Research Article Modeling and Characterization of Air Emissions from Laboratory and Industrial Fluidized Beds A fluidized bed model using several elutriation correlations was developed and tested against an operating fluidized bed used in a Fluidized Catalytic Cracker Unit (FCCU) and a 1:8.5 scale laboratory system. It was found that there was little variation between the emission rates predicted using different elutriation correla- tions, although the newly developed equations were slightly more accurate for the laboratory-scale system. Although total emission rates were predicted with rea- sonable accuracy, the actual volatility and fluctuations seen in real fluidized beds emissions were not predicted. When the model was used to predict particle emis- sion from the industrial FCCU, they preformed poorly, grossly overestimating the actual levels. It was determined that the attrition terms used in emission model- ing were inappropriate and that the model preformed better without them, but still overestimated the actual emissions. This overestimation was greater in the in- dustrial system compared with the smaller laboratory system. It was also found that the older elutriation terms were better for predicting industrial emissions compared with those of the smaller scale units. Keywords: Air pollution, Fluidized beds, Fluidized Catalytic Cracker Unit (FCCU), Modeling, Particle emission Received: January 08, 2008; accepted: June 26, 2008 DOI: 10.1002/ceat.200800007 1 Introduction The process of fluidization is widely used in many chemical and process industries. The term fluidization refers to the pass- ing of gas (or fluid) through solid particles, causing the parti- cles to expand, and therefore, allowing excellent heat and chemical transfer to occur [1]. An important industrial process utilizing fluidizations is the petroleum industry where Fluid- ized Catalytic Cracking Units (FCCU) are used to convert long chained hydrocarbons into more valuable shorter chained ones. FCCU’s typically consist of three parts, i.e., a rising main where the chemical reactions occur between the catalyst and hydrocarbons, a reactor to separate the product and catalyst, and a regenerator to recharge the used catalyst. The regenera- tor combusts deposited coke off of the surface of the catalyst particles inside a large fluidized bed, with internal cyclones to remove particles from the flue gas stream before venting to the atmosphere. The recharged catalyst then recirculates through the rising main and the process is repeated [1, 2]. The advantages of using fluidized beds are often reduced due to particle loss via entrainment. This loss of particle mate- rial results in a subsequent financial and environmental bur- den associated with the control of potential air pollution issues as well as having to continually add more catalyst to maintain the bed level [3–6]. The correct design and operation are es- sential to avoid unnecessary solid loss via entrainment in a flu- idized bed. To avoid confusion, elutriation is defined as the process of particle separation, according to particle size, as the particles are carried out of the fluidized bed, while entrain- ment is the total amount of material carried out of the bed. Although there is a level of agreement in the literature in terms of the physical aspects which affect particle elutriation, no one model or equation is available that can accurately pre- dict particle loss rates from a fluidized bed. Most models are likely to have a prediction rate of – 100 % with uncertainty re- garding the validity of assumptions used in the development of the equations [3,7]. To overcome this uncertainty, models are either simplified correlations based on a specific lab-based unit, inaccurate in different situations or comprehensive large- scale models that try to compile all possible interactions found in a given fluidized bed system [8].  2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim http://www.cet-journal.com Josh M. Whitcombe 1 Roger D. Braddock 1 Igor E. Agranovski 1 1 Griffith School of Engineering, Griffith University, Brisbane, Australia. – Correspondence: Prof. I. E. Agranovski (I.Agranovski@griffith.edu.au), Griffith School of Engineering, Griffith University, Brisbane, 4111 QLD, Australia. 1336 Chem. Eng. Technol. 2008, 31, No. 9, 1336–1341