[mchannon]

They become hamburger, basically.

Some kind of “failure bond” where the company puts up a security in case of failure might be an option.

Most cryobanks don’t seem like they’d have this problem, but when you move into the decades timescale, things like catastrophic dewar failure can drain the coolant from your samples in minutes, and if it’s not staffed 24/7 with staff trained in contingencies like this (how do you safely remove samples from the dewar fast enough and how do you even know the failure occurred fast enough?) then damage can set in. There may even not be any records made or kept of the failure.

Something to think about if you want to move forward. However, if all you want is epi-DNA, I imagine corpses a few centuries old may qualify as enough. Eggs and sperm need to stay alive and viable, which is a far higher bar.