(341f) Scaling Down Oscillatory Baffled Reactors

“Scaling down Oscillatory Baffled Reactors”

Adam P. Harvey (adam.harvey@ncl.ac.uk)*, Anh N. Phan (anh.phan@ncl.ac.uk), Valentine Eze (valentine.eze@ncl.ac.uk), Fatimah Mohd Rasdi (f.r.b.mohd-rasdi@ncl.ac.uk), Richard Abernethy (richard.abernethy@ncl.ac.uk)

*Corresponding Author:

Professor Adam P Harvey, School of Chemical Engineering and Advanced Materials (CEAM), Newcastle University, UK NE1 7RU

OVERVIEW

Historically one of the most significant issues for intensified technologies was scale-up. For oscillatory baffled reactors (OBRs) there have been many studies incorporating various aspects of scale-up over the last 20 years.  Over the last 8 years, however, one strong feature of OBR research has been instead to scale them down to so-called “mesoscale” i.e. to diameters of approximately 5mm. This allows various advantages of oscillatory baffled reactors to be exploited in new ways. For instance, the OBR’s ability to accommodate reactions of residence times of the order of hours in a relatively compact reactor with plug flow allows long reactions to be screened or optimised in continuous operation, whilst consuming very low levels of feed and producing little waste. This can be very important for expensive feedstocks or very hazardous waste. Another particular advantage of the “mesoOBR” is that the reactor can at the same time uniformly suspend solid particles. This is difficult to do in other forms of continuous screening reactor.

This paper will give an overview of the development of mesoscale OBRs, explaining their current and potential uses.

DESIGN

The development of mesoOBRs has involved investigation of a number of different baffle designs, including conventional orifice plate designs, central (axial) baffles of various geometries and helical baffles. There are strengths and weaknesses of each design. Helical baffles, for instance, exhibit an “operating window” for plug flow that is approximately an order of magnitude wider than the other designs, allowing an enormous flexibility in operating conditions. The latest compact designs of mesoOBRs are even more compact as they involve serpentine channels. They are designed to maximise residence time per unit area of laboratory bench, to allow “longer” reactions (of the order of a few hours) to be screened.

PROCESS SCREENING

The portfolio of systems that can be screened using this device is gradually being increased in terms of the number of phases present. So far examples of liquid systems, liquid-liquid, liquid-solid and liquid-liquid-solid systems have been assessed. In each case multi steady state screening has been demonstrated, followed by dynamic screening (continuous variation of one input parameter, typically residence time or molar ratio), then multivariable versions of both.

One of the themes of the paper will be the ease of experimentation allowed by such small-scale continuous screening devices. This reduced barrier to experimentation has led to more “adventurous” experimentation. This is illustrated using ongoing in-house case studies on:

(i) The kinetics of biodiesel production: biodiesel is a very well known reaction, but nonetheless the system has been used to identify a previously unrealised set of conditions that allow biodiesel to be produced in a residence time of 2 minutes. This should be contrasted with the conventional commercial residence time of 1-2h. This could be the basis of a 30 – 60 fold reduction in reactor volume.

(ii) The crystallization of L-glutamic acid: a new geometry (tetrahedral) of L-glutamic acid crystal has been identified. The size (sub 20mm) and shape of this new form are more uniform than that achieved by other production methods.

(iii) The kinetics of imine synthesis: the rate constants have been rapidly determined, from dynamic multidimensional screening, using online FTIR.