Control of the Aero-Electric Power Station - An exciting QFT application for the 21st century

The Aero-Electric Power Station is the ultimate solar power station, utilizing the dry, hot air of Earth's desert zones. By spraying water at the top of e.g. a 1200 m tall chimney with a diameter of 400 m, the air is cooled by evaporation and flows downwards through turbines at the bottom, generating 380 MW of net electric power. The Aero-Electric Power Station is still in the planning stage, and this paper belongs to a long series of feasibility studies. The current 'truth' model of the Aero-Electric Power Station is a one-dimensional partial differential equation model. The external slowly changing weather, defined as the mean air pressures, temperatures and humidity at the top and bottom of the tower, determines the optimal operating point, i.e. the optimal water spray flow and turbine velocity that give the largest net power. The gross power produced by the turbine is partly delivered to the grid and partly to pump sea water to spray water reservoirs. The reservoirs make it possible to use the pumping power and the spray flow rate as control. Wind changes cause significant deviations from the mean external air pressures, requiring closed loop regulation to keep the rotor velocity constant. The Aero-Electric Power Station may be modelled as an uncertain, unstable irrational transfer function, with two disturbances (external air pressure deviations at top and bottom), two control variables (turbine power and spray flow), and one output (rotor velocity), without a cascaded structure, giving rise to a robust load sharing control problem. A robust linear feedback regulator is designed by QFT, in such a way that the load of regulation is shared between the two control inputs. A closed loop step response simulation for one operating condition, using the 'truth' model, demonstrates the design.