Conference Dates

May 22-27, 2016


The performance of many fluidized bed reactors strongly depends on the bubble behavior since they influence the mass transfer to the dense phase where the catalyst is present. An example is the methanation in a fluidized bed that allows for conversion of unsaturated hydrocarbons in the gasification gas without catalyst deactivation [1]. The BFB reactor is a very challenging step in the process chain to produce SNG out of biomass as feedstock since next to the bubble behavior a lot of other parameters like temperature, pressure, particle size, attrition of the catalyst, internals, bed height and reactor diameter etc. affect the overall performance. The focus of this research work lies on the determination of the bubble properties which are an important factor to model a bubbling fluidized methanation reactor in order to predict and optimize its performance and to support its scale-up [2]. Tomographic methods such as X-ray measurements are often used to characterize bubbles in a fluidized bed. Compared to intrusive measurement, e.g. optical probing, this method possesses the advantage of measuring bubbles throughout the entire cross section. However, X-ray measurements cannot be applied to all installation, especially not in large scale plants. For these purpose, we have developed optical probes that can be employed to investigate the fluidization state in a hot pilot scale reactor. A main drawback of the optical measurements lies in their locally limited detection of the hydrodynamic pattern since they are only able to measure at one point in the reactor. Therefore, conclusions on the bubble behavior of the whole cross section based on optical measurements are not easy to derive. To compare the influence of the measurement method on the measured bubble properties, in the scope of this study, an artificial optical signal is created out of the existing X-ray measurement data set for a cold flow model of the pilot scale methanation reactor. The obtained bubble properties of both methods (i.e. evaluation of the derived artificial optical probe signal and image reconstruction based on the original X-ray tomographic data) are compared with regard to the hold-up, bubble rise velocity and the bubble size (for the X-ray method) or chord length (for the optical evaluation method), respectively. The process to obtain an artificial optical signal is depicted in Figure 1. The comparison shows that for the evaluation of optical probe data, statistical effects have to be considered carefully. The detected mean chord length of the optical method does not represent the mean bubble size determined by the X-ray method. Moreover, also a difference in the bubble rise velocity was detected for some fluidization states. This knowledge may be the basis for the derivation of a statistically sound method to calculate different hydrodynamic properties in fluidized bed reactors based on optical probe measurements.

Please click Additional Files below to see the full abstract.