Residence Time Distribution Measurements Research Articles

Measuring residence time distributions of wood chips in a screw conveyor reactor

In rotating screw conveyors both the average and the distribution of the residence time influence the extent and the uniformity of the transformation. Experimenters have applied two distinct experimental approaches to obtain the residence time distribution of granular solids in longitudinal reactors: 1) measuring the mass flow rate of product at the exit from the reactor in response to a step change (either positive or negative) in the mass flow rate of feedstock into the reactor or 2) measuring the appearance of a tracer in the flow exiting the reactor in response to either a pulse or a step change addition of tracer in the inlet. We found that all three methods reveal residence time distributions that are approximately normal (i.e., symmetrical and bell-shaped), but the distribution estimated from the pulse input of tracer exhibited a long trailing tail that was not detectable in either the positive or negative step changes. Second, we demonstrated that a normal probability plot proved valuable in displaying and analyzing the residence time distribution obtained by the pulse addition of tracer. Finally, we observed that all three methods yielded mean residence times that consistently differed from the nominal values. The positive step change averaged 8% shorter, the pulse addition of tracer averaged 7% longer, and the negative step change averaged 60% longer.

Characterization of an enzymatic packed-bed microreactor: Experiments and modeling

A micro packed-bed reactor (µPBR) based on two-parallel-plates configuration with immobilized Candida antarctica lipase B in the form of porous particles (Novozym® 435) was theoretically and experimentally characterized. A residence time distribution (RTD) within µPBRs comprising various random distributions of particles placed in one layer was computationally predicted by a mesoscopic lattice Boltzmann (LB) method. Numerical simulations were compared with measurements of RTD, obtained by stimulus-response experiment with a pulse input using glucose as a tracer, monitored by an electrochemical glucose oxidase microbiosensor integrated with the reactor. The model was validated by a good agreement between the experimental data and predictions of LB model at different conditions. The developed µPBR was scaled-up in length and width comprising either a single or two layers of Novozym® 435 particles and compared regarding the selected enzyme-catalyzed transesterification. A linear increase in the productivity with the increase in all dimensions of the µPBR between two-plates demonstrated very efficient and simple approach for the capacity rise. Further characterization of µPBRs of various sizes using the piezoresistive pressure sensor revealed very low pressure drops as compared to their conventional counterparts and thereby great applicability for production systems based on numbering-up approach.

Instantaneous transport of a passive scalar in a turbulent separated flow

The results of large-eddy simulations of flow and transient solute transport over a backward facing step and through a 180° bend are presented. The simulations are validated successfully in terms of hydrodynamics and tracer transport with experimental velocity data and measured residence time distribution curves confirming the accuracy of the method. The hydrodynamics are characterised by flow separation and subsequent recirculation in vertical and horizontal directions and the solute dispersion process is a direct response to the significant unsteadiness and turbulence in the flow. The turbulence in the system is analysed and quantified in terms of power density spectra and covariance of velocity fluctuations. The injection of an instantaneous passive tracer and its dispersion through the system is simulated. Large-eddy simulations enable the resolution of the instantaneous flow field and it is demonstrated that the instabilities of intermittent large-scale structures play a distinguished role in the solute transport. The advection and diffusion of the scalar is governed by the severe unsteadiness of the flow and this is visualised and quantified. The analysis of the scalar mass transport budget quantifies the mechanisms controlling the turbulent mixing and reveals that the mass flux is dominated by advection.

Investigations about the effect of fractal distributors on the hydrodynamics of fractal packs of novel plate and frame designs

