Standalone Cased-Hole Formation Evaluation and Uncertainty Propagation to Increase Well Deliverability in Challenging Environments

Abstract

In new wells with hole stability problems, which require to be cased and cemented immediately, and in old wells without a complete or reliable log dataset, a formation evaluation by means of quantitative open-hole (OH) log interpretation is not possible. Cased-hole (CH) logging can represent the only solution, despite being usually considered highly uncertain. This paper discusses the standalone use of CH logs, integrated in a probabilistic framework, for petrophysical characterization and uncertainty propagation purposes. The approach consists of a full statistical workflow aimed at a formation evaluation with only CH logs as input, also including capture cross-sections, carbon/oxygen ratios and inelastic spectra. Several wells with complete OH petrophysical characterization have been used to define the rate of success of this methodology in different scenarios. Furthermore, a Monte Carlo framework is introduced to account for the uncertainty quantification of the CH outputs. The final outcome is the set of probability distribution functions of the petrophysical properties, the most probable scenario and the associated uncertainty. Three real operative applications, in scenarios with no OH logs available, are presented: an old well without a complete/reliable OH log dataset (gas- and oil-bearing sandstone reservoir with variable salinity), and two new wells with hole stability problems (in a sandstone and in a carbonate oil-bearing reservoir at high formation water salinity). In the latter cases, numerical simulations are performed to correctly handle mud-filtrate invasion effects for a robust modeling also in the shallow zone investigated by CH logs. In all these challenging conditions, a complete formation evaluation has been obtained, and successfully used to select proper intervals to perforate. The increased hydrocarbon production driven by the outcomes of the standalone CH characterization further validates the efficiency of this method. Though CH logging is a well-known technology in reservoir monitoring, its successful standalone use for reliable formation evaluation can represent an important step forward in reservoir characterization processes, in all those cases where OH data are not available or too risky to acquire. Finally, the value of uncertainty analysis has a huge relevance for appropriate production optimization and reservoir modeling strategies.