Dry process rotary cement kiln was studied to understand its behaviour for efficient operation and for improving control.
First, an optimal control problem in heat-up operation was formulated for minimizing the quadratic performance criterion which is a function of temperature, temperature gradient in the wall and fuel flow rate. For optimal control law computations mathematical model was simplified with assumptions and then linearized by use of orthogonal collocation in radial direction. Effects of weighting function assigned to temperature and temperature gradient and final time were investigated. Results of optimal control were compared with industrial data.
Second, a steady state behaviour of the kiln and its dynamics were investigated by the use of a Spang's partial differential equation $model^{(16)}$ with a modified model of the flame. The steady state behaviour of the kiln as exhibited by the model appears to match the behaviour of the real system except for kiln outlet gas temperature. Simulation results show that most reactions occur at the end part of the kiln.
Kiln dynamics were investigated for 10% reduction of gas flow rate, long flame and 100℃ reduction of kiln outlet gas temperature. Solid temperature and composition responses were slow as expected due to long residence time of about 4 hours. The magnitude of their response was small compared with real system due to assumptions built in the model, especially the neglect of coating effect.
