Manufacturing Test Sample Size??

Good afternoon.
Not having a statistical background I am trying to figure out how many samples of a manufacturing population set I need to test to convince myself that the population size will successfully work in the field. I will be doing accelerated temperature testing on a small set of the population size. I suspect this will jeopardize this set so I will most likely not replace this subset back into the population. My population size is 250 units every quarter of a year. I would like to have 1% or less failures in the set. Can someone talk me through the math?

Thx