| dc.description.abstract | Surge and swab pressure occur while tripping in and out, respectively, of a wellbore during drilling operations. High tripping speed can lead to fracturing the well formation, whereas low tripping speed can lead to an increase in non-productive time and cost. Hence, there is a requirement to predict surge/swab pressure accurately. Several analytical and machine learning models have already been developed to predict surge/swab pressure. However, these existing models make use of numerical calculations to generate the data.
This thesis explored four different supervised machine learning models, i.e., Linear Regression, XGBoost, Feedforward Neural Network (FFNN), and Long- Short-Term Memory (LSTM). In this empirical study, real field data from the Johan Castberg field provided by Equinor is utilized to develop the mentioned machine learning models. The results indicated that XGBoost was the best performing model with an R^2-score of 0.99073. This trained model can be applied during a tripping operation to regulate tripping speed where repetitive calculation of surge/swab pressure is required. | |