CONTINUOUS FLOW REACTOR WITH REMOVABLE INSERT WITH BAFFLES

Abstract

This specification discloses a continuous flow reactor comprising a reaction vessel enclosing a rotating shaft which bears multiple impellers. A removable insert comprising baffles separates the interior of the reaction vessel into multiple interconnected chambers each enclosing at least one impeller, and a gap between the outer edge of the baffle and the vessel's interior wall facilitates insert removal for maintenance but avoids excessive backmixing.

Claims

1.-25. (canceled)

26. A continuous flow reactor comprising: a reaction vessel comprising an inlet and an outlet; a rotating shaft bearing at least two agitators enclosed by the reaction vessel; means to drive the rotating shaft; and a reaction vessel insert comprising a baffle attached to a support, the baffle comprising an opening permitting the flow of liquid through it; wherein the insert is removably fixed within the reactor such that in use a baffle separates the agitators into interconnected mixing chambers each enclosing an agitator, while maintaining a gap between 20 m and 100 m in size between the outer edge of the baffle and the inner surface of the reaction vessel that facilitates removal of the insert but minimises backmixing between the chambers.

27. The continuous flow reactor as claimed in claim 26 where the gap is between 30 m and 75 m in size.

28. The continuous flow reactor as claimed in claim 26 where the reaction vessel is cylindrical and has a length to diameter ratio between 1:1 and 100:1.

29. The continuous flow reactor as claimed in claim 26 where the inlet and outlet are positioned at opposite ends of the reaction vessel.

30. The continuous flow reactor as claimed in claim 26 comprising multiple rotating shafts each bearing at least two agitators.

31. The continuous flow reactor as claimed in claim 26 where a rotating shaft is positioned substantially centrally with respect to the reaction vessel.

32. The continuous flow reactor as claimed in claim 26 comprising between 30 and 200 agitators per metre length of the rotating shaft.

33. The continuous flow reactor as claimed in claim 26 where the agitators are substantially perpendicular to the longest axis of the rotating shaft.

34. The continuous flow reactor as claimed in claim 26 where the rotating shaft and/or the agitators are supported by a bearing.

35. The continuous flow reactor as claimed in claim 26 where the agitator is an impeller or a turbine.

36. The continuous flow reactor as claimed in claim 26 where the agitator is selected from a Rushton impeller, a pitch blade impeller, a high head closed channel Impeller, a vortex impeller, a centrifugal screw impeller, a propeller, a shredder impeller, an anchor-type impeller, and a mixed flow impeller.

37. The continuous flow reactor as claimed in claim 26 where the removable reaction insert comprises a spacer.

38. The continuous flow reactor as claimed in claim 26 where the support comprises a support rod.

39. The continuous flow reactor as claimed in claim 26 where a support rod is adapted to deliver a reaction component.

40. The continuous flow reactor as claimed in claim 26 where all adjacent agitators are separated by a baffle.

41. The continuous flow reactor as claimed in claim 26 where a baffle comprises multiple openings.

42. The continuous flow reactor as claimed in claim 26 where the means to drive the rotating shaft is coupled to the shaft magnetically or by a rotating coupler, swivel joint or rotary union.

43. The continuous flow reactor as claimed in claim 26 where the interconnected mixing chambers comprise a reagent basket.

44. The continuous flow reactor as claimed in claim 26 where the interconnected mixing chambers comprise a turbulence increasing structure.

45. The continuous flow reactor as claimed in claim 26 where the turbulence increasing structure is a baffle.

46. The continuous flow reactor as claimed in claim 26 comprising means to heat and/or cool the reaction vessel contents.

