Reservoir and costs in one loop
Diego Valdés Larraín, Ignasi Herms and Enrique Gómez Rivas have proposed a way to join reservoir simulation with techno-economic evaluation automatically. Their starting point is the gap between modelling reservoir physics and counting costs: the tools that judge profitability are quick, but they rest on simplified one-dimensional or semi-analytical models that assume homogeneous, isotropic media. As a result they do not capture three-dimensional reservoir heterogeneity, permeability anisotropy or the real geometry of the wells.
What the method involves
The authors built a Python-driven workflow that links a three-dimensional thermo-hydraulic simulation in the MRST environment with the techno-economic assessment tool GEOPHIRES-X. It runs in batch mode and replaces the simplified reservoir estimates built into GEOPHIRES-X with data from the MRST simulation: it passes across the computed thermal vectors and geometry-corrected capital expenditure. Both tools are open source, and the whole loop is meant to be reproducible and auditable.
The Catalan case
The method is shown on a hypothetical doublet made of two J-type directional wells aimed at karstified Jurassic dolostone lying in the Camp Basin in Catalonia. Accounting for the three-dimensional pattern of vertical anisotropy and for buoyancy effects brought a 39.3-percent drop in the thermal power that could be delivered, compared with the simplified model. The fully coupled model cut the net present value by 91.2 percent and raised the levelised cost of heat by 71.9 percent, to 0.1141 euro per kWh of heat. A sensitivity analysis showed that in the study area the economic outcome depends more on surface operating variables than on the geological reservoir parameters considered.
What it means
The authors stress that models which ignore three-dimensional reservoir physics can paint too rosy a picture of a project's viability and understate the levelised cost of heat. The work is available as a preprint and has not yet been peer reviewed.