Lester, you are correct in your understanding of the insitu process. Right now I'm working on connecting electric submersible pumps for multiple wellpads at the Long Lake site (Opti/Nexen). This would be the production side of the process. Each wellpad has up to 13 producing wellheads. The pumps suck the bitumen out of the ground after the steam process liquifies the molasses-like bitumen into a manageable viscosity. The sand sinks to the bottom of the natural reservoir leaving a cleaner product to be pumped out of the ground. This site is operating at approximately 75% capacity and very clean. The heat from the steam used to liquify the bitumen is reused back at the co-generation plant to help power the site.
One of the disadvantages of the steam injection process of insitu is the vast amounts of water required to carry out the procedure. It's taxing the local water sources in the area. I've heard talk of building a pipeline to bring water in from other sources but we'll see how far that idea goes.
As far as what is happening underground when the steam is injected, it makes sense if a thick, syrupy material is suddenly liquified it is going to migrate through cracks and fissures into areas it wouldn't normally go. And that could mean straight into wellwater, lakes and rivers.
As far as the open pit sites go, I've been to the Syncrude site and those tarponds are a disturbing sight to behold. The oily surface, the scarecrows, the cannons all combine to create a surreal scene. Kinda reminds me of those space movies where a distant planet is being mined and nobody lives there for any length of time, except for the grunts who are condemned there for life. You can't wait to jump in your spaceship and head back to earth.