47. The continuous flow reactor as claimed in claim 26 comprising: a vertically-installed reaction vessel comprising an inlet at its bottom and an outlet at its top; a vertical rotating shaft bearing between 30 and 200 impellers per metre length of the shaft enclosed by the reaction vessel; means to drive the rotating shaft; and a reaction vessel insert comprising a baffle attached to a support, the baffle comprising an opening permitting the flow of liquid through it; wherein the insert is removably fixed within the reactor such that in use the baffle is held substantially perpendicular to the shaft and separates the agitators into interconnected mixing chambers each enclosing an agitator, while maintaining a gap between 20 m and 100 m in size between the outer edge of the baffle and the inner surface of the reaction vessel that facilitates removal of the insert but minimises backmixing between the chambers.

48. The use of a continuous flow reactor as claimed in claim 26 to carry out a reaction.

Description

FIGURES

[0271] FIG. 1: Cross-sectional view of a continuous flow reactor according to the invention.

[0272] FIG. 2: Plan view of baffle 10.

[0273] FIG. 3: Isometric cut-away view of the modified continuous flow reactor of FIG. 1.

[0274] FIG. 4: Cross-sectional view of the reaction vessel insert from FIG. 3, showing the parallel support rods flanking the central rotating shaft.

[0275] FIG. 5: Isometric view of the reaction vessel insert shown in FIG. 4.

[0276] FIG. 6: Detailed cross-sectional view of rotating shaft 6 adapted for gas or liquid delivery.

[0277] FIG. 7: Cross-sectional view of a removable catalyst basket.

[0278] FIG. 8: Isometric cut-away view of an alternative continuous flow reactor embodying the invention.

[0279] FIG. 9: Isometric view of the reaction vessel insert used in the reactor of FIG. 8.

[0280] FIG. 10: Exploded view of the upper end of the reaction vessel insert shown in FIG. 9.

[0281] FIG. 11: Exploded view of the lower end of the reaction vessel insert shown in FIG. 10.

[0282] FIG. 12: Enhanced detail of the reactor of FIG. 8, showing the gap between the baffle outer edges and reaction vessel inner surface.

[0283] FIG. 13: Schematic diagram for the residence time distribution (RTD) experiments.

[0284] FIG. 14: Schematic diagram for the oxygen mass transfer experiments.

[0285] FIG. 15: Signal emitted by the dye detector for different tracer concentrations.

[0286] FIG. 16: Residence time distribution without baffles at a liquid flow rate of 300 mL min.sup.1 and impeller rotation speed 650 min.sup.1.

[0287] FIG. 17: Smoothed residence time distribution without baffles at a liquid flow rate of 300 mL min.sup.1and impeller rotation speed 650 min.sup.1.

[0288] FIG. 18: Residence time distribution with and without baffles at a liquid flow rate of 300 mL min.sup.1 and impeller rotation speed 650 min.sup.1.

[0289] FIG. 19: Gas-Liquid Mass transfer (k a) for different mixing rates and different gas flow rate (a) at a liquid flow rate of 200 mL min.sup.1 and (b) at a liquid flow rate of 500 mL min.sup.1. Horizontal dashed lines indicate the k.sub.La values achieved by the reactor packed randomly with glass beads and the bubble column.

[0290] FIG. 20: Solid-liquid mass transfer (k.sub.sa.sub.s) for different mixing rates and different liquid flow rate.

[0291] FIG. 21: Calculated power number of eight Rushton impellers across different Reynolds numbers.

DETAILED DESCRIPTION

Example 1: Continuous Flow Reactors

[0292] An example continuous flow reactor embodying the invention is pictured in FIG. 1, with different angles and close ups of elements and groups of elements shown in FIGS. 2, 3, 4 and 5.

[0293] The reactor comprises a vertically-installed cylindrical reactor tube 1 constructed from stainless steel. The open ends of tube 1 are sealed using flanges 2, which are used to provide a good seal and enable the reaction vessel to contain high internal pressures. An oil/heating jacket 3 is employed to give control over a reaction's thermal parameters, and the reaction vessel is provided with an inlet 4 through which reactants and reagents can be introduced, and an outlet 5 through which reaction products can be collected.

