I'm not sold on this (deep geothermal pilots such as the Geothermal Habanero project in Australia have proved expensive busts, and the productive lifespan of a given borehole is limited to a few decades). But it's worth consideration.
I've written on Habanero previously. The project consistently overran time and cost estimates, and delivered far less power than initially planned. That's not to say that the concept is fatally flawed, but it's also not the panacea first projected. By contrast, surface geothermal fields have been and are developed at commercial scale worldwide, and have been for years: The Geysers in California, in Iceland, Japan, the Philippines, New Zealand, and elsewhere. In fact most viable fields have already been tapped, with the largely untapped resources now existing in the African Rift Valley (largely within Kenya, where it could hugely bolster the country's fairly anemic generating capacity), and of course Yellowstone in the US, where there are significant environmental and political barriers. A USGS survey getting on two decades ago of US geothermal resources conspicuously excludes Yellowstone from any consideration at all. From 2007: <https://www.usgs.gov/publications/usgs-national-geothermal-r...>
In Yellowstone, if I am not mistaken, the energy is extractable at its land surface, or quite close to it. Drilling 4 km does not seem necessary at Yellowstone. Is this incorrect?
Secondly, there exist efficiencies of scale that come with drilling at ten or a hundred sites rather than just one or a few.
Efficiencies of scale from multiple drilling operations likely pale in comparison to the costs of each well. Keep in mind that some factors, such as well casings, drilling mud, drill bits, labour, and support costs remain high on a per-well basis. A 10% savings evaporates quickly if 100 wells are required to match the energy return of 1 or 10. Well depth and diameter are major determinants of drilling costs in both petroleum and geothermal operations.
The ability to achieve a high, long-term return on relatively shallow drilling operations probably trumps any learning-curve efficiency improvements in drilling itself. Sites such as Yellowstone (based on some former research I'd made) contribute significantly to US baseload electrical generation, should the US choose to exploit them.
It's also worth noting that there already is considerable expertise in drilling generally, with over 160 years of experience over millions of individual wells, and that the efficiency / improvement curve is likely fairly ... well ... exploited.
The top of the magma chamber starts ~5-17 km from the surface with another magma chamber ~20-50 km from the surface. Heat starts and goes well above that of course but if your goal is to meaningfully hook into and extract the heat of the chamber itself it's quite deep. Put from another perspective: if a significant amount of the energy of the chamber were imminently near the surface it would already fizzle itself out over thousands of years without the need of digging short holes to do it in a few.
Geothermal in Yellowstone is no better or more useful than geothermal at many other less important places. It isn't even the place with the most surface level geothermal energy in the first place. Overall geothermal technology advancement makes a lot of sense but starting said advancement via sandbagging for an outcome 1000 years after developing one of the best national parks to do so does not make sense.
Regardless of all of that, there are significantly more than the two possibilities of either starting drilling today or having catastrophe in an eruption.
AFAIU the "last mile" (or last 5--17 km) transfer largely occurs through ground-water migration. Yellowstone combines extensive geothermal energy with ample surface water flow (e.g., Lake Yellowstone, which is itself a major geothermal zone). I'm really well beyond my depth here, though looking up soem background:
Mostly addresses seismic activity, though there's some discussion of inferred structures from that. Based on the journal article by Sin-Mei Wu, Kevin M. Ward, Jamie Farrell, Fan-Chi Lin, Marianne Karplus, Robert B. Smith. "Anatomy of Old Faithful from subsurface seismic imaging of the Yellowstone Upper Geyser Basin". Geophysical Research Letters, 2017; DOI: 10.1002/2017GL075255 <http://dx.doi.org/10.1002/2017GL075255>
(The Smithsonian piece is based on the same article.)
Exactly. I am not convinced that drilling 5 km is necessary at Yellowstone. I suspect the energy is ripe for the taking much closer to the surface due to this water migration.
Yellowstone has accessible thermal gradients. Even if a remote approach is made, say, outside the National Park boundaries and intended to minimise surface impacts, necessary drilling should be minimal, and a small number of wells should provide far more energy return than a large number of deep boreholes.
That said, I'm, ahem, well out of my depth here ;-)
Via a flagged comment, a counterargument is deep geothermal, accessed via boreholes, and not strictly limited to extant surface geothermal fields:
<https://www.quaise.energy/>
Thomas Homer-Dixon is also an advocate that I'm aware of: <https://homerdixon.com/a-big-bet-on-geothermal-could-help-pr...>
I'm not sold on this (deep geothermal pilots such as the Geothermal Habanero project in Australia have proved expensive busts, and the productive lifespan of a given borehole is limited to a few decades). But it's worth consideration.
I've written on Habanero previously. The project consistently overran time and cost estimates, and delivered far less power than initially planned. That's not to say that the concept is fatally flawed, but it's also not the panacea first projected. By contrast, surface geothermal fields have been and are developed at commercial scale worldwide, and have been for years: The Geysers in California, in Iceland, Japan, the Philippines, New Zealand, and elsewhere. In fact most viable fields have already been tapped, with the largely untapped resources now existing in the African Rift Valley (largely within Kenya, where it could hugely bolster the country's fairly anemic generating capacity), and of course Yellowstone in the US, where there are significant environmental and political barriers. A USGS survey getting on two decades ago of US geothermal resources conspicuously excludes Yellowstone from any consideration at all. From 2007: <https://www.usgs.gov/publications/usgs-national-geothermal-r...>
See:
"Habanero Geothermal Project Field Development Plan". A largely sober summary of the project, noting that it's been concluded rather than expanded. <https://arena.gov.au/knowledge-bank/habanero-geothermal-proj...>
My own 2014 summary: <https://web.archive.org/web/20230601073717/https://old.reddi...>