From a recent training session, here is a snippet. Capacity is reserved, not drives. The number of drive failures would also be influenced by the number of Parity drives in addition of "preservation space" to evaculate from failed drives in the pool.
"DEFAULT PRESERVATION CAPACITY
When dynamic disk pools are created, a certain amount of capacity is preserved for emergency use. This capacity is expressed in terms of a number ofdisks in the management software, but the actual implementation is spread across the entire pool of disks. The default amount of capacity that is preserved is based on the number of disks in the pool.
After the pool has been created, you can increase or decreasethe preservation capacity, or even eliminatepreservation capacity (0 disks’ worth). Themaximum amount of capacity that can be preserved (expressed as a number ofdisks) is 10, but the capacity that is available might be less, based on the total number of disks in the pool.
Like hot spares, the preservation capacity sits idle until it is needed. However, because the disks are all used for I/O for the volume (or volumes) in the disk pool, you know that the preservation capacity is active and available. In contrast, a hot spare mightfail between the time that it is configured and the time that it is needed as a spare for a disk in a volume group."
Also, larger pools will have larger preservation capacity equivelent to more drives. That default can be decreased.
Good news is that this spare capacity means that more HDDs are in use unlike spare volumes, and rebuild times are much more likely to much faster.