Lognormally Distributed K/Th/U Concentrations ? Evidence for GeoCritical Fracture Flow, Los Azufres Geothermal Field, MX

Abstract

Well-log data for solute species K, Th and U for two wells in the Los Azufres geothermal field in central Mexico show lognormal population distribution systematics that can be understood in terms of in situ fracture-borne flow in a ?geocritical? crustal reservoir. Radionuclide solute well-log data could help build large-scale observationally-constrained models of in situ flow structures within a geocritical reservoir. Well-log, well-core, and field permeability data for crustal rock can be unified into an empirical/physical scheme in which in situ fracture systems are understood to be long-range spatiallycorrelated networks of grain-scale fractures/defects through which fluids percolate on all length and time scales. In this ?geocritical? model of in situ fluid flow, reservoir permeability is controlled by fracture-fracture interactions leading to fracture-connectivity pathways within a background of grain-scale fractures. The greater the fracture connectivity within a crustal volume the greater the permeability and the more lognormally distributed are the fracture-system flow pathway populations. It is thus plausible that significant flow structures in a geothermal reservoir appear as localized solute deposition events within an otherwise normallydistributed solute-abundance population. Lognormal distributions of K/Th/U solute abundances appear at intervals along the Los Azufres spectral gamma well-log fluctuation record, in contrast with more normally-distributed well-log fluctuations elsewhere in the K/Th/U logs and for other logging data (sonic velocity, mass density, neutron porosity). In line with generic geocritical flow phenomenology, K/Th/U lognormality in well-intervals indicates the presence of large-scale in situ flow structures which concentrate solute deposition in high-flux channels revealed by the well-log. If such flow structures are currently active, advanced reservoir modeling techniques can incorporate their flow path information into more comprehensive and reliable reservoir-scale flow models.