Search

The main purpose of this thesis work is focused on the use of PEST (Parameter Estimation) to calibrate numerical models of High Enthalpy Geothermal Reservoirs (HEGR). PEST is a parameter estimation and analysis of the uncertainties of complex numerical models tool, that can be instructed to work with a standalone simulator. So, the T2Well-EWASG was used as coupled wellbore-reservoir simulator for multiphase-multicomponent HEGR. The idea of this thesis work is that the possibility to implement some automation degrees in the wellbore-reservoir model calibration task would improve substantially the Reservoir Engineers work. To become familiar with PEST, it has been necessary a preliminary training to learn how to manage its input files, its keywords, and the utility programs having the function of verifying the correctness and consistency of the created files. Then, one of the examples of PEST manual (which Fortran source code is supplied) was reproduced and analyzed, and subsequently modified. In particular, starting from this example, a simple linear model with two free parameters, some changes have been performed: "fixing" a parameter to inhibit its change during the calibration; reading a more complex model output file respect to the original example; inserting dummy data that should not be processed and instructing PEST to consider only the data of interest; changing the model adding parameters to be calibrated, and including them in the analysis changing the PEST inputs files. Finally, these skills were applied to use PEST with T2Well-EWASG to calibrate a numerical model, relative to a real HEGR, previously calibrated via a trial and error approach in a PhD thesis work. Among the real data used there were also short production-tests done in a geothermal field located in the Dominica Commonwealth. The preliminary results show that the PEST-T2Well-EWASG calibration system works fine, and that it is a useful tool that can improve the work of reservoir engineering.

The main purpose of this thesis work is focused on the use of PEST (Parameter Estimation) to calibrate numerical models of High Enthalpy Geothermal Reservoirs (HEGR). PEST is a parameter estimation and analysis of the uncertainties of complex numerical models tool, that can be instructed to work with a standalone simulator. So, the T2Well-EWASG was used as coupled wellbore-reservoir simulator for multiphase-multicomponent HEGR. The idea of this thesis work is that the possibility to implement some automation degrees in the wellbore-reservoir model calibration task would improve substantially the Reservoir Engineers work. To become familiar with PEST, it has been necessary a preliminary training to learn how to manage its input files, its keywords, and the utility programs having the function of verifying the correctness and consistency of the created files. Then, one of the examples of PEST manual (which Fortran source code is supplied) was reproduced and analyzed, and subsequently modified. In particular, starting from this example, a simple linear model with two free parameters, some changes have been performed: "fixing" a parameter to inhibit its change during the calibration; reading a more complex model output file respect to the original example; inserting dummy data that should not be processed and instructing PEST to consider only the data of interest; changing the model adding parameters to be calibrated, and including them in the analysis changing the PEST inputs files. Finally, these skills were applied to use PEST with T2Well-EWASG to calibrate a numerical model, relative to a real HEGR, previously calibrated via a trial and error approach in a PhD thesis work. Among the real data used there were also short production-tests done in a geothermal field located in the Dominica Commonwealth. The preliminary results show that the PEST-T2Well-EWASG calibration system works fine, and that it is a useful tool that can improve the work of reservoir engineering.