Flow distributors are extensively adopted by chemical industry to distribute process fluid for the downstream equipment. Conventional flow distributors in general have limited number of outlets. Their low outlet densities cannot distribute the incoming process stream uniformly and efficiently, which consequently undermine the overall equipment performance. In contrast, fractal distributors can achieve high outlet densities because of their inherent self-similarity feature. In this study, we fabricated two fractal packs of novel plate and frame designs. Each fractal pack was comprised of a fractal distributor, a resin bed and a fractal collector. The fractal distributors/collectors in these two fractal packs have 256 and 16 outlets/inlets, respectively. In this study, we carried out both CFD modeling and experiments to understand the effect of outlet density on the hydrodynamics of the downstream resin bed. By comparing the predicted velocity fields and the measured residence time distribution (RTD) curves, we noticed that a fractal distributor of high outlet density can distribute the process stream uniformly at a low pressure drop to the resin bed. The process streams entering the resin bed reach plug flow profile within a short bed depth. As a result of the uniform flow distribution, mal-distribution such as channeling and dead space are minimized, and the process streams have a uniform RTD in the fractal pack. In comparison, the fractal distributor of low outlet density, which mimics some aspects of a conventional distributor, is found to have severe mal-distribution in the resin bed and consequently poor RTD inside the fractal packs. In addition, the RTD analysis confirms that the fractal distributor of high outlet density shows consistently good distributing performance for a wide range of operating flow rates. This study confirms the advantages of fractal distributors over conventional ones and demonstrates the benefit of fractal distributors in improving the overall performance of chemical equipment, indicating that fractal distributors are promising to enable process intensification for the chemical industry.

Measurement and simulation of the three-dimensional flow pattern and particle removal efficiencies in a large floating closed sea cage with multiple inlets and drains

This study investigated the three-dimensional flow structures in an ellipsoid, closed sea fish cage. The results are presented using computational fluid dynamics (CFD) simulations and experimental measurements. Experimental residence time distribution (RTD) measurement and CFD simulation are the best methods to study the hydrodynamics of inflow systems. Three-dimensional numerical simulations of the flow and transport characteristics of the system were conducted using a Reynolds-averaged Navier-Stokes equation approach and the results were compared to the measurements performed using acoustic Doppler velocimetry techniques. The objective of the investigation was to characterize the flow field generated in an ellipsoid, closed tank. The flow in the enclosed volume is driven by four inlets pipes integrated into the wall of the cage. The focus is on the turbulent structures and undesirable flow patterns that lead to reduced self-cleaning efficiency and a lower quality habitat for the fish through phenomena, such as recirculation zones or low velocity areas. Correlations between CFD and the experimental data confirm the adequate reproduction of hydrodynamic conditions and reinforce the predictive capabilities of numerical models as tools to simulate field scale closed containment systems or to optimize existing and future aquaculture designs. The simulation of aquaculture-like particles demonstrates that almost 100% of particles with a diameter ranging between 1μm and 3000μm are removed during a maximum of two hydraulic retention time (HRT) cycles. Smaller particles are removed via the upper-side outlets and larger particles are removed via the bottom outlet.

Radiotracer investigation on the measurement of residence time distribution in an ethyl acetate reactor system with a large recycle ratio

A radiotracer investigation was carried out on the measurement of residence time distribution (RTD) of process fluid in an industrial-scale ethyl acetate reactor system, which consists of two independent reactors with recirculation and connected in series with each other. Bromine-82 as ammonium bromide was used as the radiotracer for the RTD experiments at different operating conditions. The individual reactors and the overall reactor system were modelled using physically representative phenomenological models comprising of continuously stirred tank reactors (CSTRs). The results showed that the recirculation rate considerably affected the flow mixing behaviour and mean residence time of the process fluid in the reactor system. The results also showed that there was bypassing of the fluid in the first reactor that ranged from 12% to 22% and 40% dead volume at different operating conditions, whereas the second reactor behaved closely as an ideal CSTR. The results of the investigation can be used to optimise the process parameters and design new improved reactor systems for the production of ethyl acetate.

Crystal Population Growth in a Continuous Helically Coiled Flow Tube Crystallizer

AbstractA helically coiled flow tube crystallizer was investigated as a novel device for the shape‐selective generation of crystals with potassium alum as the model solid. The shape evolution of the crystal population was analyzed for growth‐dominated seeded cooling crystallization. The macro‐mixing behavior was characterized by residence time distribution measurements for the fluid phase and the crystals. At the crystallizer outlet, a flow‐through microscope was used to analyze the crystal size and shape distribution, where the 3D shape of individual crystals was estimated from 2D projections. The experiments showed a separation of the particle population according to the crystal size. This process allows the controlled shaping of crystals under continuous operating conditions.

