| Title | Is the Deep Geothermal Heat Source on Plate Boundaries Affected by Long-Term Deformational Heat Production and Advection? |
|---|---|
| Authors | Vikram SINGH, Ben HOLTZMAN |
| Year | 2020 |
| Conference | World Geothermal Congress |
| Keywords | transform plate boundary, deformational viscous dissipation, seismic dissipation |
| Abstract | Most models of geotherms on transform plate boundaries assume simple conductive cooling from a mantle adiabat that begins at a fixed depth (a plate model), with possible heat production by fault slip, such as the Lachenbruch and Sass (1980) model applied to the Coast Ranges of California. However, such models may be incomplete and the geothermal potential may be larger than assumed if plate boundaries have both siginificant viscous dissipation in the strong but aseismic part of the plate boundary (lower crust and upper mantle) in addition to seismic dissipation. Heat produced in plate boundaries may be transferred upwards via conduction and also advection by fluid flow. We consider transform motion of two plates on a boundary of specified width (~10-100 km) driving visco-elastic-plastic deformation in ratios governed by the geotherm and the laboratory derived flow laws, along with hypothetical temperature- and pressure-dependent yield stress laws. Viscous deformation is converted into heat continuously. We consider simplified seismic dissipation as a continuous and distributed process over time scales longer than many seismic cycles (single time steps of ~ 1e3 yrs), such that the yield stress is converted to heat where the plate boundary reaches its yield stress. Three observables from the San Andreas plate boundary are compared with the results from the model: average heat flux measured near the surface, estimates of seismic dissipation calculated from average seismicity rates across the entire Coast Ranges, and seismic depth of the brittle-ductile transition estimated roughly from observed earthquake depths. Results indicate that both seismic and viscous dissipation contribute to the heat flux and their ratio significantly depends on the yield stress function used. |