I'm by no means a flywheel expert, but I know there would an ultimate limit due to the strength of the flywheel material.  As the wheel gets bigger the velocity of the outside edge increases, eventually any material would fail.  Practically this may not mean anything, a materials expert is required there.

You would also need to worry about windage and bearing friction as your flywheel got bigger.

As for the motor, with a VSD I can't see any problem as long as the previous are considered.

When you start an induction motor, there is a lot of energy dissipated in the rotor of the motor. The energy dissipated is proportional to the full speed kinetic energy of the load at full speed. In other words, as you increase the inertia seen by the motor, the rotor heating is increased. Motors are rated in the "maximum load inertia" or the maximum DOL (start)time. Both of these are indicative of the thermal inertia of the rotor and ther is a very large variation in this characteristic of motors.
Using a VFD results in the motor never operating under high slip conditions and so the thermal mass of the rotor becomes irelevant. Check with the VFD manufacturer to ensure that the VFD is capable of controlling the heavy load.

Mark Empson
http://www.lmphotonics.com

I agree with gordonl. I think from an electrical point a view there would not be any major problems rotating a large flywheel. As gordonl advised maybe you should post your question in the materials engineering forum of this site.

Good answers all around. It's an interesting subject and I'd like to ramble a little bit...

For starting of an unloaded induction motor DOL, the total rotor heating energy is not only proportional to the kinetic energy of the rotor, it is exactly equal to the final kinetic energy of the rotor.  There's no obvious physical reason for this, but can be shown mathematicaly.

If you add a load to the motor, then this heating is increased by a quantity which resembles the average value of Te/(Te-Tm).  This factor can get quite large if Tm (load torque) approaches Te (motor torque) for the same speed.  A very interesting result is that the total rotor heating energy will increase dramatically if you use a reduced-voltage start.  This runs a little bit against our intuition which suggests that a reduced voltage start is less stressful for the motor. Reduced voltage starting actually generates more heating in both the rotor and stator, although the heating is spread out over longer period of time.

You're right that VSD applied properly will solve all of these problems.

Suggestion: It appears that a more careful electromechanical system approach, e.g. by using a suitable simulation software package, will be needed since the large flywheel will introduce the longer motor acceleration time, longer-lasting large motor starting current, vibration, consume some power from 500HP, coupling and bearing difficulties, etc.
Please, what is this actually needed for?

As gordonl touched on, important issues with large inertia are mechanical and materials. Starting by vfd is straight forward. First level software simulations are simple today -I personally have 8 different tools and the best one (brain) also.