Revolutionizing Reservoir Structure Modeling Workflow for Exponential Acceleration of Horizontal Well Placement in Complex Reservoir Models

Abstract Traditional structural modeling methods struggle with time-consuming updates and imprecise well placements within reservoir models. Current algorithms falter with closely spaced wells, complex trajectories, and geological uncertainties. Integrating these wells manually is arduous, resulting in neglected models and potential financial losses. It's crucial to overcome these obstacles for a more reliable reservoir modeling and well placement strategy in the petroleum subsurface industry. In this paper, these challenges are addressed through an innovative approach to reservoir structure modeling. The method employs the Correlated Horizon Intervals and Bayesian Analysis (COHIBA) engine, a mathematical framework adept at managing input uncertainties. It introduces four key elements: 1. Zone Log Information: Utilizes zone logs for accurate structure and well placement, surpassing traditional methods reliant solely on marker points. 2. Flexible Horizons: Allows adjustment of reference horizons in a geostatistical manner, unlike locked traditional methods. 3. Well Path Uncertainty: Incorporates vertical uncertainties along well paths via an uncertainty log, preventing exclusion of anomalous data. 4. Streamlined Automation: Simplifies automation by recording and saving the process as a user-friendly workflow, facilitating quick updates with new well data. The automated and robust nature of the method was successfully demonstrated in a real-world scenario with over 230 horizontal wells, significantly outperforming standard methods. There were several cases of challenging trajectories. These included, but were not limited to, wells that scratched unwanted zones several times along their trajectories, wells that have crossed several faults with throws, wells that were in a highly congested area, and wells with erroneous deviation surveys. By minimizing manual work, saving time, and reducing errors, the approach enhances reservoir models' value for decision-making in reserves calculations, development planning, and drilling operations. This results in significant cost reductions, improved operational efficiency, and increased profitability, making it a transformative tool for the petroleum subsurface industry. Previously, the average time it took to incorporate a single well into the structure model was 1-2 working days and involved a lot of tedious and manual process. Now, it takes less than 1 hour from loading to completion. This means that the employee is now able to get the job done in a fraction of the time (90% less time) in addition to achieving much more accurate and statistically sound result since the well placements are all true to reality. Implementation of continuous and automated updates has transformed our modeling approach into an evergreen model. Gone are the days of stagnant, outdated static models. This dynamic process enables real-time adjustments while drilling, enhancing operational efficiency and accuracy. The result is a more reliable structure model, leading to improved property models and reliable hydrocarbon volume calculations, marking a significant advancement in operational effectiveness and decision-making within the scientific community.