That's an impressive piece of kit - far larger than most of us will ever get to play with.
Alternators can and do behave differently with reactive (in this case inductive) load components, with exact results varying with the design.
SO, I'd say that the "right" answer is to ask the alternator's manufacturer for specifications of its characteristics with varying amounts of inductive load.
But as a starter ...
Mechanical load wise the alternator will not mind inductive load component. VAR (Volt-Amps reactive) are "Wattless". The energy source (diesel motor?) will not know they are there overall. You will still however have to provide for the IR distribution losses for the reactive current so they will contribute some real Watts, but it will be small compared to if they had been pure-resistive-load Watts.
Inductive load component may quite possibly affect the ability of the alternator to regulate well. I assume that you are producing a 3 phase supply. If so, you'll wan't to keep the reactive component spread reasonably evenly over all 3 phases.
You'll be able to find better references than this, but a quick search turned up this Southern Illinois University Laboratory experiment instruction sheet which seems to do a good job of covering the basics. Dr Louis Youn, whose web page this is on, should be able to advise further. They note:
The output voltage of an alternator depends essentially upon the total flux in the air gap. At no load, this flux is established and determined exclusively by the dc field excitation.
Under load, however, the air gap flux is determined by the ampere-turns of the rotor and the ampere-turns of the stator. The latter may aid or oppose the MMF (magneto motive force) of the rotor depending upon the power factor of the load. Leading power factors assist the rotor, and lagging power factors oppose it.
Because the stator MMF has such an important effect upon the magnetic flux, the voltage regulation of alternators is quite poor, and the dc field current must continuously be adjusted to keep the voltage constant under variable load conditions.
Related: If one phase of a three-phase alternator is heavily loaded, its voltage will decrease due to the IR and IXL drops in the stator winding. This voltage drop cannot be compensated by modifying the dc field current because the voltages of the other two phases will also be changed. Therefore, it is essential that three-phase alternators do not have loads that are badly unbalanced.
I suggest that you try isolating the CFL bulbs (or enough of them to make a difference) in a switchable bank and observe effects on regulation (if any) with them in an out of circuit. This may work best if you test load with a limited number of incandescent bulbs and a large percentage of CFLs. Or, "best" of all, a purely CFL load. Observe carefully / no responsibility taken / Do not bend fold staple or mutilate / YMMV / Caveat Emptor ... - but it will probably be OK.
Another possible issue is CFL startup, but it is unlikely that the load on the alternator can be worse than the horrendous cold filament surge load of incandescents.
If you do find that the inductive component is causing eg regulation problems you can add capacitive reactance to balance. At the load capacity you have it may be attractive to use a 3 phase motor adjusted to present a capacitive reactive load to the system. I think these need to be wound-rotor slip ring induction motors - Mr Google is not too forthcoming on this but I know he knows. The technique is mentioned here- but using a rather large and special induction machine to do it.
I thought plasma balls were capacitively coupled and verified that "A standard plasma lamp device uses an electric current having an oscillating frequency of 35 kilohertz and a voltage ranging from 2 to 5 kilovolts"
The light emission in the globe from ionization uses different inert gases to yield different colors. The electronic ballasts in fluorescent lamp (FL) tubes require less current to ignite than older versions with heaters, due a similar high frequency trigger and then current limit once the tube resistance drops.
The current thru the globe is low but increases with capacitance coupling to the glass sphere and brightness increases with current density from higher capacitance per square cm.
Holding the glass is safe since the capacitance across the glass insulator is low. But once the gas inside has ionized the impedance to the electrode pin is lower than the impedance thru the glass and also being a conductor, the current density from contact is much higher, so a sensation is easier to detect.
The sensation of creating static electricity walking across a carpet with a door key in hand and zapping the door the knob, which has almost no sensation yet creates a strong arc. Yet touching the door knob has a much stronger sensation. The difference here is not current here but rather current density.
But this is just a static discharge where the peak current can be more than an amp in less than a nanosecond then sustained at lower level for several uS or more. In the case of the FL contact pins, the finger contact creates a high current density sensation and also since it is AC, it can be sustained at a lower level if squeezed.
So in short ( no pun) it is the capacactive coupling of the glass tube to the globe that creates the glow with the capacitive coupling of the hand. Bypassing the capacitive impedance of the hand-to-glass with a lower resistance contact to the tube pin bypasses the flow thru the body with much higher density. The body leakage capacitance to air (free space) then completes the circuit.
Overall current is limited by the flow from the plasma globe. But human sensation is a combination of current density and current level as well as pulse rate.
Best Answer
How does a lightning strike generate thunder?
The discharge into the gas pulses the atoms, while it also heats the gas. This all happens in a minute fraction of a second.
The single layer of glass can't do much to stop the vibrations this creates and as an effect you hear the glass vibrate along, giving the ticking the characteristic sound.
In flasher tubes inside big stroboscopes the tick is less flavoured, because the glass is thicker to relation, to keep from bursting at the high energy levels of the flashes.
If a starting pulse comes more as a wave, it can sometimes be that the tick is barely audible, because the last spark-over is of much lower energy. But in normal old-fashioned lamps the balast makes a sharp pulse creating a strong flash and as such a loud tick.