RAID 5 provides fault tolerance by distributing parity information across all drives, allowing the array to continue operating if one drive fails. In a three-drive RAID 5 set, a single disk failure puts the array into a degraded state where it can still function, but it is at higher risk because losing a second drive before recovery would cause data loss. Since the system reports Drive 3 has failed after a power outage, the best next step to minimize risk is to replace the failed drive immediately and start the rebuild. Rebuilding regenerates the missing data blocks onto the new drive using parity, returning the array to a protected state.

Option B leaves the array degraded and increases exposure to a second failure. Option C (firmware upgrade) is not a first response in an active failure scenario and can introduce additional risk or downtime. Option D (restore from backup) is typically used when the array is unrecoverable or multiple drives fail; it is more disruptive and unnecessary when RAID 5 can recover from a single-disk failure. Therefore, replacing the defective drive and triggering an array rebuild is the correct next step.