| Abstract |
As recognized in the 1970s, extracting heat from the crust is limited by slow thermal conduction rates. Heat mining was thus predicated on engineering large volumes of permeable rock within hot ambient crust – primarily by hydraulic fracture stimulation. However, engineering such volumes has proved elusive. We discuss how this difficulty arises. We use data from the FORGE site to illustrate our discussion. Following oil/gas practice, crustal fluid flow properties have been taken to conform to a statistical ideal in which porosity and permeability variations are assumed to be uncorrelated (random) above a limited scale length. In this model, the crust is effectively a uniform elastic continuum in which engineered fractures are oriented along the local stress field. Above this scale porosity and permeability would be statistically unrelated and uniformly distributed – uncorrelated white noise. Instead, well-log, well-core, and well-flow data show that porosity and permeability are related and spatially correlated at all scales: permeability k(x,y,z) is related to porosity φ(x,y,z) by k=koexp(αφ). If φ’s population and spatial distributions are normal and pink instead of white noise (i.e. φ (s) = 1/s, s = spatial frequency), then κ’s population and spatial distributions are lognormal and pink. Consequently, while φ centers around an average, k has a few large values whose locations dominate the permeability field. This poro-perm relationship is found in all types of brittle rock. Permeability stimulation induced microearthquakes (iMeqs), including those at FORGE, follow these relationships. FORGE iMeqs are pairwise statistically correlated as G(r) ~ 1/r, r = the distance between iMeq pairs. This matches κ’s distributions for the case of a normal population and pink distributed φ and implies the iMEQs are mapping the permeability structures as well. Since k is lognormally distributed, there will be a few locations with highly elevated permeability. One such permeability structure is reported for the FORGE site stimulation. A location with steady 30L/s flow at 185C has been recorded for 30 days across a 100 m separated doublet. If this location were a simple, isolated fracture between the wells, it would likely show noticeable cooling in 100 days. Instead, it appears that a more complex, larger end member of the k=ko exp(αφ) distribution of permeable rock was opened for advective heat transfer from a much larger volume. |