The application of PEST to real conditions is a challenge since it involves the setup of the automated parameter estimation and the complexities of the groundwater response. We have developed an applied case of automatic parameter estimation on a groundwater flow model related to a site drainage with 6 pumps and steel sheet piles. The calibration is based on observations from a pumping well and two outer monitoring points over a period of 120 days. The case involves as well the sensitivity analysis of hydraulic conductivities from heterogeneous soil and sheet piles together with the conductance of neighboring water channels.

There are many topics to take into account when we create a groundwater flow model. One of the most important is to represent the available geological information into the distribution the hydraulic parameters. We have researched a simple and reproducible workflow to create a geological model from point shapefile and insert the modeled geological units into Model Muse with corresponding K values. This webinar cover the whole procedure to create a geological model with Gempy and Aquifer App and the necessary codes to create a xyz shapefile that is later imported in Model Muse with the Kx value extracted from the attribute table.

Aquifer App offers a friendly, clean and powerful way to create Gempy scripts for geological modeling. We have developed an applied case of a 3D petroleum system model for a part of the Williston Basin, USA that contains the Nesson Anticline. Top layer information was provided as raster format where elevation and orientation for random points were extracted with Python codes using the Gemgis package and exported in Gempy input file format. On Aquifer App the processed CSV files were inserted and the geological model was set up with the corresponding geological sequence. 

Lithium is present in some salt lakes along with other brines in different concentrations. In order to simulate the right groundwater flow regime of Lithium brine along other brines we need to model variable density flow from multiple species. This type of modeling is possible on the latest versions of Modflow 6 with the BUY package and Model Muse. 

We have developed an applied case of variable density flow modeling on a andean salt lake that is pumped from 3 wells in 10 years. The model has a grid refinement close to the lake and works with 6 brines: B(OH)3, NaHCO3, NaCl, KHSO4, LiHSO4 and NaHSO4. The example implements the BUY package with all the required parameters together with the configuration of the boundary conditions and initial conditions. Concentration and water balances are processed with Python to evaluate the brine migration over time and the impact of pumping on the water balance.