[0294] A central rotating shaft 6 runs vertically upwards through the installed reactor. The shaft is coupled to a magnetic drive housed above region 7 to provide a leak-free linkage to a motor 7a or other suitable driving device (not shown). Depending on the length and shaft properties, additional supports such as bearings (not shown) can be used to reduce vibration of both the shaft itself and/or any impellers (or other agitating means) attached to it. Additional inlets 8 may be present, and placed so that they can be used to deliver a reaction component to a suitable configured support rod (see Example 2).

[0295] Rushton impellers 9 are fixed along the central shaft at regular intervals. The number of impellers can vary from 2 to 100 depending on the length of the reactor and desired reaction parameters. The type of impeller used (Rushton, pitch blade etc.) can be selected depending on the desired mixing characteristics. Impellers may also be replaced by other agitators or turbulence generating structures such as turbines. In the vertically-installed reactor of FIG. 1, the impeller blades extend from the central shaft in a substantially perpendicular fashion.

[0296] Baffles 10 are fixed consecutively between each impeller and separate the internal reaction vessel space into a series of interconnected mixing chambers each containing a single impeller. For convenience, the baffles are attached to support rods 11 which are installed in parallel to rotating shaft 6, one support rod on each flank. This allows easy removal of the baffles for cleaning and maintenance.

[0297] FIG. 2 shows that the baffles have a flat annular or disk-like structure whose diameter is such that the outer edges of the baffle can extend across the corresponding diameter of the reaction vessel leaving a small gap (e.g. of about 20-100 m) between the outer edge of the baffle and the reaction vessel interior wall. The baffles comprise a central opening 12 which accommodates shaft 6, optionally loosely enough to additionally permit flow of the reaction liquid through the apparatus. Opening 12 may be varied in diameter according to the specific reaction residence times required, and other openings 13 may also be introduced to the baffle structure to give further control over the passage of solvents.

[0298] Additional features are shown in the reactor variant depicted in FIGS. 8 to 12. The reactor comprises cylindrical reactor tube 1 made from stainless steel. Although a variety of dimensions can be used, the internal diameter of the tube is 26 mm, making it suitable for pilot scale reactions. The reactor tube is capped with custom flanges 2, the top accommodating a magnetic drive and an outlet port 5, and the bottom an inlet port 4. The reactor houses removable reaction vessel insert 15, which has spacers 16 at each of its terminals, offsetting the position of baffles 10 and aligning them with heating jacket 3. In the example shown, the spacer comprises a two-part structure, with the part facing the reaction interior engineered to received components and provide stability. Indentations 16a are present in the concave surface of the spacer, with these features aiding mixing.

[0299] A close up of insert 15 is shown in FIG. 9. It comprises a series of 9 baffles stacked equidistantly along four rods acting as a support, arranged around and passing through each of the baffle's poles. Two of the opposing hollow support rods are engineered into feeding pipes 14 to allow passage of a liquid or gas material into a reaction, for example by having one or more holes drilled into the rod surface. These feeding tubes are aligned with additional inlet/outlet ports in the reactor structure. The other two support rods are not so engineered and therefore only perform the role of supporting the baffles by means of a welded joint, allowing them to divide the reaction vessel interior into 10 interconnected mixing chambers.

[0300] In this particular variant, rotating shaft 6 forms part of the removable insert, running centrally through apertures in baffles 10. FIGS. 9-11 show that at each shaft terminal are disks 17 that rest against bushings 18, which are in turn accommodated by bushing holders 19 engineered into spacers 16. This arrangement allows easy removal of all of the internal components of the reactor from the reaction vessel, making maintenance (or modification) far easier.

[0301] FIG. 12 shows that when the insert is removably attached to the reactor interior (not shown), there is a gap 10a between the outer edge of the baffles and the reaction vessel interior.

