Model based prediction of the clogging of an anaerobic fixed bed reactor.

This paper presents the use of nonlinear constrained optimization techniques in order to detect and evaluate the degree of clogging in an anaerobic fixed bed reactor. First, experimental results show that the validity of a mass balance model can degrade over the time. Using the available model of the process and nonlinear constrained optimization tools, it is established that these changes can be due to the decrease of the liquid volume into the reactor while the mean values of biomass concentrations increase, leading to the clogging of the reactor. These theoretical results are confirmed experimentally in evaluating the hydraulic retention time of the reactor using a tracer.