Linear and Nonlinear Control of an Upflow Anaerobic Sludge Blanket Reactor: Effects of IMC-PID Tuning and Uncertainty

Abstract

Upflow anaerobic sludge blanket (UASB) reactors are often used for wastewater treatment and microalgae production. UASB have a greater performance due to higher biomass concentrations. However, this comes at the cost of increased nonlinearities and therefore increased difficulties of control through linear methods (especially ones that make use of simplified models of the system). The methods developed to counter this type of problems include feedback linearization, which uses additional information on the system to create performance advantages. This paper performs a comparative study of both types of controllers applied to a UASB, first in their equilibrium points and their stability, secondly in the different stable regions (basins of attraction) of these points in the parameter space, and thirdly in the effects of identification uncertainty on over these regions. It is found that the two control loops have similar equilibrium points qualitatively and quantitatively. The kinetics of substrate inhibition limit the values of controller set point in the linear loop, and the uncertainty of identification gain has greater importance to that of the time constant.