[0302] A further variant of the reactors shown in the figures may comprise several impeller shafts each bearing multiple impellers separated by baffles. In an example of such a variant, the impeller shafts may pass vertically upwards through the reaction vessel as described in FIG. 1, with multiple shafts distributed evenly across the horizontal width of the reactor tube. Rotating shaft groups may be arranged such that their overall grouping is still overall placed centrally with respect to the reaction vessel volume.

Example 2: Continuous Flow Reactor Operation

[0303] During use of the reactor shown in FIG. 1, reactants are introduced to inlet 4 and flow upwards from the bottom of the reactor (an up flow configuration), where they are mixed by the rotating impellers 9 and react to form product before exiting the reactor at outlet 5. For some reactions other flow orientations (such as down flow or horizontal flow) may be preferred. In such cases, the reactor can be installed horizontally rather than vertically (such that the longest length of the rotating shaft extends horizontally and the shaft rotates around its horizontal axis), and/or the positions of the inlet 4 and outlet 5 chosen differently (for example, for a downflow configuration the same vertical installation of the reactor could be used but the inlet 4 could be placed on the top and the outlet 5 on the bottom of the reaction vessel).

[0304] For two phase reactions involving liquids and gases, suitable reactants may be added along the reactor length. A typical example is introduction of a liquid from the bottom inlet 4 while a gas may be injected along the central shaft 6 and through the rotating impellers 9, or through the vertical support tubes 13. The former configuration is shown in more detail in FIG. 6, which shows a cut-away view of an adapted shaft; but a similar arrangement can be used when the support tube(s) 13 are used to deliver a component, via additional inlets 8.

[0305] In FIG. 6, a static tube 14 engineered with perforations 20 runs internally to the rotating shaft to provide gas flow along its length. The shaft and/or impellers are similarly engineered with perforations 21 to allow passage of the reactant to the reaction vessel interior. The diameter of the perforations 20 and 21 can be altered along the tube in order to provide even delivery throughout the reactor at low gas flow rates and pressure. Although described for a gas-liquid process, this approach could also be useful for difficult liquid-liquid reactions where one fluid may enter from the bottom and the other be distributed along the reactor to (for example) manage reaction heat or formation of gas.

[0306] When carrying out catalytic reactions, catalysts can be provided to the continuous flow reactor as a suspended slurry. Alternatively, catalyst may be supported in stainless steel mesh baskets 22 which surround the impellers 9, as shown in FIG. 7. This directs the fluid in defined flow patterns 23 through the catalyst, allowing for superior mass and heat transfer.

Example 3: Continuous Flow Reactor Properties

[0307] To show the utility of the continuous flow reactors of this specification a series of experiments were performed. For each test, a reactor was constructed according to in the form of an acrylic tube 24 with an internal diameter of T=64 mm and a length of L=300 mm, the ends of which were sealed using flanges. Eight Rushton impellers 9 with an impeller diameter of D=25.6 mm (D/T=0.4) were spaced evenly along the central rotating shaft of the reactor. The distance between each impeller was C=32 mm (AC/D=1.25). In general, tests were performed in a control model without baffles, and compared to a reactor comprising horizontally arranged baffles equally spaced across the whole of the reactor height, with the distance between baffles was equal to C. Each baffle had a thickness of 2 mm. An electric motor was used to rotate the central shaft around a 6 mm rotation axis at 500 to 1500 revolutions per minute (rpm), rotating the impellers in turn. All experiments were conducted under ambient conditions.

Example 3a: Residence Time Distribution and Determination of the Number of Continuous Stirred Tank Reactors in Series

[0308] As shown in FIG. 13, water 25 was pumped through the reactor with a Knauer 80P pump at 100 to 300 (+2%) mL min.sup.1. A stirring rate of 200 rpm was used. The liquid was mixed with fluorescent dye (fluorescein) with a syringe pump 26 (World Precision Instruments, AL-1010) at 0.3 mL min.sup.1. A dye detector 27, containing a laser and detectors with light filter, continuously monitored the fluorescent dye concentration in the liquid phase at the reactor outlet, which passed in turn to liquid waste 28. The response was determined to be linear for a wide range of tracer concentrations (FIG. 15).