Superporous pellicular κ-Carrageenan–Nickel composite beads; morphological, physical and hydrodynamics evaluation for expanded bed adsorption application

Expanded bed adsorption (EBA) is an authentic separation method for the purification of bioproducts. Novel pellicular κ-Carrageenan–Nickel (KC–Ni) and superporous κ-Carrageenan–Nickel (SP-KC–Ni) adsorbent beads have been fabricated and evaluated using water in oil emulsification and granule leaching methods. The morphology, structure and porosity of the fabricated composite matrices were investigated. A proper spherical shape and porous structure with nano sized pores in the range of about 50–190nm has been detected for the KC–Ni beads and upper than 50% of the SP-KC–Ni superpores was at the range of 600–5400nm. The results from the image processing indicated that the KC–Ni particles had normal size distribution with the range of 60–320μm. As a result of physical properties measurements, a proper wet density at the range of 1.43–2.42g/mL was resulted for the adsorbent beads. The porosity and effective porosity of egg albumin (EA) for SP-KC–Ni3 was measured about 98.89% and 73.87% respectively. Residence time distribution (RTD) measurements were applied to investigate hydrodynamic characteristics of KC–Ni and SP-KC–Ni matrices in expanded bed. The results showed that proper physical and hydrodynamic characteristics of KC–Ni and SP-KC–Ni matrices confirm the potential of use in high speed expanded bed chromatography.

In-situ measurement of residence time distributions in a turbulent oxy-fuel gas-flame combustor

For improving the design of combustors, the knowledge of residence-time distributions (RTD) is important as they influence exhaust gas compositions. Measuring RTDs in combustors is challenging, due to high temperatures, chemical reactions, the presence of particles or corrosive species as well as high turbulence levels. This paper presents a technique for the in situ measurement of RTDs in combustors. Based on tunable diode laser absorption spectroscopy (TDLAS), the temporal evolution of the concentration of tracers is tracked simultaneously at the combustion chamber inlet and outlet. Using either air or mixtures of oxygen and carbon dioxide (oxy-fuel atmosphere) as oxidants, the method is applied to reacting and non-reacting operating conditions in a 20-kWth methane combustor. For reacting conditions, hydrogen chloride is used as a tracer, whereas for non-reacting conditions methane was chosen. Depending on the tracer, for a repetition rate of approximately 2 kHz detection limits of 16–40 ppmV are achieved. For deriving RTDs, low-pass filtering is compared to reactor network modeling. Different RTDs observed for varying operating conditions are discussed.

Evaluation and modeling of a spinning disc photoreactor for degradation of phenol: Impact of geometry modification

Abstract A new spinning disc photoreactor with two different designs employing baffles in its structure for degradation of phenol was studied and compared with conventional ones. TiO2 was successfully immobilized on stainless steel disc and Teflon baffles. X-ray diffraction, scanning electron microscopy, field emission scanning electron microscopy and transmission electron microscopy analysis were performed in order to give information on the crystal phase and structure of both TiO2 coatings and size of particles. The wettability of the photocatalyst film on the disc surface was evaluated by contact angle test. The results have demonstrated that complete phenol removal was achieved using baffles after 180 and 150 min of reaction time, which was 60 and 90 min faster than the reactor with smooth disc under optimum condition of the reactor (rotational speed = 290 rpm, flow rate = 2000 mL/min, disc diameter = 22 cm). Influence of rotational speed for smooth and baffled discs was studied. Results indicated that there was no considerable mass transfer limitation for new designs above approximately 150 rpm. For description of flow regime in the reactor residence time distribution (RTD) measurements was performed experimentally. According to RTD data, reactor was accurately modeled using tanks-in-series model. The estimated equivalent number of tanks-in-series in baffled discs were 8 and 10 compared to 4 for smooth disc. The Langmuir-Hinshelwood (L-H) kinetic model was combined with hydrodynamic model for prediction of degradation of phenol and rate coefficients. Predicted results were in agreement with experimental data.

