In this thesis work the hydrogenation kinetics of phenanthrene inhibited by the basic nitrogen compound acridine and the non-basic carbazole was investigated. Based on a transient reactor model a steady state plug flow model was developed and kinetic parameters were estimated through nonlinear regression to experimental data. The experimental data was previously collected from hydrotreating of phenanthrene in a bench-scale reactor packed with a commercial NiMo catalyst mixed with SiC. As a first two-step solution, the yields of the hydrogenation products of phenanthrene were predicted as a function of conversion, which subsequently was used to calculate concentration profiles as a function of position in reactor. As a second improved solution, the concentration profiles were calculated directly as a function of residence time, and these results were then used for further analysis. Reaction network 2 in figure 7 was considered sufficient to describe the product distribution of phenanthrene, with a pseudo-first-order rate law for the nitrogen compounds. Both solution methods provided similar results which gave good predictions of the experimental data, with a few exceptions. These cases could be improved by gathering more experimental data or by investigating the effect of some model assumptions. The two-step method thus proved useful in evaluating the phenanthrene reaction network and providing an initial estimate of the parameters, while the onestep method then could give a more precise solution by calculating all parameters simultaneously. As expected, acridine was shown to be more inhibiting than carbazole, both in the produced concentration profiles and estimated parameters. A possible saturation effect was also seen in the inhibition behavior, where adding more nitrogen compounds only had a small additional effect on the phenanthrene conversion. The Mears and Weisz-Prater criteria were found to be inversely proportional to the concentrations of the nitrogen compounds and otherwise only depend on rate constants, with values well below limits for diffusion controlled processes. Sensitivity analyses also supported that the global minimum had been found in the nonlinear regression solution.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:kth-279025 |
| Date | January 2019 |
| Creators | Johansson, Johannes |
| Publisher | KTH, Skolan för kemi, bioteknologi och hälsa (CBH) |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
| Relation | TRITA-CBH-GRU ; 2020:207 |
Page generated in 0.0029 seconds