The following components were analysed using specific detectors:
Solvent +RS TBPS
• Hydrogen (TCD detector) • Nitrogen (TCD detector) • Helium (TCD detector) • Hydrogen sulphide (TCD detector) • Methane, ethane, and propane (FID detector) • n-butane and isobutane (FID detector) • 1-butene and isobutene (FID detector)
Liquid trap
H He i-C i-C = 1-C = C -C H S C -SH
Solvent R-S H He i-C i-C = 1-C = C -C H S C -SH TBPS
Traces of solvent
Reactor
Solvent TBPS R-S
Helium was used as an internal standard in the GC anal- ysis and was supplied at a constant flow rate along with hydrogen as the gas feed to the reactors. The concentration of helium, being an inert gas, was utilised to calculate the total mass flow rate of the effluent gas. The gas chromatog - raphy analysis had a total duration of 15 minutes, allowing sequential sampling of the gas effluent from the 16 parallel reactors over a span of four hours. The liquid product from each reactor was accumulated during this four-hour period. Sampling of the gas and liquid started after four hours of stabilisation under each testing condition. The first sampling occurred between four to eight hours, followed by a second sampling between eight to 12 hours while operating under fixed conditions. To estimate the amount of unconverted TBPS in the product effluent, the total sulphur concentration of each liquid sample was meas- ured. It was assumed that no other sulphur-containing spe- cies remained in the liquid product for calculation purposes. This was possible since the total amount of sulphur in the feed and products (unconverted TBPS and H₂S) was known. The amount of organic sulphur present in the non-spiked solvent (Table 1) was assumed to be inert in the reactors at the tested operating conditions. Direct analysis of butyl mercaptans in the GC was not feasible due to limitations in the setup, such as the total duration of the GC method and the availability of a calibration gas for this component. Instead, the total mass flow rate of butyl mercaptan in the gas effluent was calculated by performing an elemental sulphur balance. The estimation of hydrogen consumption at each testing condition was done by weight difference. This involved using the gas flow rate fed into each reactor (measured by the mass flow controller) and the concentration of hydro - gen in the gas effluent measured through the GC analysis. Further information on the calculations is provided in the reference section. A graphic representation of the distribu- tion of the different products formed in the TBPS decom- position is presented in Figure 2 . This scheme already considers the assumptions made for the estimation of the different by-product yields: • No vaporisation of the solvent, and TBPS only present in the liquid product. • No conversion of organic sulphur across the reactors. • No dissolution of H₂S and butyl mercaptans in the liquid product. Although an ideal gas-liquid separation of the different by-products is not possible in the testing setup, the effect of the small amount of gases dissolved in the liquid product and the heavy components ‘carried’ away in the gas stream has a minimal impact on the estimation of the by-product yields.
GC
Vial
Figure 2 TBPS by-product distribution in the testing unit
Results Thermal decomposition of TBPS without catalyst present The thermal decomposition of TBPS with no catalyst pres- ent was examined in duplicate reactors using two LHSV. The LHSV was adjusted by loading a smaller amount of catalyst in reactors 9 to 16 (blocks 3 and 4). Key components, includ - ing H 2 S, isobutane, isobutene, C 4 mercaptans, and TBPS, were analysed in the online GC for their product yields. Each marker in Figure 3 represents a specific component, while different colours indicate varying space velocities. The exper - imental conditions typical for naphtha hydrotreating were sustained for a duration of 12 hours. Gas effluent samples were sequentially analysed through gas chromatography, and liquid effluent samples were collected and assessed for total sulphur content. Throughout Run01, a fixed pressure of 15 barg and a gas-to-oil ratio of 50 Nl/l were maintained. Sequential temperature changes occurred every 12 hours. After analysing the reactors without the presence of a catalyst, it was observed that temperatures below 200°C showed negligible amounts of H₂S and isobutene. This suggests that most of the TBPS remains unconverted or decomposes into non-volatile sulphur components that remain in the liquid phase and cannot be distinguished from TBPS itself. Additionally, prior research by Heller and Roberts 3 reported a decomposition temperature of approximately 160°C for TBPS, suggesting that the extent of thermal decomposition into highly volatile components would likely be negligible. In the same temperature range (below 200°C), the trends in yields indicate the formation of some volatile sulphur com- ponents. However, these fluctuations may be attributed to intrinsic variations in sample analysis and deviations from
Solvent feeds properties
Properties
Naptha 0.7549
Diesel 0.8342
Density @ 15°C, g/ml
Sulphur before spiking, ppmwt Sulphur after spiking, ppmwt
100
2,915
20,675 37,034
23,407
SufrZol 54 added, wt%
3,834
IBP, °C FBP, °C
65
133 437
196
Table 1
67
PTQ Q4 2024
www.digitalrefining.com
Powered by FlippingBook