Numerical investigations on a geothermal reservoir using fully coupled thermo-hydro-geomechanics with integrated RSM-machine learning and ARIMA models

Fully Connected Neural Network Model for Fractured Geothermal Reservoir using Supercritical-CO2 as Geofluid with THM Model

“…The fully coupled thermo-hydro-mechanical and its accuracy are validated with single fracture scenario and rock matrix without fracture. The analytical solution (44) given by the Lauwerie's [78] was used to validate the developed model .…”

“…Which includes, prediction of crude oil production, rock properties, and recognition of seismic pattern, etc. [1,40,41,42,43,44,49,50]. Neural networks are the non-traditional tactics in which they are accomplished to study system of solutions relatively than being programmed to model a specific problem in the normal way.…”

“…The application of deep learning (DL) models in the energy area rapidly increases with its predictive capability, since DL models can tackle nonlinear high-dimensional problems easily, which is much more efficient than traditional physics-based numerical simulations [30,31,32,33,34]. DL has been used in the prediction of rock properties, production performance, reservoir fluid properties, well testing applications, geological CO 2 sequestration, and production temperature from geothermal reservoirs, to name a few [31,32,33,35,36,37,38,39,40,41,42,43,44,45,46,47,48]. Recently DNN has also been applied to the prediction of the production temperature at different time nodes in geothermal reservoir modeling [1,44,49,50,51,52].…”

“…DL has been used in the prediction of rock properties, production performance, reservoir fluid properties, well testing applications, geological CO 2 sequestration, and production temperature from geothermal reservoirs, to name a few [31,32,33,35,36,37,38,39,40,41,42,43,44,45,46,47,48]. Recently DNN has also been applied to the prediction of the production temperature at different time nodes in geothermal reservoir modeling [1,44,49,50,51,52].…”

Robust optimization of geothermal recovery based on a generalized thermal decline model and deep learning

“…The geothermal recovery from HDR by EGS or HSA is a multiphysics process that couples heat transfer, fluid flow, and mechanical deformation. To simulate the geothermal recovery process, it is important to consider the variation of fluid properties, rock mechanical and thermal properties at different operational conditions [4,11,12]. Existing numerical simulation models help geothermal engineers to predict the thermal front movement and production temperature behavior both in EGS and HAS.…”