In any major project like this, there are a number of figures. One of them is a basic estimated cost, taking into account overruns of previous projects. Bridges typically cost X, track costs Y, tunnels cost Z.
But this is always an estimate. No major survey has been done. 50% of the time the project will be under this budget and 50% it will be over. It seems less because we rarely hear of projects completed on time and under budget. It's useful to work out how much it will cost on average because in aggregate, all major projects will work out fairly close to this.
We have a second estimate. Perhaps when we do more thorough surveys, we'll find that bridges need better foundations, or tunnels are going through some particularly difficult rock. We can estimate the probability here, and come up with a higher figure. This is a useful figure because it tells us how much we might conceivably need for this particular project, and gives us a point at which we know it's time to take action.
Since this is a federal risk analysis, I presume this is going to be as pessimistic as possible. The point of these analyses is to identify where things might go wrong, after all.
So yes, it could cost that much. But is anyone - aside from the usual negative spin the media loves to put on major infrastructure projects - suggesting it actually will?