Vapor pressure estimation for polar compounds is very important in environmental chemistry. Most vapor pressure estimation equations are based on the Clausius-Clapeyron equation. Such equations are extremely accurate above the boiling point, but cannot be used for polar compounds around 1 Pa-100 Pa. A new polynomial expansion type of vapor pressure was developed. Such a non-linear expression has many solutions and the global minimal answer is difficult to obtain. The Genetic Algorithm (GA) was applied to determine the coefficients of the polynomial expansion equation. The accuracy of this polynomial expansion equation for vapor pressure is much better than the Riedel equation under the boiling point for polar compounds. This method was also applied for acentric factor estimation and an estimation equation with better accuracy than the Edmister equation is proposed.

View full abstract

Net hydrogen off-gas from the continuous catalyst regeneration type catalytic reforming process (CCR process) contains inorganic chlorides. In order to prevent potential problems such as corrosion in the downstream processes, such chlorides are commonly removed by using fixed-bed chloride traps. We have extensively investigated the chloride removal properties of various materials for the chloride traps and have developed effective and practical zinc oxide based chloride traps. Firstly, we found that net hydrogen off-gas from the CCR process contains not only inorganic chlorides but also organic chlorides. Secondly, we found that the widely used activated alumina based chloride traps have the major disadvantages of formation of organic chlorides from inorganic chlorides on the surface and leakage of the organic chlorides to the downstream processes. These organic chlorides may be decomposed by heating and may cause corrosion in the downstream processes. Conversely, we found that zinc oxide based chloride traps have high potential to remove both inorganic chlorides and organic chlorides through the presumptive mechanism that organic chlorides are converted into inorganic chlorides on the surface and are then trapped by reaction with zinc oxide. However, zinc oxide based chloride traps have problems with pellet breakage and pressure drop buildup due to the deliquescence of zinc chloride derived from the reactions of chloride compounds and zinc oxide. To solve these problems, the chemical and physical properties have been improved by appropriate reduction of zinc oxide content and increase of pore volume with addition of porous inorganic materials to increase the contribution of zinc oxide inside pellets to chlorine removal and to enhance zinc chloride retention capacity. Consequently, we developed a new zinc oxide chloride based chloride trap, JCL-1. Demonstration tests of the JCL-1 showed stable operation with effective removal of both inorganic chlorides and organic chlorides without pressure drop buildup or pellet breakage.

View full abstract

CP-MAS ¹³C NMR measurements were carried out on mixed gas hydrates containing CH₄-C₂H₆. The changes in NMR chemical shift values for CH₄ and C₂H₆ clearly corresponded to the structural changes in the hydrate structure. The encaged gas compositions estimated by the integrated ¹³C NMR signal intensities agreed well with the dissociated gas compositions measured by gas chromatography. Therefore, the gas composition in mixed gas hydrates can be directly estimated from the ¹³C NMR spectra. The cage occupancies of the small and large cages of the hydrates were estimated from the ¹³C NMR spectra on the basis of a statistical thermodynamic model. The large cages were almost fully occupied with guest molecules, whereas small cage occupancy decreased with increasing C₂H₆ concentration. Therefore, large cages are highly preferentially occupied by C₂H₆ molecules rather than CH₄ molecules.

View full abstract

The effect of sulfidation on ¹²⁹Xe NMR spectra of Co-Mo/Al₂O₃ hydrodesulfurization catalysts was investigated. Before sulfidation, catalysts containing cobalt such as Co/Al₂O₃ and Co-Mo/Al₂O₃ showed remarkable ¹²⁹Xe NMR peak broadening, whereas one sharp ¹²⁹Xe NMR peak was observed for Mo/Al₂O₃ and Al₂O₃, mainly caused by the paramagnetic effect of cobalt oxides such as CoAl₂O₄ and CoO. In contrast, after sulfidation, the remarkable broadening of the ¹²⁹Xe NMR peak did not occur. We consider that the paramagnetic effect of cobalt was much reduced due to the transformation of paramagnetic oxides such as CoAl₂O₄ and CoO into antiferromagnetic sulfides such as the Co-Mo-S phase and Co₉S₈. In addition, XPS of the sulfided catalysts showed that the Co 2p binding energy of the sharp peak for Co-Mo/Al₂O₃ was 0.7 eV higher than that for Co/Al₂O₃. This result strongly suggests that cobalt was mainly present as the Co-Mo-S phase on the surface and is closely related to the observation that the ¹²⁹Xe NMR peak of sulfided Co-Mo/Al₂O₃ was shifted further downfield in comparison with sulfided Mo/Al₂O₃. ¹²⁹Xe NMR spectroscopy is sensitive to the formation of the Co-Mo-S phase on Co-Mo/Al₂O₃ hydrodesulfurization catalyst.