Influence of NH3 on secondary organic aerosols from the ozonolysis and photooxidation of α-pinene in a flow reactor

This study presents detailed characterizations of a newly-developed flow reactor including (1) residence time distribution measurements, (2) relative humidity (RH) and temperature control, and (3) OH radical exposure range (i.e., atmospheric aging time). Hydroxyl (OH) radical exposures ranged from 8.20 × 1010 to 7.22 × 1011 molecules cm−3 s (0.5–4.9 d of atmospheric aging). In this study, the effects of NH3 gas on the secondary organic aerosol (SOA) formation of α-pinene by dark ozonolysis and photooxidation were investigated using the newly-developed flow reactor. For both dark α-pinene ozonolysis and photooxidation, higher SOA yields were observed in the presence of NH3 than in the absence of NH3. At RH of ∼50%, the SOA yield for ozonolysis and photooxidation in the presence of NH3 increased by 23% and 15% relative to those in the absence of NH3. Similar effects were observed at lower and higher RH conditions. Fourier transform infrared spectroscopy analysis confirmed the presence of nitrogen-containing functional groups in SOA formed in the presence of NH3. The α-pinene SOA formed in the presence of NH3 showed higher absorption and fluorescence for UV-visible radiation than those formed in the absence of NH3.

Residence time distribution measurements and modelling in industrial flotation columns

This paper presents the results of measurements and modelling the residence time distribution (RTD) in industrial flotation columns from several plants, using radioactive and salt tracers. Columns under study had different geometry (i.e. square, rectangular and circular transverse sections) and size (from ca. 2.5m3 to 180 m3). Data were obtained for liquid (both radioactive and salt tracers), floatable and non-floatable solids by size classes (radioactive tracers). Model structures including the axial dispersion, perfect mixer (PM), large and small tanks in series (LSTS) and two parallel perfect mixers were evaluated and compared. Based on the results, it is concluded that the mixing regime in flotation columns cannot be described by a unique model structure nor can they be related to the cross-sectional area or size of the columns.All industrial datasets presented in this paper showed poor fitting with the perfect mixer model. On the other hand, the LSTS as well as the two parallel perfect mixers models were outstandingly flexible compared to the other structures.The tests allowed the identification of the flow regime and the effective residence times of the different phases, which ranged from 9 to 41min. The RTD characterization has been useful to detect process failure such as bypass flow identification as well as unbalanced flow distributions in parallel columns at industrial scale.

Experimental determination of residence time distribution in continuous dry granulation

With increasing importance of continuous manufacturing, the interest in integrating dry granulation into a continuous manufacturing line is growing. Residence time distribution measurements are of importance as they provide information about duration of materials within the process. These data enable traceability and are highly beneficial for developing control strategies. A digital image analysis system was used to determine the residence time distribution of two materials with different deformation behavior (brittle, plastic) in the milling unit of dry granulation systems. A colorant was added to the material (20%w/w iron oxide), which did not affect the material properties excessively, so the milling process could be mimicked well. Experimental designs were conducted to figure out which parameters effect the mean residence time strongly. Moreover, two types of dry granulation systems were contrasted. Longer mean residence times were obtained for the oscillating mill (OM) compared to the conical mill (CM). For co-processed microcrystalline cellulose residence times of 19.8–44.4s (OM) and 11.6–29.1s (CM) were measured, mainly influenced by the specific compaction force, the mill speed and roll speed. For dibasic calcium phosphate anhydrate residence times from 17.7–46.4 (OM) and 5.4–10.2s (CM) were measured, while here the specific compaction force, the mill speed and their interactions with the roll speed had an influence on the mean residence time.

