Early kick detection and well control decision-making for managed pressure drilling automation

Abstract

Early kick detection and blowout prevention is a major concern in any drilling operation. Usually, conventional well control procedures rely merely on the pit gain and variations in pump pressure as the primary indicators to detect and handle a kick. These indicators are rudimentary and unreliable for drilling costly offshore wells while advanced well control techniques demand more reliable indicators for early detection of kicks. For instance, early kick detection plays a vital role in deploying Managed Pressure Drilling (MPD) techniques successfully.Moreover, in conventional drilling, the only widely accepted response to a gas influx is to shut the well in on the BOPs. The Constant Bottom Hole Pressure (CBHP) variation of MPD allows for circulating small and medium-sized influxes out without the need to shut in the well, either by increasing casing backpressure and/or by adjusting pump flow rate. Evident benefits include a faster response to the kick (with clear safety benefits) and a considerable time and cost savings compared to the conventional approach.Unfortunately, there is currently no method available to decide upon the best MPD-CBHP response during an influx, because such a response depends on many parameters, such as Maximum Allowable Surface Pressure (MASP), wellbore geometry, fracture pressure of any weak zones exposed in open hole, type of drilling fluid, influx size, etc. To remedy this situation, we present a fast and accurate decision-making method that overcomes the traditional difficulties and automatically selects a best well control response. The method incorporates a transient multi-phase model based on mass and momentum conservation. A numerical simulator was developed for this purpose. Comparing the actual well response with simulated results during an influx event and following the proposed algorithm allows the best response to be taken in a quick, convenient, and accurate manner. Moreover, the approach lends itself well to fully automated well control.Here, we describe the new method and its validation using available two-phase experimental data for annular air–water and air-mud flow for a wide range of flow rates and flow regimes. We believe that the method provides a powerful new decision-making and design tool in MPD operations that ensures delivery of safer wells with lower non-productive time and associated trouble cost spent on well control operations.