distillation but also to the crude oil type. Studies with a hydrodesulfurization simulator incorporating these factors indicate that a reactor volume about four times larger is necessary for a reduction from 5 ppm to 5 ppm if the catalyst activity is not changed. Moreover, a reactor volume about eight times larger is needed to achieve ppm. 2.2. Strategies In order to produce deeper desulfurized diesel oil, several measures must be combined for minimizing the total investment. NOC has adopted the following strategies for deeper desulfurization of diesel oil: 1. Achieve 5 ppm operation in 22 and ppm in the near future. 2. The priority order of the measures is as follows: (1) Improvement of catalyst activity without increasing the cost per weight. (2) Purification of recycle gas to increase the hydrogen partial pressure. (3) Addition of extra reactor with proper size for decreasing the LHSV. (4) Lowering of the end point of the feedstock to increase the reactivity. On the basis of these strategies, we have discussed the following methodologies for improving the catalysts: 1. Improve CoMo and NiMo catalysts by enhancing the support surface area and the dispersion of metals. 2. Choose a suitable type of catalyst system from among CoMo and NiMo for each of the six units depending on the H 2 partial pressure. 3. Consider the stability of the catalyst life. 4. Consider the reusability of the regenerated catalyst. 5. Maintain the cost of catalyst production at a similar level. -4-
2.3. Features of Improved Catalysts As shown in figure 7, NOC has been developing hydrodesulfurization catalysts for 5 ppm operation. The NHS-92 catalyst was released for 5 ppm hydrodesulfurization, and NHS-99 catalyst was developed for 5 ppm. NHS-1 and NHS-5 were improved on the basis of NHS-99 and were loaded into six hydrodesulfurization units at NOC refineries. Their total loaded volume amounts to more than 1, m 3. NOC recently developed a new CoMo catalyst and a new NiMo catalyst. The relative activities of those catalysts are shown in figure 8. product sulfur massppm NHS-92 NHS-99 NHS-1 NHS-5 (NHS-92=) for 5ppm for 5ppm for ppm base + +2 +3 +4 +5 reaction temperature deg-c Figure 7 Result of HDS bench tests over NHS series CoMo catalysts Feed: d=.8635, S=1.42%, N=227 ppm, T/9=285/361 deg-c, pph 2 =4.9 MPa, LHSV=2. h -1, H 2 /oil=2 Nl/l NHS92 NHS1 NHS5 New New CoMo NiMo Figure 8 Comparison of activity of developed NHS series, and of newly developed CoMo and NiMo. The relations between the relative activities and H 2 partial pressure for each catalyst are shown in figure 9. base New CoMo New NiMo NHS-5(CoMo) 2 3 4 5 6 7 8 H 2 partial pressure, MPa Figure 9 Effect of H 2 partial pressure on new CoMo catalyst and new NiMo catalyst. It is observed that the hydrodesulfurization activity of the NiMo catalyst is superior especially under higher pressure; on the other hand, under lower pressure, the hydrodesulfurization activity of the CoMo catalyst is even better than that of the NiMo catalyst. This can be explained because the hydrodesulfurization of difficult sulfur, -5-
represented by 4,6-DMDBT, is properly performed by the NiMo catalyst, which has high hydrogenation activity, under higher H 2 partial pressure. In other words, sufficient H 2 pressure is necessary to hydrodesulfurize via the hydrogenation route, as well as catalytic hydrogenation activity (Figure ). This phenomenon was not observed during studies on 5 ppm hydrodesulfurization under H 2 pressure between 3 and 7 MPa. In order to reduce the sulfur content of diesel oil to 5 ppm, only the desulfurization of easy sulfur via the direct route was nearly sufficient. CoMo catalyst is more effective at direct route desulfurization than at hydrogenation route desulfurization. 4,6-DMDBT(stericaly hindered Direct Route Hydrogenation Route Figure Desulfurization routes of 4,6-DMDBT There are several hydrodesulfurization units (HDS units; HDS-A to F) operated under various H 2 pressures between 3 and 7 MPa, so an appropriate catalyst must be chosen for each HDS unit. The results of catalyst life tests under the conditions for each HDS unit are shown in figure 11. The new CoMo catalyst was tested under the same condition as HDS-A (low pressure and low LHSV), and the new NiMo was tested under the same conditions as HDS-C (high pressure and high LHSV). The HDS-C unit has a post-treating reactor for decolorizing, so high temperature operation up to 4 deg-c is allowed. The reaction temperatures of both operations were maintained at a proper level for each unit all through a few hundred days on stream, and the catalyst life was determined to be more than two years. reaction temperature, deg.-c +6 +5 +4 New NiMo for HDS-C High LHSV) +3 +2 New CoMo for HDS-A(Low LHSV) + base 2 3 4 5 days on stream Figure 11 Bench plant life test for new CoMo and new NiMo catalysts under each condition. New CoMo: tested for HDS-A (low pressure and low LHSV type HDS unit) -6-
New NiMo: tested for HDS-C (high pressure and high LHSV type HDS unit) 5. CONCLUSION Based on the results of this study, appropriate improvement plans for the six HDS units were decided as shown in table 2. The maximum LHSV was estimated using the hydrodesulfurization simulator, and extra reactors with proper volumes were or will be installed following the existing reactors in each unit. Most of the measures were implemented before the commencement of 5 ppm operation, and the others will be adopted in the near future. Table 2 Improvement plan for 5 and ppm HDS operation Developed Catalyst Add. Reactor Increase pph 2 Decrease feed T9 HDS-A Yes Yes No No HDS-B Yes Yes Yes No HDS-C Yes Yes Yes No HDS-D Yes Yes Yes No HDS-E Yes Yes No No HDS-F Yes Yes No No NOC already achieved 5 ppm hydrodesulfurization operation in our six refineries by 22. As an example, the 5 ppm operation of HDS-F, one of the HDS units in the NOC group loaded with NHS-5, is shown in figures 12 and 13. The 5 ppm operation began in September 22, and stable operation is continuing today. reaction temperature, deg-c + +8 +6 +4 NHS-5 loaded 5ppm operation +2 base 2/7 2/9 2/11 3/1 3/3 3/5 3/7 3/9 Figure 12 5 ppm hydrodesulfurization commercial operation (temperature) S in product oil, massppm 5 4 3 2 '2/7 '2/9 '2/11 '3/1 '3/3 '3/5 '3/7 '3/9 Figure 13 5 ppm hydrodesulfurization commercial operation (sulfur in product oil) We are continuing to study the development of catalysts in order to achieve ppm sulfur diesel oil production in the near future. -7-