I'm pretty wary of proposals like this (largely because in previous designs, geothermal plants built outside on natural upwelling sources tend to be net energy losers) but let a thousand flowers bloom, etc.
Anyway, if he's right, the scale of energy available to us is simply enormous:
The figure for the whole world is on the order of 100 million exojoules or quads [a quad is one quadrillion BTUs]. This is the part that would be useable. We now use worldwide just over 400 exojoules per year. So you do the math, and you know you've got a very big source of energy.No word anywhere on what the projected costs would be, but I can't help but think these plants would have some of the cost characteristics of a nuclear plant - high initial investment, lower operating costs. This might mean the best way to build these plants would be with public financing (again, assuming it works at all.)
How much of that massive resource base could we usefully extract? Imagine that only a fraction of a percent comes out. It's still big. A tenth of a percent is 100,000 quads. You have access to a tremendous amount of stored energy. And assessment studies have shown that this is thousands of times in excess of the amount of energy we consume per-year in the country. The trick is to get it out of the ground economically and efficiently and to do it in an environmentally sustainable manner. That's what a lot of the field efforts have focused on.
Still doesn't help us with liquid fuels, though.
UPDATE: Well, looking through this PDF I'm seeing a few problems. (Note, the PDF is from 1994 and may not reflect Tester's more recent work.) The biggie is that the plant life Tester cites is only 20 years, which is a pretty short life span for a major investment like that. Secondly, the price per kwh is pretty high, even by today's standards. Thirdly, the vast majority of the "energy" released by this plant is heat, which the document assumes would be sold on to a district heating grid. The problem is that most of the world isn't Denmark, and doesn't have a well-developed cogeneration grid.
But that first problem - the 20 year lifespan - seems really problematic to me. Another article by Tester (here, PDF) states that "Typically, within a period of time less than 10 times the production period, essentially complete recovery of original temperatures will occur." So a 20-year lifespan would mean the rock the plant was drawing heat from would take "less than" 200 years to warm back up again, before we could use the area again for generation.
Maybe people will be willing to finance this kind of investment, but I'm not sure I buy it. The lifespan is short (for a generating station) and the costs are still high. If the lifespan issue weren't an issue of physics, maybe I could see a work-around, but until then color me unimpressed.