Earth Model
Add text (From Rob)
Additional images???
Native State Modeling
A numerical implementation of the conceptual and earth models was developed of the FORGE reservoir and surrounding area to estimate the spatial distribution of native state pressure, temperature, and stress conditions. The primary goal of this effort was threefold.

 Incorporate detailed 3D parameter distributions and complex boundary conditions identified from characterizing the site.
 Better understand the spatial distribution of stress at the site and how it may influence reservoir stimulation.
 Establish a reference (or baseline) set of parameter and property distributions that can be used among the team (and modeling community at large) to ensure consistency and comparability of simulation results.
Model Location and Dimensions
The FORGE numerical model domain was sized to accommodate the geothermal reservoir intersected by Well 5832 and future injection and production wells along with their predicted stimulation volumes created during FORGE Phase 3. A region box 2.5 km x 2.5 km x 2.75 km, located approximately between depths of 400 m to 3200 m below the surface aligned with the calculated principle stress direction The depth to the top of the model domain was chosen so that the entire top of the model consisted for the alluvium materials. This was chosen for two reasons, the first being to be deep as possible in the alluvium to avoid potential shallow lateral flow originating from the east side of the Opal Mound Structure on the southeast side of the domain and the second to facilitate calibration of the vertical stress in the model. A uniform mesh spacing of 50 m was used, which resulted in a total of 137,500 grid cells. 

The lithology was divided into the two broadly defined units from the conceptual model, consisting of granitic basement rocks (granitoid) and the overlying basin fill sedimentary deposits. The model domain was constructed in earth model (LeapFrog Geothermal) so that property distribution estimates from the FORGE characterization and historical studies could be directly extrapolated onto the numerical grid. The resultant numerical domain and mesh used the same reference Universal Transverse Mercator (UTM) coordinate system (global coordinate system or “global model”), which allows for importing of numerical results back into the earth model for display, comparison, and archiving. 
Boundary and Initial Conditions
Boundary conditions for pressure, temperature, and stress were based off characterization results performed during Phase 2A and B, which relied on both new data collection and information obtained from the literature. Specific boundary conditions and the values used will be discussed in detail in the following sections. A “local” model domain was also created. This was accomplished by developing a local coordinate system (which has a base point equal to the bottom left corner of the global model mesh) and rotating the mesh by 25 degrees counterclockwise. This was done to align the mesh with the principle stress direction, thereby allowing direct assignment of the estimated stresses from Phase 2B testing in Well 5832. Due to the complex nature of the distributions of pressure, temperature, and stress, all boundary conditions in the native state model have varying degrees of spatial variability, which was implemented using Dirichlet conditions. Values for Shmax were obtained from stress estimates determined during Phase 2B. A stress gradient of 0.77 psi/ft (0.0174 MPa/m) was applied to the back side of the model domain, while the front was fixed at zero displacement in the Shmax direction. Values ranged from approximately 11 MPa to nearly 55 MPa. 

