How to Calculate RAID 6 Usable Capacity
The formula, a worked example, and why RAID 6 has become the default choice over RAID 5 for arrays built on large modern drives.
The formula
RAID 6 stripes data across all disks with double distributed parity, tolerating any two simultaneous disk failures. Usable capacity is:
Usable Capacity = (N − 2) × Disk Size
where N is the total number of disks in the array. The minus-2 accounts for the space reserved for double parity, spread evenly across all disks rather than dedicated to two specific drives.
Worked example
Take an 8-disk array, each disk 8 TB:
(8 − 2) × 8 TB = 48 TB usable
Raw capacity is 64 TB, so storage efficiency here is 48 / 64 = 75%. Compare that to RAID 5 on the same 8 disks — (8 − 1) × 8 TB = 56 TB usable, 87.5% efficiency — RAID 6 costs you an extra 8 TB of raw capacity compared to RAID 5, in exchange for tolerating a second simultaneous disk failure.
Why RAID 6 over RAID 5 on large arrays
The core issue isn't RAID 5's math — it's rebuild time. Rebuilding a failed disk in an 8-disk RAID 5 array means reading every remaining disk in full to reconstruct the missing data. On 8 TB+ drives, that read can take the better part of a day depending on controller and workload, and the array runs with zero redundancy the entire time. If a second disk fails or even develops an unrecoverable read error during that window — a real risk on large drives with non-trivial URE (unrecoverable read error) rates — RAID 5 loses data outright. RAID 6 survives that same scenario because it can absorb a second failure during rebuild.
This is why most storage vendors and experienced admins now default to RAID 6 (or RAID 10) for any array with more than 5-6 disks, or any array built on drives 4 TB and larger, reserving RAID 5 mainly for smaller arrays or workloads where rebuild windows are short.
Minimum disks and practical sizing
RAID 6 requires a minimum of 4 disks — below that, the "two disks of parity overhead" math stops making sense (a 4-disk RAID 6 array is 50% overhead, no better than RAID 10, without RAID 10's faster rebuilds). Efficiency improves as disk count increases: a 4-disk RAID 6 array is 50% efficient, a 12-disk array is 83% efficient. There's a practical ceiling, though — very wide RAID 6 arrays (16+ disks) start to reintroduce long rebuild times again, which is where RAID 60 (striped RAID 6 sub-arrays) or object/erasure-coded storage systems take over.
Want to run your own numbers? Use the RAID Capacity Calculator — it covers RAID 0, 1, 5, 6, 10, 50 and 60 with the same math above, plus fault tolerance and minimum disk requirements for each level.