| Abstract |
We present a beginning-to-end case history of PFI mapping of potential drilling targets at Newberry Volcano, Oregon, in 2023. The PFI method uses hundreds of time-synced seismic recorders to observe episodes of small seismic movements in connected permeability structures – the fluid filled joints, fractures, fracture zones, faults, and fault zone that are geothermal drilling targets. PFI signals are initiated by local changes in crustal stress. Such perturbations can be caused by near and distant earthquakes, industrial activities, rapid moving air masses, tidal loading, fault creep, and the like. These seconds-to-hours long disturbances superimpose stress changes on the longer-term tectonic stresses. The resulting PFI signals can continue for tens of seconds to minutes as the permeability structures readjust to the new stress field. Our 2023 Newberry survey was designed to locate these signals for potential drilling targets. It was supported by ARPA-E grant AR000166 with supplementary funds from Quaise Energy. Early in 2023 we modeled the location and resolution tradeoffs for several distributions of seismic velocity, PFI signal depths, recorder spacings and network size. Based on them we deployed 982 can-sized seismic recorders, nominally 170 m apart, over a 27 km2 area west of Lake Paulina. This network captured PFI-signals above a 2.2 km deep, 2.72 km2 patch centered on the 3 km deep 55-29 well. The recorders were deployed by a crew of 12 over 6.5-days in Oct 2023 with station accuracy of 2 m. The completed network ran for 5 days. On day 4, a 3-hour long, low pressure hydraulic stimulation was done in 55-29. Retrieval took 5 days, just days before a 6†snow would have hidden their exact locations. The initial phase of data download, quiet time selection, trace editing, and noise suppression was completed in Nov. PFI signal locations were then found by first searching for high semblance episodes in the recordings. Using the velocity model, these episodes were then traced back to their originating 10x10x10 m voxel in a 165x165x185 volume of 5,036,625 voxels. In late Jan, maps and cross sections through this PFI signal volume were made using color bar, isosurface, and maximum semblance track methods and used for geological interpretation. The PFI maps and sections were compared to geological ones, 55-29 well logs, and a magnetotelluric resistivity profile. The PFI maps have trends matching known NNW striking, east-dipping normal faults. Near-well PFI horizons correlate with the well logs and lithology. A low resistivity body in the MT profile correlates with an absence of PFI signals, interpreted as a clay cap with limited permeability. Three other high semblance targets were also observed beyond 55-29. From funding to establishing potential drilling targets, the Newberry PFI study took ~1-year, the bulk of which was for permitting. Points along the project’s critical path included permitting, establishing a velocity model, testing tradeoffs of PFI resolution, depth, recorder numbers and spacing, and processing. With a recent, follow-on ARPA-E award we are planning a second, deeper PFI study at Newberry for 2025. |