[0309] The residence time distribution density, E(t), was calculated as follows:

[00001]$E\left(t\right)=\frac{C_I\left(t\right)}{{}_0^{}{C}_I\left(t\right)\mathrm{dt}}$

where C.sub.I(t) is the tracer concentration over time. The residence time distribution (RTD) of n continuously stirred tank reactors CSTRs in cascade was also determined by combining the RTD of the CSTR n times itself [Toson P., Doshi P., Jajcevic D.; Processes 7, 2019]. MacMullin and Weber [MacMullin R., Weber M.; Trans. Am. Inst. Chem. Eng. 31, 1935, 409] were the first to propose the function for n CSTRs in connected in series:

[00002]$E\left(t\right)={\left(\frac{N}{\stackrel{}{t}}\right)}^N{t}^{N-1}\frac{\exp \left(-\frac{\mathrm{Nt}}{\stackrel{}{t}}\right)}{\left(N-1\right)!}$

where N is the number of CSTR and t is the mean residence time.

[0310] First, the RTD in the reactor without baffles was measured and used as a reference experiment. It can be seen in FIG. 16 that the data obtained contains noise and is ascribed to the presence of gas bubbles in the fluid stream. In order to remove the noise, a moving average was applied to the data (FIG. 17).

[0311] In FIGS. 16 and 17, it can be observed that a fraction of the tracer molecule flows instantly out of the reactor without baffles. After the peak at ca. 70, the tracer concentration decreases following an exponential decay. This type of a curve is typical for a CSTR: the smaller the concentration of the tracer left in the reactor, the longer it takes for this fraction to flow out. As a result, the residence time distribution is broad resulting in poorly controllable continuous processes. Indeed, some molecules spend little time in the reactor, as they pass immediately through the system, whereas others spend longer times.

[0312] The same experiment was then performed with baffles separating the impellers into interconnected mixing chambers according to the present specification. FIG. 18 shows that the addition of baffles to the control apparatus results in a narrowing of the distribution, as well as shifting the peak to 150 s. This means that the baffles allow more effective control of the residence time of the elements passing through the reactor, in turn improving reaction control and timing. The experimental residence time distribution of the multiple impeller/baffle system accurately matched the theoretical model of 8 ideal CSTRs connected in series, despite the marked improvement in structural simplicity. Even further improvements in residence were observed for the reactor shown in FIGS. 8 to 12, with the system matching around 30 ideal CSTRs connected in series (i.e. 3 CSTRs in series for each real mixing chamber in the reactor).

Example 3b: Oxygen Mass Transfer Experiments and Determination of the Gas-liquid Mass Transfer Coefficient

[0313] Mass transfer experiments were carried out according to the basic scheme outlined in FIG. 14. Deoxygenated water 29 was pumped at 200 to 500 (+2%) mL min.sup.1 through the reactor. Air 30 was introduced into the inlet of the reaction and the flow rate was controlled using a Bronkhorst mass-flow controller 31 at 100 to 200 (+0.2%) mL min.sup.1. The oxygen concentration in the liquid phase was measured continuously with a (PyroScience, PICO.sub.2) oxygen probe 32 at the reactor inlet (+0.02% accuracy). The oxygen probe was calibrated with air (21% O.sub.2) at ambient temperature.

[0314] Gas-Liquid mass transfer measurements were also measured in a bubble column and packed bed reactor (PBR) configuration. To turn the reactor into a bubble column, the baffles and the impellers were removed. The baffles and impellers were also removed for the PBR configuration, and the empty reactor tube was filled with spherical glass beads with a diameter of 3 mm. The length of the packed bed was 100 mm.