View full abstract

Resid hydrodesulfurization catalyst was regenerated after commercial operation and the effect of vanadium accumulation during resid hydrotreating reaction on the following catalyst regeneration was investigated. Vanadium accumulation accelerated the aggregation of active metals such as NiMoO₄. Moreover, vanadium accumulation lead to the formation of Al₂(SO₄)₃ on the catalyst during regeneration. Although sulfur-containing species such as MoS₂ and vanadium sulfide were present all over the catalyst particle before regeneration, Al₂(SO₄)₃ only formed around the exterior of the catalyst particle after regeneration. Since the distribution of Al₂(SO₄)₃ corresponded well to the vanadium distribution, vanadium species probably catalytically oxidized SO₂ into H₂SO₄, and then transformed the alumina carrier around the vanadium to sulfate. The effect of vanadium accumulation on the resid hydrotreatment activities of the resulting regenerated catalyst was also investigated. The hydrodesulfurization, hydrodenitrogenation and hydrodemicrocarbon residue activities over the regenerated catalyst decreased with increasing amount of vanadium accumulation. Since the XRD results indicated that the formation of NiMoO₄ increased with vanadium accumulation, the deactivation of HDS activity probably resulted from the decrease in the number of active sites due to aggregation of nickel and molybdenum. On the other hand, hydrodemetallization and hydrodeasphaltene activities increased with increasing amount of vanadium accumulation. Vanadium sulfide accumulated on the catalyst may have formed new hydrodemetallation and hydrodeasphaltenization active sites.

View full abstract

The differences between MoS₂, CoMoS and NiMoS HDS catalysts supported on γ-alumina and high SSA titania are investigated based on the results of [³⁵S]DBT HDS experiments. Previous studies of MoS₂ and CoMoS are reviewed, discussed and compared with new results for NiMoS. Introduction of Ni or Co to MoS₂/Al₂O₃ catalysts classically yields a significant increase in HDS performance. Irrespective of the promoter, an increase in S₀, the number of labile sulfur atoms, is observed. In contrast, k_RE, the H₂S liberation rate constant, plotted as a function of the Ni/Mo ratio, presents a volcano profile on Ni-promoted catalysts, but k_RE reaches a plateau from low Co/Mo ratios on Co-promoted catalysts. The 'TiMoS' phase, which is formed in-situ during HDS on Mo/TiO₂ catalysts, promotes sulfur mobility and makes Mo/TiO₂ catalysts more active than Mo/Al₂O₃ catalysts. Nevertheless, CoMo/TiO₂ catalysts are less active than CoMo/Al₂O₃ catalysts because further promotion of 'TiMoS' phase with Co might yield excessive weakening of the metal-sulfur bonds, and/or some Co atoms might be 'lost' in the TiO₂ matrix without interacting with MoS₂. In contrast, introduction of Ni to Mo/TiO₂ catalysts yields significant increases in both k_RE and S₀. The NiMo/TiO₂ catalysts exhibit HDS performances close to those of Al₂O₃-supported catalysts. Clearly catalytic behavior over Co- and Ni-promoted catalysts is different.

View full abstract

The model for the bench plant of internally heat-integrated distillation column (HIDiC) with structured packings was developed by a rate-based model and the separation of benzene-toluene binary mixture was simulated. The developed model has been applied as bench-plant for HIDiC, and their separation and operation performance was discussed. The validity of the simulation results was confirmed with the experimental data of the HIDiC bench plant. In addition, the simulation results were also compared with those obtained with the conventional equilibrium model. It was found that the present model can predict the behavior of the bench plant more precisely than the conventional equilibrium model. Moreover, the energy-saving ratio in the HIDiC system has also investigated using this simulator with rate-based model.

View full abstract

Oil recovery enhancement of gasflood by oil film flow was verified. Film flow phenomena involve the formation of a thin film by residual oil blobs in water-wet media which becomes continuous between water and gas under certain conditions, and consequently recovers mobility. A core flood experiment assuming film flow was carried out under the same conditions as actual reservoirs. Since the oil film is supposed to be formed after the interfacial tensions are balanced, a soak period was set during gasflood to balance the interfacial tensions. The oil production behavior was different between before and after the soak period and the oil production rate rose after the soak period, which indicates that the film flow phenomena occurred. History matching was performed using typical three-phase relative permeability models which reproduce the experimentally observed production behavior before the soak period, but could not reproduce the behavior after the soak period because these models did not consider film flow. Therefore, a modified relative permeability model considering film flow was developed to simulate the production behavior. By applying this model during the soak period when film flow was formed, the simulation could reproduce incremental oil recovery after the soak period.

View full abstract