Improved Economic Performance of Municipal Solid Waste Combustion Plants by Model Based Combustion Control

Abstract

The combustion of municipal solid waste (MSW) is used for its inertisation, reduction of its volume and the conversion of its energy content into heat and/or electricity. Operation and control of modern large scale MSW combustion (MSWC) plants is determined by economic and environmental objectives and constraints that have to be fulfilled in the presence of large unmeasured disturbances resulting from a large variation in waste composition, which make the fulfillment of these objectives and constraints very difficult, and under varying market conditions and, thereby, economic priorities. MSWC plant operators and managers are under an increasing pressure to operate economically more optimally, within the operating envelope determined by the environmental limits, due to the increasing business character of the environment they have to operate in, with market forces and competition increasingly dictating the plant operation. A direction with high potential for obtaining a more optimal economic MSWC plant performance is the application of model based combustion control, which has not been done so far at MSWC plants where PID type of combustion control is the standard. This potential is due to the fact that the combustion control system of an MSWC plant has a major influence on its overall economic performance and that model based control strategies allow for a systematic handling of the main characteristics of the MSWC plant combustion control problem, more specific the presence of constraints and the multivariable, interacting nature of the MSWC plant combustion dynamics. In particular the model based control strategy called Model Predictive Control (MPC) allows for this systematic handling and has potential for improvement. Motivated by the observed need for an improved economic MSWC plant operation and the potential of model based combustion control for achieving this goal, the main research objective addressed in this thesis is to explore the opportunities of this type of control for improving the economic performance of these plants. The main conclusion from this thesis is that the economic performance of an MSWC plant can significantly be improved by means of model based combustion control, in particular by means of MPC based combustion control. More specific, usage of the latter type of combustion control particularly allows for an improved constraint handling of controlled variables not subject to setpoint tracking. This ability can be used to significantly reduce MSWC plant downtime and maintenance costs, e.g. by maintaining furnace temperatures below a certain maximum level to increase the lifetime of furnace components. MPC also allows for a significant reduction in process variability in MSWC plant combustion control applications even when constraints do not come into play, which can be used to decrease operational and maintenance costs and to operate the MSWC plant closer to the economically optimal operating point. Additionally, the ability to include constraints in MPC together with the flexibility in formulating its optimal control problem allows for application of constraint pushing type of MSWC plant combustion control problems, which also allows for operating the MSWC plant closer to the economically optimal operating point.