[0315] For a physical absorption in a continuous flow, the overall O.sub.2 mass flow rate absorbed .sub.O.sub.2, was calculated using the following equation:

[00003]${}_{O_2}=\frac{Q_L{\mathrm{dC}}_{O_2}}{S_c\mathrm{dZ}}=k_{L}a\left(C_{O_2}^{*}-C_{O_2}\right)$

where Q.sub.L is the liquid flow rate, Sc is the reactor cross section,

[00004]$\frac{{\mathrm{dC}}_{O_2}}{\mathrm{dZ}}$

is the oxygen concentration variation over the reactor length, k.sub.L is the local gas-liquid mass transfer coefficient, a is the effective interfacial area of gas bubble,

[00005]$\frac{{\mathrm{dC}}_{O_2}}{\mathrm{dZ}}$

the solubility of O.sub.2 in the liquid phase and C.sub.O.sub.2 is the oxygen concentration in the liquid phase. By integrating the equation along the reactor height and isolating the k.sub.L a value, the gas-liquid mass transfer coefficient was determined as follows:

[00006]$k_{L}a=\frac{Q_L}{L_{\mathrm{reac}}{S}_C}\ln \left(\frac{C_{O_2}^{*}-C_{O_2}^{\mathrm{in}}}{C_{O_2}^{*}-C_{O_2}^{\mathrm{out}}}\right)$

where L.sub.reac is the reactor length, C.sub.O.sub.2.sup.in and C.sub.O.sub.2.sup.out are respectively the oxygen concentration in the liquid phase at the reactor inlet and outlet. The oxygen solubility was determined with the Winkler table [Montgomery, H.A.C.; Thom, N. S.; Cockburn, A.; J. Appl. Chem. 14, 1964, 280], which has been shown to be suitable for oxygen analysis [Carpenter, H.; Limnol. Oceanogr. 10, 1965, 135].

[0316] The gas-liquid mass transfer coefficient, k.sub.La, was measured and compared for different configurations: with baffles, without baffles, as a bubble column and a PBR. For the PBR configuration L.sub.reac was equal to 100 mm. The results for various mixing, gas and liquid flow rates are shown in FIG. 20.

[0317] Firstly, it can be seen that by increasing the mixing rate, i.e. the speed of the impellers, the mass transfer was increased at both flow rates and for with and without baffles. Mass transfer was also improved in every experiment using higher gas flow rates and higher liquid flow rates. Comparing the difference between the use of baffles and without baffles, it is clear the system with baffle offers a greater k.sub.La value compared to without baffles. This difference is even more noticeable at high mixing and liquid flow rates. The vertical baffles control the RTD more effectively by maintaining more molecules in the reactor, as seen above. Therefore, the gas and the liquid remain in contact for a longer amount of time, thus increasing the performance of the reactor.

[0318] The graph also displays the mass transfer for a bubble column and a PBR. As there is no mixing for those systems, the k.sub.La values are represented by dashed horizontal lines in the wide mixing range to compare with the others configuration involving mixing. As shown in FIG. 19, the mass transfer is much greater with the baffle even at low impeller speeds. The rotary motion of the impeller reduces the gas bubble size, increasing the interfacial area and thus the mass transfer. Furthermore, compared to the bubble column arrangement, the baffled arrangement imparts a greater gas hold-up time which also increases mass transfer in the baffled reactor.

[0319] Finally, the continuously stirred baffle reactor outperformed the PBR when the impeller speed reached approximately 1000 rpm. Hence, high mass transfer rates (ka) can be achieved by simply increasing the rotational speed of the impellers.

Example 3b: Benzoic Acid Mass Transfer Experiments and Determination Solid-liquid Mass Transfer Coefficient

[0320] For this experiment, the baffles were removed and only one Rushton impeller was used. The impeller was surrounded by a basket with a height of 32 mm and an internal mesh arrangement that protected the solid particles therein from the rotating impeller.

