Bubba, in reply to your questions,

(1) As Stuart said above, "they estimate their EROEI is 3.5."  I checked the math, and the 8 GW powerplant (which requires 24 GW of heat input) does roughly correspond to the 1 mbd crude production figure, at this ratio. This is a great demonstration of the implications of the declining EROEI as we have produced the easy oil and turn to unconventional oil.

(2) The purpose of the freezewall or ice dam is to protect the aquifer from the heated hydrocarbons. (Yes, they do put refrigerators down holes surrounding the heating elements down other holes, which in turn surround the extraction hole. BTW, the EROEI needs to count all this drilling!) Mut was quite proud of the fact that they got out the light hydrocarbons, while leaving the heavy carbon sequestered in place. I would also hope that the aquifer and the "cooking zone" are at completely different depths.

He claimed that the freezewall was actually a tiny percentage of the overall energy budget (they leave a significant gap between it and the heated zone). I wonder about what happens to the residual hydrocarbons after they've moved on the groundwater starts to come back in (given hydrocarbon recovery rate is presumably not going to be any better than it usually is for a conventional reservoir).

Air quality issues are going to depend on how they generate all that power. In fairness, Mut didn't present the 5mbpd figure as part of his presentation. It was an off-the-cuff remark when pressed by a questioner on how likely was the 10mbpd number the questioner had heard Washington types assuming could be done from oil shale by 2025. He said he couldn't imagine the basin ever handling 10mbpd, but he thought maybe just maybe 5mbpd. I think we can probably guesstimate that this is at a very high environmental price in Colorado and Utah. Whether that would stop us from doing it is not clear...

I think what the Swede's were doing was really near the surface, whereas Shell is proposing to do this under 1000' of overburden if I recall correctly.

The fact that this is more electricity than Colarado now uses, just to get 1mbpd of fluids, is certainly sobering. Welcome to the world of EROEI = 3.5...

From my talks with HO about this issue, we're looking at a multi-step process to get from here--basically nowhere--to there, not knowing what that last step looks like in terms of production or date vis-a-vis these "oil shales" (kerogen--immature hydrocarbons--in a marl rock geological context).

The point is this: each incremental step potentially (and almost certainly) involves new technical problems to be solved and therefore, this is simply not a good energy solution in any time frame we care about at this point. By the way, when people talk about "Swedes" here, it is my understanding that they should really be talking about the Baltic states (Estonians, Latvians...). Perhaps somebody will clarify this for me but that's what my research shows.
You are so right. A technology chain is as strong as it's weakest link.
Oh yeah, you can also use windmills to produce power for this shale oil heating technology, but that's not so good because it's not dependable power. You have to drill more holes so that they don't get overloaded when the wind blows. It takes time for heat to diffuse away from the shale near the heaters to the shale away from the heaters.
Bear in mind that the electric heat is off peak power. The peaking power price is much higher than the off peak price. Essentially they get power at the cost of fuel and the coal plant capital cost is paid for by the peaking power output. This is because the shale oil deposit doesn't care whether it gets a kilowatt hour per hour 24 hours per day, or two kilowatt hours per hour twelve hours per day.
The cost comparison of peaking to non-peaking electric utilitzation for shale oil is a bit misleading in that if we double the state's electric capacity, the new capacity will be all baseload and all the capital costs will have to be captured in the price to the end-user - Shell Oil.  

That means that they should figure on $50 to $75 per megawatt-hour.

Did Shell say whether or not one could cycle the heat input?  If one does, one needs more wells and bigger heaters - I'd guess the economics point towards continous heat input.

For a small scale pilot project, there might be off-peak savings but only if you cycle your heat input.