TBM's can fall into a pit of despair. Seattle had a problem with a massive unexpected steel bar down in the pathway. Snowy 2.0 had a problem with grade, with soil and rock density, with water incursion. For Snowy it's beyond 2x cost overrun.

These problems can add years, and in construction, time is money: the cost could balloon. Hard to imagine $16B getting worse but it's not impossible. There's no "reverse gear" on a TBM because the liner it extrudes behind it precludes backing it out. Maybe you get a couple of meters of space in the header to work with. They have hatches so you can clean the teeth on the rotating cutter thingamajig.

Not that this isn't super fun, and interesting, and worth doing, but at this stage in the game, its very early. Across the life of this project a lot of awkward things can happen.