Solids flow patterns in large-scale circulating fluidised bed boilers: Experimental evaluation under fluid-dynamically down-scaled conditions

Tove Djerf,David Pallarès,Filip Johnsson

doi:10.1016/j.ces.2020.116309

Tove Djerf, David Pallarès + Show 1 more

Open Access

https://doi.org/10.1016/j.ces.2020.116309

Copy DOI

Abstract

This work aims at gaining novel knowledge of the mechanisms governing the solids flow pattern in the furnace of large-scale Circulating Fluidised Bed (CFB) boilers. A fluid-dynamically down-scaled unit resembling an existing 200-MWth CFB boiler was built and validated against full-scale data. The extensive experimental campaign showed, among others, that the presence or absence of a dense bed governs the entrainment of solids from the bottom region of the furnace, and that the back-flow of solids at the exit region is negligible at low gas velocities although it quickly becomes significant with an increase in gas velocity. Thus, it is shown that the estimation of the external solids flux by the top flux in the furnace is not generally valid.

Highlights

The circulating fluidised bed (CFB) is an established technology for the large-scale combustion of solid fuels in boilers (Koornneef et al, 2007; Cai et al, 2018)
Based on the simplified set of scaling laws described by Glicksman et al (1993) in Eq (4), a fluid-dynamically downscaled unit was built to resemble an existing ~200-MWth CFB boiler – hereinafter termed the ‘reference boiler’
The pressure drop profiles measured in the scale model show good agreement with the corresponding measurements obtained in the 200-MWth CFB boiler

Summary

Introduction

The circulating fluidised bed (CFB) is an established technology for the large-scale combustion of solid fuels in boilers (Koornneef et al, 2007; Cai et al, 2018). Given the complex nature of the in-furnace processes in CFB boilers with their two-phase gas solids flow, much of our current knowledge of CFB has been acquired from experimental data (see, for example (Johnsson and Leckner, 1995; Werdermann, 1993; Couturier et al, 1991; Johnsson et al, 1995; Lafanechere and Jestin, 1995; Leretaille et al, 1999; Johansson, 2005; Mirek, 2016; Yang et al, 2005)) This is because modelling from first principles either requires too great a computational effort (for a Lagrangian description of each solid particle) or entails terms with high levels of uncertainty in the governing equations (Eulerian description of the solids phase).

Objectives

Methods

Results

Conclusion