In‐line measurement of residence time distribution in melt extrusion via video analysis

Residence time distribution (RTD) is an important factor in melt extrusion as it strongly influences the product quality. It can cause degradation of the active pharmaceutical ingredient (API), which can be reduced by shorter mean residence times, and the content uniformity, which can be improved by broadening the RTD caused by axial mixing. In our study, we developed an inline video analysis that to date has mainly been applied offline. For extrusion, hydrophilic polymer and a surfactant were used as a molten matrix, in which an API was embedded. The video analysis consisted of four steps: (1) segmentation of the image to mask the strands, (2) calculation of average color values for the segmented pixels of each frame, (3) fitting of the obtained color curve with an analytical solution, and (4) calculating several RTD measures (e.g., the minimal and mean RT) for better comparison of the results. The RTD measures were responses in the design of experiments (DoE), which included the process parameters screw design, screw speed, and throughput. This analysis clearly showed that screw design had the strongest effect on the RTD measures. The results suggest that video analysis is a suitable tool for inline RTD measurements. POLYM. ENG. SCI., 58:170–179, 2018. © 2017 Society of Plastics Engineers

Study of adsorption characteristics of Au(III) onto coal particles and their application as radiotracer in a coal gasifier

This paper describes a systematic study carried out to investigate adsorption characteristics of 198Au onto the surface of coal particles using a radiotracer technique. The main objective of the study was to optimize labeling conditions and utilize the labeled coal particles as a radiotracer for tracing coal particles in fluidized bed coal gasifiers. The effect of various experimental factors such as pH, amount of adsorbent, initial gold concentration, temperature and contact time between adsorbate and adsorbent were studied to obtain optimum conditions for maximum adsorption of 198Au on coal particles from an aqueous solution. Analysis of the data showed that the Langmuir model was found most suitable to describe the adsorption phenomena. The thermodynamic analysis showed that the adsorption process is feasible, spontaneous and endothermic in nature. The results of the kinetics suggested that the adsorption presumably follows second order kinetics and chemisorption in nature. Based on the optimized conditions, the amount of coal particles and activity to be used for tracing the coal phase in pilot-scale gasifiers was estimated. Accordingly, the radiotracer was prepared and successfully used for measurement of residence time distribution (RTD) in a pilot-scale coal gasifier.

Measurement and modeling of residence time distribution of overflow ball mill in continuous closed circuit

Comparison between anticipated and measured mean residence time (MRT) along with analysis of the shape of residence time distribution (RTD) curves provides the most useful and valid information about mixing properties within a ball mill operating at industrial scale. In this work, the RTD measurements were carried out for a primary ball mill in closed circuit with hydrocyclones at two feed rates (i.e., 280 and 230 tons per hour (t/h)) by means of tracer technique. Also, perfect mixer, N-Mixer, and Weller models were studied to describe the RTD curves. It was found that the mixing condition inside the ball mill was very far from the perfect mixing circumstance. While, N-perfect mixer in series (N-Mixer model) showed better results compared to the large and small tanks in series (Weller model). According to the obtained results of the N-Mixer RTD model, when the ball mill throughput was enhanced from 230 t/h to 280 t/h, the MRT was decreased from 9.92 to 7.39 min, respectively. Moreover, relative variance of N-Mixer model diminished approximately 18% and particle cumulative passing from 75 μm of ball mill discharge increased 9%. Finally, it was concluded that operating in over-filled condition and poor mixing circumstance of slurry were the main reasons of low grinding efficiency.

Investigation of holdup and axial dispersion of liquid phase in a catalytic exchange column using radiotracer technique

Holdup and axial dispersion of liquid phase in a catalytic exchange column were investigated by measuring residence time distributions (RTD) using a radiotracer technique. RTD experiments were independently carried out with two different types of packings i.e. hydrophobic water-repellent supported platinum catalyst and a mixture (50% (v/v)) of hydrophobic catalyst and a hydrophillic wettable packing were used in the column. Mean residence times and hold-ups of the liquid phase were estimated at different operating conditions. Axial dispersion model (ADM) and axial dispersion with exchange model (ADEM) were used to simulate the measured RTD data. Both the models were found equally suitable to describe the measured data. The degree of axial mixing was estimated in terms of Peclet number (Pe) and Bodenstein number (Bo). Based on the obtained parameters of the ADM, correlations for total liquid hold-up (HT) and axial mixing in terms of Bo were proposed for design and scale up of the full-scale catalytic exchange column.