[0321] Water was pumped through the continuous flow reactor at 100 to 500 (+2%) mL min.sup.1. Benzoic acid (Sigma-Aldrich, >99.5%), in pellet form (5 mm in diameter and 10 mm in length) was placed in the basket. Samples were collected from the reactor outlet and the benzoic acid concentration in the liquid was measured with a Gas Chromatograph (GC).

[0322] The overall Benzoic Acid (BA) mass flow rate absorbed @BA was calculated as follows:

[00007]${}_{\mathrm{BA}}=\frac{Q_L{\mathrm{dC}}_{\mathrm{BA}}}{S_c\mathrm{dZ}}=k_s{a}_s\left(C_{\mathrm{BA}}^{*}-C_{\mathrm{BA}}\right)$

where k.sub.s is the local solid-liquid mass transfer coefficient, a.sub.s is the effective interfacial area of the solid particle, C.sub.BA* the solubility of benzoic acid in the liquid phase and C.sub.BA is the benzoic acid concentration in the liquid phase. The solid-liquid mass transfer coefficient was then determined by integrating the preceding equation along the basket height:

[00008]$k_S{a}_S=\frac{Q_L}{L_{\mathrm{basket}}{S}_C}\ln \left(\frac{C_{\mathrm{BA}}^{*}-C_{\mathrm{BA}}^{\mathrm{in}}}{C_{\mathrm{BA}}^{*}-C_{\mathrm{BA}}^{\mathrm{out}}}\right)$

where L.sub.basket is the length of the basket, C.sub.BA.sup.in and C.sub.BA.sup.out are respectively the benzoic acid concentration in the liquid phase at the reactor inlet and outlet. The benzoic acid solubility was taken from Swarbrick and Carless [Swarbrick J.; Carless, J. E.; J. Pharm. Pharmacol. 16, 1964, 596].

[0323] FIG. 19 compares the solid-liquid mass transfer coefficient k.sub.sas for different mixing, gas flow, and liquid flow rates. It can be seen that increasing mixing rate improves the liquid velocity in the reactor, enhancing the k.sub.sas value and improving the solid-liquid mass transfer. Indeed, at the liquid flow rate (QL) of 200 mL min.sup.1, the solid-liquid mass transfer almost doubled compared to QL=100 ml min.sup.1. The results demonstrate the possibility of using baskets with solid materials in the continuous flow reactor. The solid-liquid mass transfer rate was positively correlated to impeller rotation speeds, this demonstrates that the continuously stirred baffle reactor decouples mass transfer from fluid viscous forces by varying impeller rotation speeds.

Example 3c: Determination of Impeller Power Number

[0324] The overall power consumption of eight Rushton impellers was determined by subtracting the measured power consumption of impellers rotating in water from the power consumption rotating in air for a specific rotation speed. Afterwards, the power number, Np, was determined with the following equation:

[00009]$N_p=\frac{P}{N^3{D}^5}$

where P is the power consumption, is the density, N is the rotation speed and D is the impeller diameter. The power number is used to determine the power consumption, considering the impeller geometry.

[0325] FIG. 21 shows the calculated power number of eight Rushton impellers across different Reynolds numbers. It can be seen that for Reynolds numbers greater than 8000, the power number is on average ca. 23. In the model system described above, there are eight Rushton impellers, therefore each impeller has an approximate power number of between 2 and 3. This is surprisingly much lower than the typical power number for a Rushton impeller which is generally between 4 and 6 (Chapple D. et al., Chem. Eng. Res. Des. 80, 2002, 364; Jaszczur M. et al., Energies 13, 2020; Nouri R. M. et al., J. Chem. Eng. Japan 31, 1998, 848; Furukawa, H. et al., J. Chem. Eng., 2012; Jafari, R. Solid Suspension and Gas Dispersion in Mechanically Agitated Vessels, 2010. Hence, the overall power consumption in this system is substantially lower for the same overall performance, and the energy consumption on mixing may be 30-55% lower compared to conventional reactors.

