An Advanced Well and Reservoir Model for Supercritical and Saline Geothermal Applications, the Example of IDDP-2
Benoit LAMY-CHAPPUIS, Alina YAPPAROVA, Thomas DRIESNER
[ETH Zürich, Switzerland]
The use of deep supercritical resources for geothermal energy production is desirable given the high energy density and low viscosity of supercritical fluids. However, the lack of practical experience and the usually poor characterization of such geothermal reservoirs makes predictions on the resource producibility and recharge particularly challenging. Salty fluids add a layer of complexity due the varying phase relations ranging from Vapor (V) + Halite (H) near magmatic intrusions through V + Liquid (L) + H or V + L at shallower levels to possibly L and/or V near the surface. In the reservoir, these relations will be sensitive to cold meteoric water injection and well production. Well hydrodynamics will also be affected, especially if multiple interacting feedzones exist at different depths. We upgraded to 3D an existing transport simulator (Weis, 2014) that was previously used for supercritical resource simulations (Scott et al., 2015). We coupled it to a newly developed multisegment well model. Mass, energy and momentum conservation equations for the well are coupled and solved via an iterative Newton-Raphson procedure. The phases possibly present include L + V + H + Air. The thermodynamic properties in both the reservoir and the well are evaluated with the equation of state from Driesner (2007) which is valid in wide temperature (0-1000 °C), pressure (1-5000 bars) and salinity (0 to 100 wt%) ranges. We conducted a comprehensive modelling for the IDDP-2 well in Reykjanes, a preliminary step was to reproduce the natural state of the surrounding system which necessitated the incorporation of sufficient geological constraints (permeability layering, presence of a magmatic intrusion, etc…). We let the system evolve for thousands of years from a range of hypothetical initial conditions until a good pressure and temperature match was found between the model and the field data. Using the results from this preliminary study as a starting point, we assessed various scenarios of how the well could most likely be started with airlift procedures and then utilized. The modelling results suggest possible strategies for starting the well and highlights how implementing a good approximation of the actual fluid properties is essential to understand well and reservoir behavior.
|        Topic: Software for Geothermal Applications||Paper Number: 33035|