| Resistivity logging tools are run in boreholes to locate and quantify the presence of hydrocarbons (oil and gas) in porous rocks. In order to get better reservoir description, three key formation parameters need to be obtained from resistivity logging: (1) formation resistivity Rt; (2) invaded zone resistivity Rxo; (3) invaded length Lxo. None of the conventional resistivity logging tools, normal, lateral, and dual laterolog (DLL), can provide all three parameters with a single run. The newly developed array resistivity logging tool or High-Definition Lateral Log (HDLL, a product developed by Baker Atlas) resistivity tool addresses limitations of existing dual laterolog systems and provides all three parameters with a single run.; The key developments in this dissertation are: (1) deep (up to 50 inches from borehole wall) Synthetic Focused Resistivity (SFR) logs from HDLL measurements; (2) shallow (close to 10 inches from borehole wall) SFR logs from HDLL measurements; (3) robust computation of SFR via array processing; (4) inverse vertical resolution matching (IRM) method for SFR; (5) inversion of HDLL array data in both vertical and highly deviated wells.; The wellsite interpretation for array resistivity logging is developed by synthesizing focused resistivity logs from the raw HDLL data. The developed SFRs provide a better resolution of the resistivity distribution in both vertical and radial directions than DLL logs. The array processing method is developed to detect and correct errors in the raw measurements, and to evaluate the quality of the data. With the newly developed inverse vertical resolution matching method, we succeed in the resolution matching for the HDLL SFR logs.; The HDLL inversion process developed in this study has three major steps: (1) estimating shallow resistivity structure; (2) estimating deep resistivity structure; and, (3) combining the results of the previous steps to produce the entire true resistivity structure. The results in a final model have a constant Rxo, Lxo, and Rt values in each layer. In the presented cases, HDLL-based interpretations show not only improved delineation of the known reservoirs but also extra pay intervals overlooked by DLL-based interpretations. |
