If you want to generate a "rock" surface from depth data points, and end up with a rock surface that can be used for further volume calculations - yes, you can.
First, you have to have horizontal coordinates for the wells. create points at the well locations, with the bedrock depth as elevation - but positive (borehole at 10000,10000 with rock 8 ft down = point at 10000,10000,+8) and create a temporary surface from these points.
Second, create a volume surface using your temporary surface as base, and your existing ground surface as the comparison surface. All the values of this volume surface (and even contours if you display them) represent the interpolated bedrock elevations. In other words, it's a bedrock surface, but you can't USE it other than to display the contours.
Third, let's fix that. Set the surface style to display those contours at a fairly small interval, so you get a lot of contours. Explode the surface (twice) and you end up with contour polylines that you can use to create a new TIN surface, usable for cut/fill calculations against your grading surfaces to determine rock excavation.
Mathematically, this works, but you have to keep in mind that a great big assumption is made here - that bedrock depths make straight transitions from boring to boring. In the real world the bedrock could come nearly to the surface, or drop way down, between one boring and the next. Unless you have a pretty dense network of boreholes, I wouldn't use any calculations based on your bedrock surface without a page-long weasel clause.