280125 Experimental Investigation and Simulation On the Phase of the Product Water in Low Temperature Fischer-Tropsch Synthesis

Monday, October 29, 2012: 9:50 AM
315 (Convention Center )
Xiaojun Lu, Center of Material and Process Synthesis (COMPS), University of the Witwatersrand, Johannesburg, South Africa, Diane Hildebrandt, Centre of Material and Process Synthesis (COMPS), University of the Witwatersrand, Johannesburg, South Africa and David Glasser, Chemical Engineering, University of the Witwatersrand, Johannesburg, South Africa

Introduction

Water is the main by-product in the process of converting the synthesis gas (mixture of CO and H2) to a liquid fuel by applying Fischer-Tropsch Synthesis (FTS). The phase of this by-product water inside the reactor is critical in both the catalysis and reaction engineering aspects as it is an important factor affecting the deactivation of the catalyst, the mass transfer in the catalyst and the heat transfer in the reactor. Fischer-Tropsch Synthesis is presumed to take place within catalyst pores that are believed to be filled with waxy liquid hydrocarbon products [1], while the phase of the water is either regarded as a gas under the reaction conditions or often avoided by the researchers. In this work, an FT experiment was designed and performed and the simulation taking into consideration of the non-ideality behaviour of the mixture of the hydrocarbons and water was conducted. Both the results suggest that a considerable proportion of the water product is in liquid phase in a low temperature FTS reactor.

Experimental

The investigation of the phase of the water in the low temperature (190-230oC) Fischer-Tropsch Synthesis was carried out both from the experimental work and simulation. The experimental investigation was designed using an indirect method by means of monitoring the pressure of the reactor system when the FT experiment was performed in a batch reactor. By comparing the pressure reading of the reactor and the system pressure calculated from reactants' conversion, the phase of the water then could be suggested.

Flushing experiments were further designed and performed. At the end of batch operation, the gaseous material (including products and un-reacted reactants) in the reactor tank (with catalyst inside) was replaced rapidly with inert gas argon. The reactor then was flushed with argon at a low flow rate and the flushed stream was sampled and analysed continuously. As the gaseous material has been replaced, the products in the flushed out stream were from the liquid phase product remained in the reactor.

Results and discussion

The reactor system pressure at different reaction durations in the batch reactor are given in Figure 1. The corresponding CO conversions are given as well. The Figure illustrates three total pressures in the reactor based on different assumptions. The CO conversions for the range of reaction durations are also given as a reference. Curve A is the experimental pressure reading of the reactor. To be conservative we assumed all the hydrocarbons except the methane were in the liquid phase. If we assumed all the water was in the gas-phase Psystem = PCO+PH2+PN2+PCH4+PH2O (Curve B) we obtained a pressure-time curve that was significantly higher than the measured curve. If we assumed all the water was in the liquid-phase, Psystem = PCO+PH2+PN2+PCH4, we obtained Curve C which did give a graph much lower than the measured pressure curve. We therefore conclude at least some of the water must have been in the liquid-phase. We notice there is a crossing in curves A and B at around 3 hours. This could be corrected if we took the phase of C2 and above hydrocarbons into consideration. The information these results provide strongly suggest that a considerable proportion of the water formed by the reaction is in the liquid phase.

 

Figure 1 The pressure in the reactor at different reaction time (CO conversion is plotted as a reference.)

Figure 2 The vapour content in the flushed out streams

when the FT reaction had been conducted at different temperatures

Figure 2 presents the vapour content in the flushed out streams when the FT reaction had been conducted at different temperatures. As it has been explained above, the material flushed out were from the liquid phase product remained in the reactor, so that the concentration of the vapour in the flushed out stream could tell the concentration of the water in the liquid. By integrating the flushed out water with flushing period, the amount of water in the liquid at different reaction conditions could be suggested.

References

1.    Madon, R. J.; Reyes, S. C.; Iglesia, E. The Journal of Physical Chemistry. 1991. 95, 7795-7804.


Extended Abstract: File Not Uploaded
See more of this Session: CO Hydrogenation I
See more of this Group/Topical: Catalysis and Reaction Engineering Division