SIMPLE ROCK PHYSICS TEMPLATE: A MODIFIED FAWAD – MONDOL MODEL

• Main Author: Agra Pranatikta, Krishna
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/86834
• Institution: Institut Teknologi Bandung

The process of describing physical properties in reservoir characterization is generally done by estimating the two main target physical properties, namely porosity (????) and hydrocarbon fluid saturation (SFL). Fawad and Mondol provide a simple approach in creating an RPT (Rock Physics Template) as a quantitative approach to perform the desired target physical property description. The RPT is simply constructed using a combination of Wyllie's Time Average, bulk density, and Lee velocity ratio equations and has the flexibility to fit the RPT in AI vs ????????/???????? domain to the data by adjusting the model parameters G (controls the vertical static shift) and N (controls the slope). Despite its simplicity and flexibility, the RPT has incompatibilities with the concept of rock physics. These include the relationship of velocity ratio that only changes with porosity, the absence of critical porosity as the end-point of the RPT model, and the lack of a workflow to obtain model parameters, G and N, that fit the data. Therefore, this study focuses on modifying the Fawad and Mondol RPT model to be more in line with rock physics concepts and establishing a workflow to apply this approach appropriately. The research results show a modified Fawad and Mondol RPT model that is consistent with rock physics concepts. These include a modified velocity ratio relationship that considers both porosity and fluid saturation in accordance with the Gassmann concept, determination of critical porosity using the Nur Model to complete the end-point information in the RPT, and a workflow to obtain model parameters, G and N, that fit the data. This workflow utilizes the integration of various approaches, namely CPEI (Curved-Pseudo Elastic Impedance) attributes to regularize the random pattern of fluid saturation, histogram matching to align the statistical behavior of the model to the data, and forward modelling to obtain the best model parameters, G and N, based on the smallest misfit. Testing this workflow on synthetic and field well data shows the reliability of the proposed workflow based on the best model parameters, G and N, obtained, where the resulting model has a high correlation with the reference data