Residence time distribution studies in a reactor with recycle

Abstract

Residence time distribution (RTD) studies are of immense importance in process industries as they provide valuable information regarding the flow behavior in process equipment and can help with their real time diagnosis. In the present work, two sets of RTD studies were carried out in an industrial-scale ethyl acetate reactor system consisting of two reactors in series with a large recycle ratio and recirculation as a mean of external mixing. 82Br as ammonium bromide was used as the radiotracer for the RTD experiments. The individual reactors and the reactor system were modeled using basic RTD building blocks, like, continuously stirred tank reactor (CSTR) and plug flow reactor (PFR) to map the experimental RTD curves. RTD experiments were also performed on a laboratory-scale reactor system to study the effect of recirculation and recycle on the flow behavior. In the first set of industrial RTD studies, the results showed that the recirculation rate had a significant effect on the flow mixing behavior and mean residence time (MRT) in the reactor system. The experimental RTD curves showed that there was bypassing (12% - 22%) of the fluid in the first reactor at different operating conditions. MRT of the reactor system 1 (comprising of reactor R1 and reactor R2), decreased from 17 h to 10 h with decrease in recirculation flow. A stagnant volume of 40% inside the first reactor, exchanging fluid with the active volume, gave the best fit between experimental and model predicted RTD curves. The second reactor, however, behaved very closely to a CSTR at different operating conditions. Before the second set of RTD experiments was conducted, the reactor system was modified by the industry to accommodate any future increase in the production capacity of the plant. In the new reactors system (reactor system 2), reactor R1 had a horizontal orientation and reactor R2 had vertical orientation. Flow rates, working volume and other process parameters were kept the same as for the reactor system 1. The second set of industrial experiments was performed on the new reactor system (reactor system 2). The results showed the presence of bypassing (18% - 32%) of process fluid in the first reactor, and bypass of 2% - 15% was observed in the second reactor, at different operating conditions. MRT for the reactor system 2, decreased from 26 h to 18 h when the recirculation to both the reactors decreased. The model results predicted the presence of stagnant volume (with fluid exchange) in both the reactors corresponding to the best fit between the experimental and model predicted RTD curves. In the first reactor, 5%-15%, and in the second reactor, 5%, stagnant volumes (with fluid exchange) were predicted corresponding to the best fit between the experimental and model predicted RTD data at different operating conditions. In the laboratory-scale reactor system, RTD experiments were performed with varying recycle and recirculation rates to see their effect on mean residence time (MRT), flow bypassing and stagnant volume in the reactor. A computer program was developed to solve the model equations using fourth-order Runge-Kutta method. A low bypass flow (<5%) was observed from the experimental RTD curves obtained at different operating conditions. A change in the MRT from 1.2 h to 1.8 h was observed at different recycle and recirculation rates. At maximum recycle and maximum recirculation, in the study ranges, a 37% stagnant volume (with exchange) was predicted. In the absence of recycle and recirculation, a 53% stagnant volume (with exchange) was predicted corresponding to the best fit of the experimental RTD data. The results of the laboratory-scale RTD studies were also compared qualitatively with the results obtained from industrial RTD experiments. Fig. 1 shows the schematic of the methodology followed for overall thesis.