Finally: Dyno charts of de-snorked 986S!!!

[maxster]

I think the alternator theory may be somewhat suspect; a typical alternator is rated for 70amps. This would equate to (70amps)(12volts)=840wats/746wats/HP=1.13 HP, a typical alternator is between 50% & 62% efficient, this means the mechanical HP required would be between 1.13HP/.5% = 2.26HP and 1.13HP/.62% = 1.8HP. If the car has a good battery and the lights are off the draw is most likely 50% output or less. Remember this 70amp alternator is sized to charge the battery and run with all electrical equipment on. Add the loss due to the drive belt and it normally would draw 2HP. Under the test conditions as described I don’t see how the alternator could cause more than a tenth of a HP change in load.

The test should be performed in the reverse order; snorkel off then on, to verify the results are the same.

[pecivil]

Topping the battery back up after starting would be a small load, and is essentially repeatable. The large HP drains you're quoting would only be present if you were running many peripherals such as headlights, brake lights, ventilation fans, rear window defogger, etc. As long as the test were done with the same loads on the alternator each time, the difference should be insignificant.

Not a statistician here, but this makes sense to me.