Annulus flow behavior of Geldart Group B particles in a pilot-scale CFB riser

Highlights

• Analysis of annulus flow behavior of Geldart Group B particles in a CFB riser

• Larger particle diameter dictated upflow annulus behavior in monodisperse systems.

• Binary mixtures largely showed upflow annulus behavior.

• Impact of polydispersity pronounced at lower solids loadings

• Applicability of available riser flow regime maps limited for Geldart Group B

Abstract

Solids flow direction near the wall of a CFB riser was studied for three monodisperse, two binary mixtures and one continuous particle size distribution (PSD) of Geldart Group B particles. Measurements were taken at five axial positions along a 0.3 m diameter, 18 m tall pilot-scale CFB riser at four operating conditions. The data were compared to predictions available from flow regime maps. Results showed that (i) for monodisperse systems, particle diameter was more dominant in dictating upflow annulus behavior; (ii) the binary mixtures exhibited largely upflow annulus behavior, presumably due to the dominant influence of the larger constituent particles in dictating upflow annulus behavior; (iii) for the continuous particle size distribution (PSD), the impact of this type of polydispersity was more pronounced at lower solid loading conditions; and (iv) the applicability of the available regime maps, which were largely developed based on Geldart Group A particles, was found to be limited for the Geldart Group B and non-monodisperse particle systems investigated in this study.

Introduction

Gas-solid circulating fluidized beds (CFBs) have wide-ranging industrial applications [1], [2], including fluid catalytic cracking (FCC), chemical synthesis, coal gasification and more recently chemical looping. Despite the widespread usage, gaps in the knowledge base persist, one of which is the annulus flow behavior in the riser, which is the focus of this study. In particular, core-annulus flow (i.e., a dilute upflow core and a dense downflow annulus) in a CFB riser of Geldart Group A [3] particles is well-acknowledged. The core-annulus phenomenon was first proposed by Bierl and Gajdos in 1982 [4]. They proposed a flow structure, which consisted of a dilute core of upward moving particles, which occupied most of the cross-section, and a relatively dense annulus at the riser wall, which consisted mainly of downward-moving solids. Weinstein et al. [5] also acknowledged the core-annulus structure for describing the radial solids distribution in a CFB. Dry [6] used a technique based on the equivalence of heat and mass transfer to determine the radial solids profiles in a 7.2 m tall fast-fluidized bed and found that the core-annulus structure developed above a height of 0.4 m from the entrance zone. Johnsson et al. [7] concluded that the concentration of solids in the annulus at the wall may be as high as the bulk solids density. Zhang et al. [8] showed that the radial voidage profile became steeper with increasing solids circulation rate or decreasing gas velocity. This corresponded with the results of Monceaux et al. [9] who had shown that the core-annulus flow was enhanced, i.e., the riser had greater solids upflow mass flux at the core and greater downflow solids mass flux at the annulus with increasing solids circulation rate. Herb et al. [10] also observed a core-annulus profile in the riser but they found some downward solids flux in the core, which they attributed to occasional particle clusters, which moved downward due to their greater characteristic mass. They noted that the thickness of the annulus was approximately an order-of-magnitude lower than the column diameter, or column width in the case of a square riser, and that risers with different diameters gave similar radial voidage profiles. Rhodes and Laussmann [11] showed significant variation in the solids flux in the annulus along the perimeter of the riser, while Yang et al. [12] also found greater variation of the local gas velocity in the annulus. Qi and Farag [13] found higher solids flux near the wall, which exceeded the overall solids flux. Zhou et al. [14] observed a core-annulus flow in a CFB riser of square cross-section and noted that the particle concentration was not always minimum at the central axis of the column. The ‘M’ shaped concentration profiles observed agreed with the voidage data for circular risers by Bai et al. [15]. Zhou et al. [14] also found that the core had a more homogeneous flow than the wall. Several authors have noted that the radial solids concentration became more uniform with riser height [16], [17], [18]. In addition, coarser particles have been observed to preferentially segregate to the annulus [19], [20], [21]. Bi and Grace [22] were the first to include the core-annulus phenomenon as a separate regime in a regime map of riser flows involving Geldart Group A and smaller Geldart Group B particles, constructed using semi-empirical correlations published earlier [23], [24], [25].

Tsuo and Gidaspow [26] predicted the core-annulus structure using multiphase flow equations to simulate a two-dimensional CFB riser. They obtained a general quantitative agreement between predicted and experimental values of solid velocity, with some discrepancies at the transition of the core and annulus. Yang et al. [27] studied the core-annulus flow in a CFB riser via a CFD simulation using the energy minimization multi-scale (EMMS) approach and varying drag coefficients of different structures in the core and annulus. Outlet solid flux values, and radial and axial voidage distributions obtained via the simulations were found to quantitatively agree with experimental results. Benyahia et al. [28] employed a two-dimensional model using the kinetic theory of solid particles to simulate a CFB riser and predicted a core-annulus flow from solid flux and granular temperature profiles. While the simulated solid flux data quantitatively agreed more with experimental results at the riser center rather than at the wall, the simulated granular temperature and solid viscosity trends agreed qualitatively with experiments. Gao et al. [29] found a core-annulus flow structure from Eulerian-Eulerian simulations of a CFB riser using a steady-state multi-phase model. The simulated solid concentration and velocity profiles agreed well quantitatively with experimental results, and the inlet boundary conditions had negligible effect on the core-annulus flow. Also the solid motion was found to be dominated by individual particles in the core and by particle clusters in the annulus. Recently, Hou et al. [30] simulated a CFB riser using a drag force based model and obtained the core-annulus structure. While a general quantitative agreement was found between simulated and experimental radial voidage profiles, slight deviations from experimental values were obtained at the riser center.

