Study On Hydrodynamics And Mass Transfer In An Internal Loop Airlift Reactor

Airlift reactors have been widely applied in chemical andbiotechnological industries due to their special advantages. However, becauseof the complexity of hydrodynamics and mass transfer in the reactor, togetherwith the influences of reactor structure, operating conditions and physicalproperties, there are still many remaining problems in designing, scaleup andoperating of the reactor. The demands on airlift reactors are increasing withthe development of industrial technologies. Thus, how to design an airliftreactor with high performance and realize the operation optimization of airliftreactors are key problems needing to be solved.This work focuses on the hydrodynamics and mass transfercharacteristics in an internal loop airlift reactor. The influences of reactorstructures, such as sparger structure and sieve plate, on hydrodynamicbehaviors and mass transfer in the reactor were investigated. A macro masstransfer model of the reactor was established based on the axial dispersionmodel. The dynamic mass transfer process between gas and liquid phases wasanalyzed based on experimental data and simulation results. The pressurefluctuation phenomenon in the airlift reactor was studied and the relationshipbetween pressure fluctuations and hydrodynamic behaviors were revealed.Several new flow regime identification methods based on the pressurefluctuation signal were proposed.The research content and main conclusions of this study including:1. The influences of the sparger structure on hydrodynamics and masstransfer in the reactor were investigated. Three different spargers, namely2-orifice nozzle, rotary-cut4-orifice nozzle and O-ring distributor, were tested.It's found that under the precondition of the same gas outlet velocity of eachorifice, the sparger with the smaller orifice diameter was more efficient for themass transfer, because it produced a smaller mean bubble diameter but a higher liquid circulation velocity, and therefore a higher volumetric masstransfer coefficient. At the given airflow rate, the lager number of orificesleads to a low gas outlet velocity of each orifice, which weakens the impactbetween gas and liquid phases, resulting in the decrease of mass transferefficiency. The rotary-cut4-orifice nozzle was confirmed to be the bestsparger.2. The flow regime transitions have a large influence on hydrodynamicbehaviors. In the homogeneous flow, the overall gas holdup and thedowncomer liquid velocity increased with the superficial gas velocity, whilethe downcomer liquid velocity tended to be stable and the increase of gasholdup slowed down in the heterogeneous flow. The volumetric mass transfercoefficient almost increased linearly with the superficial gas velocity. Thesuperficial gas velocity strongly affected the specific interfacial area, whilehad less effect on the liquid-side mass transfer coefficient. Empiricalcorrelations were proposed to predict the gas holdup and the volumetric masstransfer coefficient for different spargers in different flow regimes. Based onthe material conservation and pressure balance principles, a fluid model wasestablished to predict the liquid velocity in the downcomer, and itseffectiveness was validated by experiment data.3. Effects of the sieve plate on flow and mass transfer characteristics inthe reactor were studied. It's found that the sieve plate played an importantrole in breaking up bubbles, and thus made the flow field more uniform andenhanced the mass transfer efficiency. The installing of sieve plates in theriser could significantly enhance gas holdup and volumetric mass transfercoefficient. The sieve pore diameter and the free area ratio are importantdesign parameters of the sieve plate. Smaller sieve pore diameter and freearea ratio were more benefit to increase the gas holdup, while larger sievepore diameter and free area ratio were more efficient in enhancing the masstransfer efficiency. Moreover, the number and mounting position of the sieveplate also have strong effect on the performance of the reactor. The resultsindicated that the mass transfer coefficient was larger when more sieve plates were used. The installing of the sieve plate at the bottom section of the riserwas found to be more efficient. Empirical correlations were proposed topredict gas holdup and volumetric mass transfer coefficient for different sievestructure parameters. A fluid model was established to predict the downcomerliquid velocity in the airlift reactor with sieve plates.4. A macro mass transfer model of the airlift reactor was established byapplying the axial dispersion model to gas and liquid phases. The partialdifferential equations of mass transfer model were transformed to differenceequations using the finite difference method, and a numerical method wasdeveloped to solve the mathematical model. Experiments were carried out tovalidate the effectiveness of the model. Based on experiment data andnumerical simulation results, the dynamic mass transfer process in the reactorwas illustrated in details. Based on the conclusion obtained from numericalresults, the mass transfer model was further simplified and the analyticalsolution of the simplified model was obtained by using a coordinate transformmethod. The results indicated that the simplified model could well predict thedissolved oxygen concentration when axial dispersion coefficients and axialconcentration gradients in the reactor were negligible.5. The origin of pressure fluctuations in airlift reactors was discussed.The pressure fluctuations in the reactor were divided into two categories:global pressure fluctuations and local pressure fluctuations. Based on thecoherence analysis, the pressure signal was decomposed into three differentparts: coherent part, joint incoherent part and exclusive incoherent part. Eachpart was related to pressure fluctuations from different sources. Therelationship between evolution of energy ratios of these three parts with thesuperficial gas velocity and flow regime transitions in the reactor wasinvestigated. The results indicated that the pressure signal was capable torepresent flow regime transitions. The power spectral and coherence analysismethods were applied to study the relationship between pressure fluctuationsand hydrodynamic behaviors in different flow regimes.6. The flow regime identification based on pressure fluctuation signals was deeply investigated and several new flow regime identification methodswere proposed. The wavelet transform, Hilbert-Huang transform, higher orderstatistics, Wigner-Ville distribution, chaos and fractal theories were applied toanalyze pressure fluctuation signals, and thus the flow regime characteristicswere extracted from the pressure signal. By using the wavelet entropy,average bispectrum, generalized average frequency, Hurst exponent andchaotic parameters as characteristic quantities, the three typical flow regimesin the reactor were successfully identified. These results provide a newthought for the flow regime identification based on pressure fluctuationsignals.