Early gas kick detection-inversion-control integrated system: The significance of applications of managed pressure drilling: A review

With the development of oil drilling technology and equipment, oil exploration is moving towards deep water high temperature and high pressure (HTHP) formations. However, deepwater drilling operations often encounter problems such as shallow gas, narrow formation pressure windows, and long non-productive time. As a new drilling method, managed pressure drilling (MPD) technology has been applied in deepwater drilling fields after upgrading conventional drilling equipment and construction technology. This study describes the technical characteristics of MPD operations, draws conclusions about several variants of MPD and their applications, and compares different states of equivalent circulating density (ECD) and bottom-hole pressure in conventional drilling and MPD. Early gas kick detection (EGKD) is at the core of the drilling safety issues. This paper summarizes the EGKD methods, which are classified into two categories: conventional gas kick detection and unconventional gas kick detection. The two major categories are systematically reviewed. The review of the existing literature shows that for conventional gas kick detection the sensitivity of mud pit gain ranges from 2.79 to 8.18 bbl, the accuracy of delta-flow measurements ranges between 25 and 50 gpm. Combining the MPD technology, we put forward a downhole condition identification model based on the physical mechanism neural network (PMNN) algorithm. In addition, it is difficult for drilling to accurately describe wellbore fluid flow by relying on forward modeling, and quantitative interpretation and analysis of downhole parameters need to be further realized. Moreover, conventional single-measurement point data cannot precisely reflect and explain the change of downhole parameters, thus combining the forward modeling, a multi-measurement point and multi-parameter inversion (MPMPI) method within the intelligent drill pipe technology is proposed. Finally, this paper summarizes the automatic control technology of MPD using nonlinear model predictive controllers (NMPC) algorithms for real-time intelligent optimization and decision-making.