| Abstract: |
Roughly 7-10 geothermal wells are drilled in order to provide 1 well producing enough steam to drive a turbine. This well-drilling sunk cost is, however, far from unique to geothermal reservoirs. Rather it is endemic to crustal reservoirs of all fluid types: convention and non-conventional hydrocarbon, groundwater, fossil flow mineralisation, as well as geothermal. E.g., representative of the vast statistical data for onshore oil field well productivity, in the year 2009 the state of Texas had 60 wells producing an average of 645 barrels/day, 2984 wells producing an average of 88 barrels/day, and 80770 oil field wells producing an average of 1 barrel/day. As there is no intrinsic reason to consider geothermal fields significantly different from oil fields (except that the pay fluids are different), the multi-state multi-year US oil field well producer statistics are strong indicators of how rare it is to drill crustal reservoir \'producers\': for purely blind drilling, for every geothermal producer, the intrinsic odds are for 50 non-producers (with an additional 1300 wells that would never be considered for drilling). At $100/barrel value of oil, oil field well sunk costs are manageable. At $1/barrel value of hot water, geothermal drilling sunk costs requires a fundamentally different well drilling strategy. Historically geothermal wells are drilled with reference to surface manifestations and geological interpretations of subsurface structure. Given what lognormal-distribution well productivity data tell us about the strong heterogeneity of reservoir flow systems, we can assert that we need greatly improved subsurface flow structure data to guide the drill bit to achieve more cost-effective hydrothermal heat extraction. We can take the necessary step by looking below the reservoir surface via innovative use of standard multi-channel seismic sensor array technology. Unlike oil field flow systems (which are non-flowing until disturbed by reservoir operations), geothermal reservoir flow systems are naturally acoustically noisy, with the noise levels scaling in rough proportion to the volume of flow. We have demonstrated that large-scale surface seismic array monitoring of acoustic noise generated by shale body fracking practice can collect sufficient data that specialised use of otherwise standard seismic processing technology can reliably locate a large range of reservoir subsurface flow structures, most importantly the large-scale reservoir flow patterns associated with large-scale but spatially elusive fracture channels endemic to crustal rock. Thus, while the dismal oil field well productivity statistics amply reflects the deficiency of drilling without valid information about subsurface flow targets, by identifying the fundamental physical reason behind poor well productivities, and taking technical steps to acquire the necessary information, we can actively upgrade the drilling deficiency by mapping large scale reservoir fracture-flow channels in geothermal fields to guide the drill bit to the main flow structures in the reservoir fracture system. |