Fall-off Test Pressure Derivatives to Weight Geological Model Centroids for Simulation of Polymer Injection
Research output: Research › Master's Thesis
After a polymer pilot indicated encouraging incremental oil production in the 8 TH reservoir of the Matzen field, gained insights are deployed for a first field-rollout. A probabilistic workflow was developed to forecast incremental oil production, which comprises a geological sensitivity study followed by the extraction of a representative subset of models (model centroids) for full-physics dynamic simulations. While clustering is currently performed based on virtual tracer responses obtained using streamline simulations in order to select a subset, which preserves maximum geological diversity, dynamic properties are treated as tuning parameters during the model calibration step. Lacking a detailed understanding of in-situ polymer properties, currently broad ranges of the relevant parameters are considered. This thesis studies the applicability of pressure fall-off test data to be incorporated into the clustering step to weight model centroids and decrease both the uncertainties associated with geological and dynamic parameters. Primary focus is set on investigating the in-situ viscosity of the polymer solution in horizontal well applications. While in-situ polymer properties represent one of the main sources of uncertainty in the current workflow, they constitute one of the key design parameters for optimizing displacement efficiency as well as project’s economics. Fall-off tests have been simulated for a wide range of generic reservoir models and examined using pressure transient analysis (PTA) with a particular focus on pressure derivatives. For single-phase water models, PTA provides a reliable tool to characterize both geological features and essential polymer properties. An approach is developed to yield the in-situ polymer viscosity based on the characteristic shape of the pressure derivative plot. The degree of reservoir heterogeneity was found to have a strong influence on interpretation reliability. Reservoir quality index (RQI) depending polymer adsorption and viscosity have been implemented to investigate the effect of different rock types and to approximate the pseudoplastic behaviour of the polymer solution. For multi-phase models, the analysis of simulated fall-off tests has not yielded a consistent reservoir description; however, in-situ viscosity can still be approximated with the proposed workflow in case a pure water injection response is provided as baseline. This finding is of importance for the operational planning of future fall-off tests in polymer injection wells. Although the applicability of PTA to weight actual geological models could not be confirmed at this stage, outcomes of this thesis will help to constrain the interpretation of planned fall-off tests to be conducted in recently drilled horizontal polymer injectors.