Elucidation and visualization of solid-state transformation and mixing in a pharmaceutical mini hot melt extrusion process using in-line Raman spectroscopy

Mixing of raw materials (drug+polymer) in the investigated mini pharma melt extruder is achieved by using co-rotating conical twin screws and an internal recirculation channel. In-line Raman spectroscopy was implemented in the barrels, allowing monitoring of the melt during processing. The aim of this study was twofold: to investigate (I) the influence of key process parameters (screw speed – barrel temperature) upon the product solid-state transformation during processing of a sustained release formulation in recirculation mode; (II) the influence of process parameters (screw speed – barrel temperature – recirculation time) upon mixing of a crystalline drug (tracer) in an amorphous polymer carrier by means of residence time distribution (RTD) measurements. The results indicated a faster mixing endpoint with increasing screw speed. Processing a high drug load formulation above the drug melting temperature resulted in the production of amorphous drug whereas processing below the drug melting point produced solid dispersions with partially amorphous/crystalline drug. Furthermore, increasing the screw speed resulted in lower drug crystallinity of the solid dispersion. RTD measurements elucidated the improved mixing capacity when using the recirculation channel. In-line Raman spectroscopy has shown to be an adequate PAT-tool for product solid-state monitoring and elucidation of the mixing behavior during processing in a mini extruder.

Single phase mixing in coiled tubes and coiled flow inverters in different flow regimes

The present study focuses on the measurement of residence time distribution (RTD) in coiled tubes and coiled flow inverters (CFI) over a wide range of Reynolds numbers (NRe), in contrast to earlier studies that have reported the RTD over a very narrow range, mainly restricted to the laminar region. The analysis of our experimental data reveals the existence of three distinct flow regimes in coiled structures, viz. laminar flow with prominent Dean circulation, laminar flow with less contribution of Dean circulation, and fully developed turbulent flow. Even though the turbulent transition is known to be delayed in coiled geometries, good radial mixing brought in by Dean circulation results in a narrow RTD of the exiting fluid even when the flow is clearly laminar. For coiled flow inverters, the 90° inversion allows further enhanacement in overall mixing, albeit at the cost of extra pressure drop. The effect of different design parameters on the role of secondary flow for coiled geometries of varying number of inversions (n) and curvature ratios (λ) has been investigated. Finally, a master plot is developed between the dispersion number (D/Udt) and Reynolds number (NRe) showing the marked improvements in cross-sectional mixing in coiled flow, over that in straight tubes. Based on this, a correlation is presented.

Kinetic investigations of the steam reforming of methanol over a Pt/In2O3/Al2O3 catalyst in microchannels

A kinetic study of methanol steam reforming over bimetallic Pt/In2O3/Al2O3 catalyst was carried out. The kinetic measurements were performed in a microstructured monolithic reactor with an external recycle free of temperature and concentration gradients. By the help of residence time distribution measurements it could be verified that the reactor showed the behaviour of an ideal continuous stirred tank reactor (CSTR). The absence of external and internal concentration gradients could be proven by corresponding experiments and theoretical diagnostic criteria.The kinetic measurements performed by variation of the reactant inlet partial pressures revealed that in the temperature range from 310°C to 355°C the molar rate of methanol consumption mainly depends on the methanol partial pressure, especially at higher temperatures, whereas there is only minor dependence on the water partial pressure. Carbon dioxide has no inhibiting effect, whereas hydrogen showed a weak inhibiting effect.Two power laws and three Langmuir-Hinshelwood rate equations were created for the modelling of the kinetic data. Power laws could not be fitted to the measured values. Therefore the uses of Langmuir-Hinshelwood rate laws with temperature dependent sorption constants are inevitable for the modelling. The model discrimination revealed that the rate law derived from a mechanism, which assumes the dehydrogenation of an adsorbed methoxy-species as rate determining step, described the measured kinetic data second best. Optimum agreement between observed and predicted molar rates of methanol consumption was obtained when applying a Langmuir-Hinshelwood rate law assuming dissociative methanol and molecular water adsorption on the catalyst surface. Dissociative adsorption of water and methanol at the same active site may be excluded. This leads to a better evaluation of the models that assume molecular water adsorption at the same site where methanol ties or the ones with no participation of water in the rate determining step (RDS) under discrimination.