A Data Assimilation-Based Method for Real-Time Monitoring and Estimation of Gas Influx Size and Distribution During Gas Migration in a Marine Drilling Riser

Abstract

Effective well control is essential in marine drilling operations to prevent catastrophic blowouts that endanger human lives and cause severe environmental damage, including long-lasting impacts on marine ecosystems. Accurate real-time monitoring and management of well conditions after taking gas influxes are critical for maintaining safety and environmental protection during offshore drilling operations.When a gas influx becomes trapped in a closed marine drilling riser, its upward migration can result in excessively high surface pressure, presenting significant risks to offshore drilling and well control. Therefore, the accurate estimation of downhole gas influx size and distribution during gas migration events is crucial for effectively managing these situations. The complexities of wellbore two-phase flow and gas migration make direct measurements challenging, leading to potential inaccuracies in traditional modeling approaches. This study introduces a Data Assimilation (DA) technique, employing the Ensemble Kalman Filter (EnKF), to improve the real-time estimation of gas influx rates and distribution, thereby enhancing the accuracy of two-phase flow models in marine drilling.This research utilizes full-scale gas influx migration experiments conducted at Louisiana State University, where gas was injected from the bottom of an experimental well to simulate a riser gas event in a Water-based Mud (WBM) system. Key real-time measurements, including downhole pressure gauges at multiple depths and Distributed Acoustic Sensing (DAS) data, were employed to validate the estimated gas influx size and distribution within the riser using the EnKF. The overall gas influx distribution was captured through the real-time estimation of a dimensionless distribution index (DI), offering additional insights into the spatial characteristics of the migrating gas. Additionally, a Drift Flux Model (DFM) calibrated online through DA was used to predict flow parameters such as gas void fraction profiles within the riser during migration. The validation of these predictions demonstrated a strong correlation between the estimated and measured values.The outcomes of this study underscore the effectiveness of the proposed method in estimating parameters that are otherwise difficult to measure directly, thereby improving the predictive accuracy of gas influx behavior in marine drilling risers. This method offers significant practical value by enhancing real-time decision-making and risk management in marine drilling operations that involve gas migration, contributing to the broader goals of offshore drilling safety and marine environment protection.