Distributed parameter modeling and optimal control of the oxidation rate in the iron removal process

Generally, the iron removal process is modelled as a lumped parameter system that does not provide information about the distribution of reactants in the steady state. In this paper, we investigate the distributed parameter model and control for the iron removal process. By analyzing the process properties, we study the mass balance over a differential volume element, and the spatiotemporal distributions of the Fe2+, Fe3+ and H+ concentrations are derived by partial differential equations. An optimization problem is constructed to estimate the unknown parameters. Then, an optimal control problem for the oxidation rate of the ferrous ions in the steady state is proposed to achieve process requirements that have the lowest cost of oxygen and zinc oxide and obtain high goethite quality. To eliminate the impact from inevitable disturbances, an expert-based correction mechanism is constructed to compensate for the optimal control when the outlet ferrous ion concentrations are out of the desired range. Finally, the simulation results demonstrate the good performance of distributed parameter model. Industrial experiments demonstrate the satisfactory control performance of the optimal control strategy. Regarding manual operation and PI control, the control strategy increased the qualified ratio of the #4 reactor outlet Fe2+ concentrations by 8.4% and 3.4%, respectively. Additionally, on average 17760 m3 of oxygen and 109.68 t of zinc oxide per month were saved compared to manual operation. The mass percent of iron in the goethite increased from 34.31% (manual operation) and 35.12% (PI control) to 35.83%.