Computers and Chemical Engineering 35 (2011) 2326–2333 Contents lists available at ScienceDirect Computers and Chemical Engineering j ourna l ho me pag e: w ww.elsevier.com/locate/compchemeng Rational design of heating elements using CFD: Application to a bench-scale adiabatic reactor Pablo Marín, Salvador Ordó ˜ nez ∗ , Fernando V. Díez Department of Chemical Engineering and Environmental Technology, University of Oviedo, Facultad de Química, JuliánClavería 8, Oviedo 33006, Spain a r t i c l e i n f o Article history: Received 7 May 2010 Received in revised form 17 November 2010 Accepted 19 November 2010 Available online 27 November 2010 Keywords: CFD Heat losses compensation Un-steady state Lab-scale testing a b s t r a c t This article explores the use of computational ﬂuid dynamics (CFD) modelling for designing an oven which enables the adiabatic operation of a chemical reactor at bench-scale. For accomplishing this scope, the oven consists of electrical heating elements, air circulation system and a control loop that uses the temperature inside the reactor as set-point for the reactor wall temperature. Depending on the spatial conﬁguration of the air ﬂow and the heating elements, as well as the air ﬂow rate, different temperature proﬁles within a given oven section are obtained, being appropriate those leading to uniform reactor wall temperatures and fast dynamic response. The use of CFD allows, by obtaining temperature maps within the oven, the selection of appropriate conﬁgurations. The optimal conﬁguration adopted has been experimentally validated in a lab-scale adiabatic reactor working with both particulated and monolithic catalyst beds. © 2010 Elsevier Ltd. All rights reserved. 1. Introduction Many industrial processes, such as chemical reactors, are designed to work at adiabatic or near-adiabatic conditions. This is easily achieved in large devices, i.e. cylindrical apparatus with large diameter, and hence low surface–volume ratio, which can be easily designed to produce very low heat losses to the ambient. However, in small devices, such as lab- or bench-scale experimen- tal units, heat losses may be signiﬁcant, and operation at adiabatic conditions becomes very difﬁcult. As a consequence, scale-up is not straightforward, and additional experiments or sophisticated mod- elling may be required. The most usual way of achieving adiabatic conditions at lab-scale is by using thermal insulation, but using this method heat losses are not completely eliminated, and thermal insulation increases signiﬁcantly the thermal inertia of the system, which strongly affects the un-steady state behaviour of the system (Luss, 1997; Menzingera, Yakhnina, Jareeb, Silvestonc, & Hudginsc, 2004): when temperature increases or decreases, insulation acts as a heat sink or source, and temperature proﬁles may be markedly modiﬁed. Both factors (heat losses and thermal inertia), must be taken into account when modelling the process (Beld & Westerterp, 1996). Both thermal inertia of the insulation and heat losses can be partially solved by substituting the insulation layer by a vac- uum jacket (overall radial heat transfer coefﬁcients of 1.5 W/m 2 K using jacket pressures of 0.02 mbar), but heat losses are still rele- ∗ Corresponding author. Tel.: +34 985 103 437; fax: +34 985 103 434. E-mail address: sordonez@uniovi.es (S. Ordó ˜ nez). vant and must be taken into account in the process modelling (Bos, Vandebeld, Overkamp, & Westerterp, 1993; Beld, Borman, Derkx, van Woezik, & Westerterp, 1994; Neophytides & Froment, 1992). To overcome the disadvantages of thermal insulation other alternatives have been proposed, based on the compensation of the heat losses using electrical heating around the experimental appa- ratus which is desired to operate adiabatically (Cunill, van de Beld, & Westerterp, 1997; Nieken, Kolios, & Eigenberger, 1994; Züﬂe & Turek, 1997). However, it was not possible to compensate com- pletely all the heat losses, and in most cases the use of a single electrical furnace in dynamically operated devices creates impor- tant disturbances in the axial direction with respect to adiabatic conditions. A further improvement of this compensating system was proposed by Fissore et al. (2005) (Hevia, Ordonez, & Diez, 2006). According to this design, heat losses from an experimental apparatus are compensated by a multi-section oven, every section controlled independently, so that only the required amount of heat is compensated in each section. Dynamic processes with impor- tant axial temperature proﬁles, such as catalytic oxidations (Liu et al., 2007), were found not to be signiﬁcantly disturbed when using an oven with 7 heating sections. Nevertheless, it was found that the compensation of the heat losses around the apparatus was not totally uniform, resulting in the occurrence of radial gradients inside the reactor. A procedure for a reliable numerical solution of transport equa- tions (system of partial differential equations) is needed for the simulation of temperature proﬁles within each oven section. In the last years, different computational ﬂuid dynamics (CFD) codes based on the ﬁnite element method have been developed to solve 0098-1354/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.compchemeng.2010.11.005