(716b) Validated Hydrodynamic CFD Model for Catalytic Fast Pyrolysis

Catalytic fast pyrolysis (CFP) is a promising technology for converting lignocellulosic biomass into renewable transportation fuels and chemicals. In CFP, small particles of biomass are injected into a circulating fluidized bed (CFB) of hot catalyst particles. Inside the reactor, the biomass particles are rapidly heated and pyrolyzed, and the vapors are catalytically deoxygenated. As in Fluidized Catalytic Cracking (FCC), coke combustion in the regenerator supplies the heat required by the endothermic process. The advantages of CFP over competing thermochemical biofuels processes are twofold. First, a single particle is both heat carrier and catalyst, which greatly simplifies the solids handling and reactor engineering, in turn reducing capital requirements and operational complexity. Second, the resulting bio-oil is much lower in oxygen than typical pyrolysis oil, and therefore, much more stable. When combined with a hydrogen addition upgrading step, the overall yield and economics of the CFP process are superior to its competitors.

Inaeris Technologies (formerly KiOR) is a world leader in the development and commercialization of the CFP technology. Reactor design is a crucial element of the Inaeris technology. Depending on the design concept employed, it is possible to have multiple hydrodynamic regimes in the same reactor, ranging from moving bubbling bed, to dense phase riser, to dilute phase riser (fast fluidization). Furthermore, simultaneous feed of solid biomass and catalyst particles, and evolution of pyrolysis gases, complicates the hydrodynamics analysis. Both of these factors make it difficult to use standard CFB scale-up correlations and methods. As a result, Inaeris decided in late 2014 to start a program to develop computational fluid dynamics (CFD) models capable of supporting scale-up from pilot scale. Other future uses such as optimization, troubleshooting, and ultimately, blank sheet reactor design were also contemplated as a justification for this effort.

In early 2015, Inaeris selected the Barracuda Virtual Reactor® software package, and embarked on a collaboration with CPFD Software, for two reasons. The first was CPFD’s experience in rapid model development; the software’s incorporation of the MP-PIC numerical method was the second.

After two years, 1000+ cold flow CFB tests, and 500+ simulations, the collaboration has produced a single CFD model which has been experimentally validated for bubbling, dilute phase and dense phase regimes across a wide range of reactor configurations [1-4]. Equally important is that the model accurately predicts hydrodynamics for catalyst-only circulation, as well as those for catalyst and biomass co-feeding. This is important, because somewhere inside the reactor, the hydrodynamics transition from catalyst plus biomass to catalyst only (with some residual char particles). Thus the model must be able to accurately predict the extreme cases.

In this presentation, details of the model will be presented, and application of the model to reactor scale-up will be illustrated. Further recommendations for model refinement, as well as guidelines for general CFD model development for CFP, will also be given.

References

[1] B. Adkins, N. Kapur, J. Pendergrass, J. Parker, P. Blaser, KiOR Update: Incorporating Barracuda in our CFP Development Process, Proceedings of the Inaugural Barracuda Virtual Reactor Users’ Conference, Santa Ana Pueblo, New Mexico, 2015.

[2] B. Adkins, N. Kapur, T. Dudley, S. Webb, P. Blaser, Experimental validation of CFD hydrodynamic models for catalytic fast pyrolysis (CFP), Fluidization XV, May 22-27, 2016, Quebec, Canada.

[3] N. Kapur, B. Adkins, T. Dudley, S. Webb, P. Blaser, “Cold Flow Validation of Catalyst-Biomass Hydrodynamics in Catalytic Fast Pyrolysis”, 2016 AIChE Annual Meeting, San Francisco, California, November 2016.

[4] B. Adkins, N. Kapur, T. Dudley, S. Webb, P. Blaser, “Experimental Validation of CFD Hydrodynamic Models for Catalytic Fast Pyrolysis”, Powder Technology, published online, December 2016.