The Magmatic–Hydrothermal System Hosted in the Campi Flegrei Caldera with Emphasis on the Solfatara

The available information relevant to reconstruct the conceptual models of both the magmatic–hydrothermal system hosted in the Campi Flegrei caldera and the Solfatara magmatic-hydrothermal system was summarized and reviewed. This information includes (but is not limited to): the main characteristics of the deep geothermal wells drilled in the area, the hydrothermal alteration mineralogy and the geochemistry of the fluids encountered by the AGIP-ENEL geothermal wells, the tomographic modeling of active and passive seismic data, the chemical and isotopic characteristics of onshore thermal springs and shallow wells, onshore fumaroles and gas vents, offshore fumaroles and gas vents, and sub-lacustrine hydrothermal discharges. Taking into account the general conceptual model of volcano-hosted magmatic-hydrothermal systems (Fournier, 1999) and adopting the methodology commonly used by geothermal scientists (Cumming 2009, 2016), the conceptual model of the magmatic-hydrothermal system hosted in the Campi Flegrei caldera was reconstructed. The local stratigraphy comprises: (a) water-saturated volcanic products above 0.6 km depth; (b) volcanic and marine deposits extensively affected by propylitic, phyllitic, and argillic hydrothermal alteration, between 0.6 and 2.7 km depths; (c) a layer of thermometamorphic rocks at depths between 2.7 and 4 km approximately; (d) Mesozoic carbonate rocks (such as those cropping out around the Campanian Plain) and the crystalline bedrock from ca. 4 to ca. 7.5 km depths (e) a magma reservoir whose top is situated at ca. 7.5 km depth. Levels (c) and (d) host over-pressurized gases and hypersaline brines that are expelled from the crystallizing magma and CO-rich gases produced in-situ by different reactions. Actually, CO-rich fluids were discharged by well San Vito 1. Thus, levels (c) and (d) are the “engine” governing the bradyseism as proposed by several authors. Levels (b) and (c) are separated by a relatively impermeable zone produced by self-sealing, mainly quartz deposition, at temperature close to 400°C (Fournier 1999). Actually, this quartz-bearing zone was encountered in well San Vito 1, at depths between ca. 2.5 and 2.8 km and temperatures of ca. 360 to 385°C. This quartz-rich zone represents the boundary between the underlying deep-magmatic portion of the magmatic-hydrothermal system, where fluid pressure is controlled by the lithostatic regime, and the overlying shallow-hydrothermal portion of the magmatic-hydrothermal system, where fluid pressure is governed by the hydrostatic regime (Fournier 1999). In the shallow-hydrothermal portion, the circulation of hydrothermal fluids of marine origin (at least for the most) is restricted to the brittle rocks of the propylitic zone, where permeability is fracture-controlled, whereas the overlying impermeable, plastic rocks of the phyllitic and argillic zones act as the clay cap of the geothermal system. Similar to the basinal-type geothermal systems, the extension of the Campi Flegrei shallow geothermal reservoir is marked, to a first approximation, by the neutral chloride thermal waters emerging at the surface, or intercepted by shallow wells, close to the coastline. However, the onshore thermal manifestations (both thermal waters and gas emissions) cluster in the two distinct areas of Mofete and Solfatara-Agnano. Hydrothermal circulation might be restricted, almost completely, to these two areas, which would host two different, comparatively small hydrothermal systems. Alternatively (and this is our preferred interpretation), the two areas of Mofete and Solfatara-Agnano mark two separate upflows of a unique, relatively large hydrothermal system, extending between them, southwards (as indicated the offshore gas vents) and northwards (as proven by the deep boreholes of San Vito). Finally, the two conceptual models of the Solfatara magmatic-hydrothermal system proposed by Cioni et al. (1984) and Caliro et al. (2007) are recalled. Both conceptual models are realistic, in that they describe the system of interest during the 1982-1984 bradyseismic crisis and after its cessation, respectively.