Kick Detection and Influx Size Estimation during Offshore Drilling Operations using Deep Learning

Abstract

An uncontrolled or unobserved influx or kick during drilling has the potential to induce a well blowout, one of the most harmful incidences during drilling both in regards to economic and environmental cost. Since kicks during drilling are serious risks, it is important to improve kick and loss detection performance and capabilities and to develop automatic flux detection methodology. There are clear patterns during a influx incident. However, due to complex processes and sparse instrumentation it is difficult to predict the behaviour of kicks or losses based on sensor data combined with physical models alone. Emerging technologies within Deep Learning are however quite adapt at picking up on, and quantifying, subtle patterns in time series given enough data. In this paper, a new model is developed using Long Short-Term Memory (LSTM), a Recurrent Deep Neural Network, for kick detection and influx size estimation during drilling operations. The proposed detection methodology is based on simulated drilling dataand involves detecting and quantifying the influx of fluids between fractured formations and the well bore. The results show that the proposed methods are effective both to detect and estimate the influx size during drilling operations, so that corrective actions can be taken before any major problem occurs.