Model and control strategy design of sludge combusiton in an industrial bed combustor
Résumé
The treatment and disposal of sewage sludge is an economical and environmental sensitive problem. Among different removal processes, the fluidized bed combustor process is industrially largely used for coal, wastes and sewage sludge combustion. This technology is very interesting for its fuel flexibility, its high combustion efficiency, its low nitrogen oxides (NOx) emissions and its high sulphur capture efficiency. At the same time sludge amount to be burnt increases and legislative norms on pollutant emissions are more and more drastic. Incomplete sludge combustion leads to the emission of carbon monoxide (CO). Formation of nitrogen oxides (NOx) is a highly complex and nonlinear phenomenon, mainly linked to the bed temperature of the furnace. Recent development of on-line gas analyser allows following real-time evolution of gaseous composition, and so combustion quality can be controlled. As the sludge feeding the combustor is coming from the wastewater treatment plant, variations of its composition and its inflow disturb the process. Thus the aim of this study is to propose an efficient control strategy regulating the operating conditions at the optimum point and minimizing pollutant emissions. Firstly, the sludge combustion in fluidised bed is introduced. An industrial plant which treats in a fluidized bed combustor sludge coming from an urban wastewater treatment plant is presented (figure 1). The sludge characteristic and the phenomena of sludge combustion are then presented. In a second phase, a dynamic model based on mass and energy balances is designed. In the proposed model, only major chemical reactions which are combined with chemical kinetics (taken from literature) have been worthy of attention. The fluidized bed combustor is assimilated to two perfectly mixed reactors in series. The first one is the dense bed where the fluidisation takes place with some other phenomena as devolatilization (decomposition of sludge into three phases: solid phase or char, gas phase and ash). The second one is the diluted bed where only gas combustion occurs. Thirdly, the model has been validated with data issued from the industrial plant. The model parameters have been tuned around the set point of the industrial plant. Dynamic events, as rapid variations of quality or properties of sludge supplies, are taken into account in order to validate our model. The model behaviour fit well the experimental data variations. Finally, in a third phase, this model is used for control strategy design. Simulations are carried out to determine the most efficient way for regulating the process behaviour and therefore preventing NOx formation. A simple PID control to be potentially implemented in a real industrial furnace is proposed. Evaluation is made on results accuracy and its consequences on industrial furnaces control. At last, the opportunity of proposing a more advanced control law is discussed