Example 4: Effect of Gap on Insert Removal and Backmixing

[0326] As already mentioned, the reactors described in the present specification comprise a gap between the outer edges of a reaction insert's baffles and the reaction vessel's interior wall, making the insert more easily removable for maintenance or modification (and insertable thereafter). Table 1 shows the approximate insertion force (measured using an FSS015WNGX Honeywell Load Cell calibrated against known weights) required for inserts with varying gap distances in the example reactor described in FIGS. 8-11 and constructed using a 24 cm long, 26 mm internal diameter reaction vessel and an insert comprising 9 baffles 1 mm thick.

TABLE-US-00001 TABLE 1 Effect of Reactor Insert Gap on Insertion Force Gap Size (=distance between outer edge of the baffle and the inner surface of the reaction vessel, m) Approximate Insertion Force 10 >>1 Kg (Damage Caused) 20 0.5 Kg 35 0.1 Kg 60 0.05 Kg 90 Negligible 125 Negligible 200 Negligible 300 Negligible

[0327] It can be seen that a relatively small gap increases the ease by which the support/spacer/baffle construct dramatically: a 10 m gap insertion caused major damage to the insert, but 20 m results in a reasonably workable insertion (and by analogy removal) force.

[0328] As already noted, allowing gaps between internal structures such as baffles and a reaction vessel's interior wall is known to cause backmixing. To see the effects in the reactors of the present specification, tests were carried out according to the method of example 3b on the example reactor described in FIGS. 8-12, constructed using a 24 cm long, 26 mm internal diameter reaction vessel and featuring an insert hosting 9 baffles 1 mm thick. A control with 0 gap size was prepared by applying heated viscous paraffin onto the 20 m gaps to seal them completely when cooled.

TABLE-US-00002 TABLE 2 Effect of Reactor Insert Gap on Backmixing Gap Size (=distance between outer edge of the baffle and the inner surface of Number of Ideal Stirred Tanks in the reaction vessel, m) Series 0 27 20 27 35 25 90 14 200 6.7 250 2.3

[0329] The results show the number of ideal stirred tanks that would be needed to replicate the same performance as the reactor of the specification, with a higher number representing a better performance in terms of reaction residence time control. Reactor performance can be seen to substantially deteriorate as the size of the gap increases, due to unwanted backmixing. However, despite teaching to the contrary in the art, small gaps appear to be reasonably well-tolerated and do not unduly reduce reaction control.

CONCLUSIONS

[0330] The experimental data above show that the continuous flow reactors disclosed herein have several advantages compared to earlier alternatives. Using baffles to create multiple mixing chambers each enclosing an agitator leads to a more uniform flow and a narrower residence time distribution compared to a control CSTR system without baffles, improving reaction control. Furthermore, the RTD curve for a system comprising 9 baffles (10 mixing chambers) accurately matches the theoretically calculated RTD curve of up to 30 ideal CSTRs in sequence. This shows that the RTD of the instant reactor can be conveniently controlled in relation to the number of CSTRs required by changing the number of baffles.

[0331] The gas-liquid mass transfer results show the advantage of the mixing impellers for increasing mass transfer (k.sub.La) rates compared to typical packed bed reactors and bubble columns with no mixing. These results also show that having baffles significantly increases the gas-liquid mass transfer compared to a control system lacking them or the mixing chambers they provide. The solid-liquid mass transfer experiment further compounds the advantage of the use of impellers at low and medium liquid flow rates. Indeed, in the reactors of the present specification, the overall power consumption is significantly less than expected for a given number of impellers, leading to improved efficiency.

[0332] Finally, it can be seen that the removable nature of the insert and the gap maintained between the outer edge of the insert baffles contributes to the reactor being easy to clean and maintain, while avoiding excessive backmixing and a concomitant loss of reaction control.