Initial pressure, temperature, and stress conditions used for the native state model were based on estimates in the earth model and assigned to the numerical model cells in a similar fashion used to assign the boundary conditions.Values for pressure were estimated based on the top boundary condition and interpolated downward as a function of depth. Temperate was directly interpolated from the earth model to the numerical model cells. In FALCON, stress is a derived quantity based on calculations of the displacement of the rock matrix, and as such, is difficult to assign a priori. A value of zero displacement was assigned as an initial condition, letting the model iterate a few extra times to come to a converged solution.Initial reservoir properties used in the native state model were taken directly from characterization data. In many cases, a range of possible values were available, and the mean or median was used, with the values being adjusted within the measured range during model calibration. In all cases, uniform reservoir properties were used within the alluvium. For the granite heterogeneous property distributions were used where appropriate and data were available. Tables below summarize the property values used. These are the “reference” values for the current state of FORGE. 
Modeled Granitoid Parameters
Parameter  Units  Min  Max  Source/Comment  Link to data 
Compressibility  1/kPA  2.52E12  8.51E08  Upscaled DFN  
Kii  m2  1.75E21  1.20E16  Core and reservoir testing, upscaled DFN  1161 
Kjj  m2  2.44E21  1.28E16  Core and reservoir testing, upscaled DFN  1161 
Kkk  m2  2.93E21  1.10E16  Core and reservoir testing, upscaled DFN  1161 
Porosity  —  1.00E07  0.0118  Core and cuttings analysis, upscaled DFN  
Rock grain density  kg/m3  2750.00  Core and cuttings analysis, native state calibration  1052  
Specific heat capacity  J/kg K  790.00  Literature  
Grain thermal conductivity  W/m K  3.05  Core and cuttings analysis, native state calibration  5832 thermal conductivity data  
Young’s Modulus  Pa  5.50E+10  6.20E+10  Core analysis  1140 
Drained Poisson’s Ratio  —  0.26  0.3  Core analysis  1140 
Undrained Poisson’s Ratio  —  0.35  0.4  Assume B=0.8  
Biot coef  —  0.5  0.7  Literature  
Thermal expansion coef  —  2.00E06  Literature  
Fracture asperity (Lognormal)  m  µ=1e4, s=1e8  Literature  
Cohesion (Lognormal)  MPa  µ=3, s=0.5ˑ10^6  Literature  
Frictional angle (Lognormal)  MPa  µ=0.6, s=0.001  Literature  
Mode 1 fracture toughness  —  2.48MP√M  Core analysis  1140 
Modeled Basin Fill Parameters
Parameter  Units  Value  Source/Comment  Link to data 
K  m2  1.70E14  Aquifer test  1140 
Porosity  —  0.12  Core cuttings analysis, native state calibration  1052 
Rock grain density  kg/m3  2500.00  Core cuttings analysis, native state calibration  1052 
Specific heat capacity  J/kg K  830.00  Literature  
Grain thermal conductivity  W/m K  2.00  Core cuttings analysis, native state calibration  5832_thermal conductivity data 
Young’s Modulus  Pa  3.00E10  Literature  1140 
Drained Poisson’s Ratio  —  0.30  Literature  1140 
Biot coef  —  0.60  Literature  
Thermal expansion coef  —  2.00E06  Literature 
Native State Model Results
The Native State Model Calibration presents the modeled pressure, temperature, and stress along the trajectory of Well 5832, along with the data from pressuretemperature logging collected in November 2018. The simulated pressure and temperature match the field measured data quite well. The pressure distribution is largely linear along the length of Well 5832 within the model domain and shows little to no differentiation between the overlying sediments and granitoid reservoir. The modeled temperature distribution also matches the field measured data quite well and shows a break in slope at the sedimentgranitoid contact. The stress gradient data from Phase 2B were collected in the toe of Well 5832. The gradient estimate is plotted over the entire length of Well 5832, along with the modeled native state stress. The numerically modeled stress distributions show some minor perturbations at the sedimentgranitoid contact. The vertical stress was calibrated by adjusting the sediment density and porosity, as well as the density of the granitoid, within the range of measured values, until the modeled vertical stress matched the field measurements at the toe of Well 5832. The simulated minimum and maximum horizontal stresses were slightly overestimated the 0.62 psi/ft and 0.77 psi/ft (respectively) but were within the range of measured values. 
Reference DFN and Upscaling
The FORGE reference DFN model was constructed using FracMan software (Golder Associates, 2019). The DFN incorporates measured surface and well log site data to create planer fractures that communicate as a single hydrological and mechanical system. The reference DFN consists of a deterministic set of fractures intersecting Well 5832 where fracture locations and orientations are known, plus a stochastic set of fractures away from well control. Fracture apertures, permeabilities, and compressibilities were calibrated using measured bulk rock values once the fractures were generated. Hence, the fracture sizes and intensities were established. Once generated, the DFN is exported for use in the DEM simulations and upscaled to provide properties for continuum modeling simulations such as those run by FALCON. These include 3D properties such as fracture porosity and directional permeability. Multiple realizations can be generated to show a range of possible reservoir natural fracture set populations. Fracture orientations are based on Formation Micro Scanner (FMI) log interpretation of Well 5832 (EGI, 2018). These measured orientations have been weighted to account for the bias introduced by sampling from a vertical well. The fractures in the DFN were generated by randomly selecting values from the Terzaghi weighted population and so mirror the measured values quite well. 
Fracture Set Parameters
EW Vertical  NS Inclined Dipping West  NE Steeply Dipping SE  
Set Intensity  P_{32}  7.80E01  1.41E+00  0.31 
[%]  3.10E+01  5.60E+01  12  
Mean Set Orientation  Strike  9.60E+01  1.85E+02  215 
Dip  80 S  48 W  64 SE 