Cyclone Reactors for Selective Non-Catalytic Reduction of NOx

Casper Stryhn Svith

Research output: Book/ReportPh.D. thesis

50 Downloads (Pure)

Abstract

The objective of this thesis is to investigate the performance of cyclone reactors, focusing on Selective Non-Catalytic Reduction of NOx (SNCR), outlining the existing knowledge and improving the understanding.

Cyclone reactors provide an opportunity for process intensification, by facilitating gas-particle separation, gas-solid heat exchange, and reaction in a single unit simultaneously. One important application of cyclone reactors is the NOx abatement by SNCR in combustion processes. Cyclone SNCR is commonly utilized in the cyclones for CFBC (Circulating Fluidized Bed Combustor) boilers, and in cyclone preheater systems for cement and mineral wool production.

A literature study on cyclone reactors has been conducted to describe the current state-of-the-art of cyclone reactors as well as the current knowledge gap in understanding and modelling cyclone reactors, as well as the SNCR process. Analysis of the available literature on industrial applications of cyclone reactors showed that cyclone reactors are primarily used for high temperature processes, such as combustion, gasification, pyrolysis, pyro-metallurgical smelting processes, and flue gas cleaning by SNCR. Many other potential applications have been investigated in laboratory or pilot scales. The available literature on modeling of cyclone reactors showed a lack of detailed reactor models and a lack of experimental verification of existing models. From the SNCR literature, the published results on ammonia slips are scarce both experimentally and theoretically.

A pilot scale cyclone reactor has been modified and used to conduct a series of SNCR experiments at different reactor conditions. The overall performance of SNCR was tested. The temperature profiles and the concentration profiles inside the cyclone reactor were mapped at various conditions. The temperature in the cyclone inlet varied between 866 and 1023 °C, while the inlet NO concentration was kept at ~500 ppm. Gaseous NH3 was injected in the cyclone inlet using nitrogen as carrier gas, with the NH3/NO molar ratio varied between 0 and 7. The carrier gas flow rate was varied between 0 and 1.5 Nl/min and the injection tube position in the inlet was also varied. The optimal conditions for NOx reduction were found to be at an inlet temperature of 982 °C, NH3/NO molar ratio of 1.6, and carrier gas flow of 1.5 Nl/min. At these conditions, a NO reduction of 69 % was achieved with only 4 ppm ammonia slip. The temperature profiles showed large temperature drops in the cyclone reactor of close to 400 °C from the top to the bottom of the cyclone, suggesting that the main reaction zone of the cyclone was the inlet section and the upper parts of the cyclone chamber. This was confirmed by the concentration profiles measured inside the cyclone reactor, which showed very little or even no NO reduction in the lower part of the cyclone. Despite the low temperatures (600-800 °C) in the lower part, a significant ammonia conversion was still observed in this region. Based on the concentrations measured in the outer and inner vortex and in the outlet, it was concluded that a large fraction of the flue gas bypass’ the majority of the cyclone volume due to lip leakage.

The pilot scale experiments were modelled using both simple and detailed kinetics combined with simple reactor models, and a compartment reactor network model was also developed. The qualitative trends of NO reduction and ammonia slip were captured by the models using detailed kinetics. However, NO reduction and the ammonia slip were significantly over-predicted for most of the experimental results. The modeling results were insensitive to the choice of reactor model and mixing model, while the influences of temperature, gas composition and kinetics model were significant. The simple kinetics of Duo et al. [1] showed the best quantitative fit with the experimental results. The developed compartment model were able to adjust the flow pattern and represent the temperature distribution in the cyclone reactor. Furthermore, the model was able to qualitatively reproduce the internal concentration profiles caused by the double vortex flow in the cyclone reactor, using both detailed and simple kinetics. The simple kinetics generally showed a good quantitative agreement for the NO concentrations, but unsatisfactory for the ammonia.

Industrial scale measurements of the SNCR process were conducted in a preheater cyclone system. In the measurement campaign, raw NOx emissions, flue gas composition, temperatures, and the performance of SNCR were measured. The system were modelled using the measurement data, simple reactor models and detailed chemical kinetics. Experimentally the SNCR process achieved a NOx reduction of up to 70 % without exceeding ammonia slip limits of 40 ppmv for the standard preheater conditions. When operating at low temperatures in the preheater system, the SNCR performance was insufficient for meeting the emission limits, due to high ammonia slips. The mixing limitations in the riser was found to be severe and only 0-15 % NOx conversion occur in the riser, meaning the cyclone chamber is the largest contributor to the conversion. The models could predict the overall trends of the SNCR process using the measured temperature and gas composition data as input. However, the NOx removal and the ammonia slip were over-predicted, as also observed in the pilot scale results.
Original languageEnglish
Place of PublicationKgs. Lyngby
PublisherTechnical University of Denmark
Number of pages172
Publication statusPublished - 2020

Fingerprint Dive into the research topics of 'Cyclone Reactors for Selective Non-Catalytic Reduction of NO<sub>x</sub>'. Together they form a unique fingerprint.

Cite this