Parametric Model for the Optimisation of Flywheel Energy Storage Systems

An analysis on the fundamental components of a partially integrated flywheel energy storage (FES) system was conducted. These components included the flywheel and supporting shaft (rotor); supporting bearings; motor/generator unit and containment structure. A set of key system design parameters were established to which the component properties were modelled from. The analysis included three-dimensional stresses exerted on the rotor, required containment structure thickness and bearing power losses for both active magnetic and mechanical contact types. The component models were applied to the existing designs from Hebner, Beno and Walls (2002), Jordan, Herbst and Hayes (2002) and Hearn et al (2007). The results from these comparisons found energy storage capacities in the near vicinity of the expected values with design stresses within material limitations. Justifications involving inaccurate modelling assumptions, material choices and the possibility of shafts with varied diameters were pinpointed for the discrepancies. An optimisation model including both a global heuristic and derivative-dependant local scheme was then developed to determine the design parameters values which achieved an energy storage capacity at an optimised system mass or cost. Specified bounds were set specific to the required design and constraints created from the modelled components. The transit bus FES system from Jordan, Herbst and Hayes (2002) was optimised for the purpose of validation and comparison. The mass-focussed optimisation returned a carbon fibre composite flywheel paired with a steel shaft, achieving a specific energy storage capacity of 12.06 Wh/kg at a specific power output of 860 W/kg. The optimiser was found to minimize both the flywheel height and angular velocity in an attempt to reduce the mass of the greatest contributing component: the containment structure. The cost-optimised model was achieved with an aluminium alloy flywheel paired with a steel shaft and returned a specific energy storage capacity and power output pair of 5.52 Wh/kg and 393.43 W/kg respectively, placing both designs in the expected FES performance bracket specified by Briat et al (2007). The lower steel shaft cost of $1/kg vs $2/kg for the aluminium alloy flywheel resulted in the optimiser increasing the size of the shaft to within specified bounds to reduce the requirement of the more expensive flywheel material