Reactor length lR 0.5 m

Reactor diameter dR 21 mm

Inlet  temperature T0 673.15 K

Reactor wall temperature Tw 673.15 K

λr ¼ λbed þ K1Peh

bedð  ÞDAB þ  2 ð30Þ

 uðr ¼ 0Þ

u f   ð  —  Þ ð   Þ

Inlet molar fraction of  C4H10 yC4 H10 ;0 0.02 – Inlet molar fraction of O2 yO2 ;0 0.2 – Inlet molar fraction of H2O yH2 O;0 0.03 – Gas hourly space velocity GHSV 2000 h— 1

Y.  Dong et al. / Chemical Engineering Science 142 (2016)   299–309 303

dimensional model is necessary to follow the pronounced radial temperature profile and capture the exact location and magnitude of the hot spot. The accuracy of the ‘pseudo-’ family model is heavily dependent on the reliability of the effective transport parameters especially for the radial heat transport (Dixon, 2012). A wrongly calculated temperature profile affects the concentration profile via the exponential term in the reaction kinetics. Also, the

hot spot temperature is a  critical value to check while    evaluating

compared to the center are caused by lower temperature and higher flow rate near the  wall.

The pseudo-heterogeneous model calculates the source term at each point of the reactor from the pellet mass balance. Therefore, one can plot the profiles inside the pellets at every point (r and z-coordi- nate) in the reactor. For illustration purposes, five sampling points at position 1 (r¼ 0 m, z¼ 0.1 m), position 2 (r¼ 0 m, z¼ 0.25 m), position 3 (r¼ 0 m, z¼ 0.4 m), position  4 (r ¼ 5:25 · 10— 3 m, z¼ 0.1 m) and

3  m, z¼ 0.1  m) were taken. Fig. 5 shows the

the results owing to varying pore structure parameters. A too high hot spot temperature is not only dangerous for the operation (run- away), but also has negative effect on the selectivity and life-time of the catalyst (Alonso et al.,  2001).

Fig. 4 shows the simulated concentration profiles of C4H10       and

C4H2O3 for the reference case. Concentration of C4H10 decreases while C4H2O3 increase along the axial direction with reactions proceed along the reactor. No large radial concentration gradients are observed for both species concentration profiles. The slightly higher   concentrations   of   C4H10      and   C4H2O3      near   the      wall

position 5 (r ¼ 10:5 · 10—

concentration profiles of selected species and temperature profile inside the pellet at the five sampling points in the reactor. All species participating in the reaction show intra-particle concentration gra- dients. The temperature gradients are negligible. This  behavior  can also be found in the literature (Froment, 2011). The concentration gradient indicates the presence of  diffusion–reaction  interplay  with the pore structure chosen for the reference case. By comparing the concentration plots in positions 1, 2, 3 which are at different axial positions but same radial position, one can see that the concentration profiles inside the pellet change considerably along the reactor. With