A Peaceman-type well model for the 3D Control Volume Finite Element Method and numerical simulations of supercritical geothermal resource utilization

Abstract

While supercritical geothermal resources receive increasing attention, it has so far remained unclear if and how they could be utilized. In order to provide a tool that can simulate both the natural transient evolution of a supercritical resource near a magma intrusion and its response to possible operation schemes, we augmented the CSMP++ simulation platform with a Peaceman-type well model. For the purpose of generic porous medium simulations of the supercritical resource’s response to direct production, injection or other operation schemes, only a single in-/outflow interval per well is considered and flow in the wellbore is not simulated. The model’s semi-analytical source/sink function accounts for the rapid change in pressure near the wells in the reservoir simulation. As the implementation is based on the 3D Control Volume Finite Element Method it allows the use of unstructured computational grids in order to be able to include geologically realistic geometries. We validate that the well model provides robust solutions of well pressures and rates. Such robust solutions are also obtained for supercritical conditions where water is highly compressible and, therefore, does not strictly fulfil the assumptions used in the derivation of the Peaceman model. To illustrate how a magma-related supercritical resource responds to well operation, we first simulate the evolution of a geothermal system around and above a 2 km deep, explicitly represented magma body. During the system’s hottest phase, ca. 2700 years after its initiation and during progressive inward cooling of the magma body, a well completion interval is activated at 2.1 to 2.2 km depth where temperatures then reach 420 to 490 °C. For the relatively low host rock permeability used in the model (10−15 m2), injection and production rates remain quite limited at several kg/s. Yet, the evolving patterns of temperature, enthalpy, density and flow rates provide some first insights into supercritical resource response to operation.