Riser geometry is widely acknowledged to exert a considerable impact on solids flow in a CFB riser. Regarding type of riser exit, it has been found that constricted [31] or abrupt exits like the blinded T-shaped exit [32], [33], [34], [35] increased solids concentration at the riser top due to solids accumulation [32], [35] which yielded a C-shaped concentration profile with respect to the riser height [15], [35]. Gradual exits like the elbow-shaped exit [36] or a 90-degree exit with a 45-degree baffle [15], [32] were found to affect solids concentration at the top of the riser to a lesser extent. An L-shaped (90-degree) exit had an intermediate effect on solids concentration [15], [32], [33]. Simulations [37], [38] agreed well with the above experimental observations on riser exits. Regarding riser entry, Bai et al. [15] found that a weak inlet restriction to the riser (e.g., fully opened valve) led to an increased solids concentration at the riser bottom while a strong inlet restriction (e.g., valve minimally opened) decreased solids concentration at the riser bottom. De Wilde et al. [39] further noted that a bottom gas inlet hindered gas-solid mixing while a side gas inlet enhanced mixing. Regarding riser diameter (D), it has been observed that with increasing D, the uniformity of axial solids concentration [15], the cross-sectional average solids concentration [40], length of the acceleration region [41] and the extent of influence of an abrupt exit [42], [43] increased, while axial solids mixing decreased [44]. Regarding riser height (H), it has been found that shorter risers exhibited only the acceleration region [45] such that the region of fully developed flow did not manifest [15]. Further, the section of riser influenced by the gradual exit decreased with decreasing H [15].

Most of the literature describes the core-annulus structure as a radially segregated solids flow with a lean core consisting of upflowing particles and a thin dense annulus consisting of downflowing particles. However, in the case of dense phase flow, a core–annular solids distribution persists, but, instead of a downflow annulus, an upflow annulus was consistently observed [46], [47], [48], [49], [50], [51]. Bai et al. [52] found that the statistical and chaotic analyses from the signals obtained from such high-density CFB (HDCFB) risers were markedly different than the ones from more dilute CFB risers. Gupta and Berruti [53] demarcated ‘dense phase flow’ and ‘fast fluidization’ by an empirical correlation developed from experimental data on Geldart Group A particles. Grace et al. [54] subsequently proposed a new flow regime, using experimental data obtained mostly from Geldart Group A particles. They called it ‘Dense Suspension Upflow’ (DSU). The key feature of the DSU regime is an upflow annulus, while retaining the variation of radial solids concentration as in a conventional core-annulus flow (i.e., dilute core and dense annulus). Issangya et al. [55] further carried out measurements of local voidage in an HDCFB riser using Geldart Group A particles and found that, while voidage fluctuations were low at the core, they reached a peak value (i.e., maximum heterogeneity) at some distance away from the wall, which is in contrast with the conventional core-annulus flow whereby maximum heterogeneity was found at the wall. Subsequently, Kim et al. [56] demarcated three different flow regimes for CFB risers, based on the flow direction of the annulus and the solids holdup across the riser cross-section at a given height, ϕ. They used previously published experimental data, most of which was from Geldart Group A particles. Many authors have investigated the ‘Dense Suspension Upflow’ (DSU) regime [57], [58], [59], [60], [61], but with the focus largely on Geldart group A. Table 1 lists the relevant characteristics of the experimental systems used in the studies referenced above. Although there has been some research on demarcating flow regimes in CFB risers containing Geldart group B particles [62], none has focused on investigating the core-annulus behavior of these systems. Geldart Group B particles are well-acknowledged to fluidize differently from Geldart Group A particles, with the most significant difference being the lack of cohesion in the former [3]. In particular, a previous study has shown that Geldart Group B particles surprisingly exhibited a reverse core-annulus behavior (i.e., a dense core and a dilute annulus) due to the higher Stokes numbers (St) of the particles [63]. Because behavioral trends of Geldart Group A particles cannot be directly extrapolated to predict those of Group B, an investigation into the hydrodynamics of the latter warranted especially with the increasing prominence in its use as inert materials to improve the fluidization quality of biomass [64], [65], [66]. Accordingly, this study is focused on understanding the annulus flow behavior of Geldart Group B particles.

The lack of knowledge on the impact of polydispersity, which is an ever-present aspect of solids processing in industry, has been flagged as being a critical gap to bridge [67], [68], [69], [70]. Polydisperse systems have been shown to exhibit distinctly different behaviors from monodisperse systems [21], [70], [71], [72], [73], [74]. Distinctly different behaviors among monodisperse systems and different types of polydispersity (e.g., binary mixtures and continuous particle size distributions (PSDs)) have been reported in bubbling fluidized beds [74], [75], [76], [77], [78] and circulating fluidized beds (CFBs) [79], [80], [81], [82]. The impact of polydispersity on the annulus flow behavior has, however, not been studied.

The present work aims to experimentally study the annulus flow behavior of Geldart group B particles in a pilot-scale CFB riser. Six particle systems, namely, three monodisperse, two binary, and one continuous PSD, were studied at four operating conditions. Measurements were done at five axial positions along the riser. The annulus flow behavior was compared to the available flow regime maps [22], [53], [54], [56].