First, the average MTBF (Mean Time Between Failures) for HDs historically is
FIVE years when operated 24/7.
Second, they are electro-mechanical devices and parts do fail. Even a brand new HD
will be shipped with sectors that are flagged by the MFG as do-not-use
and when/if you surface test, you may create still more. In the early days
('80s) it was common to HDs with 100s of bad blocks while few came with ZERO
Factors that contribute to HD failure: Overheating, physical abuse
(eg dropping a laptop), forced shutdown which then need CHKDSK /F
to get running again, power outages or unstable voltages, and like people, aging.