DataDRILL: Formation Pressure Prediction and Kick Detection for Drilling Rigs

Abstract

Accurate real-time prediction of formation pressure and kick detection is crucial for drilling operations, as it can significantly improve decision-making and the cost-effectiveness of the process. Data-driven models have gained popularity for automating drilling operations by predicting formation pressure and detecting kicks. However, the current literature does not make supporting datasets publicly available to advance research in the field of drilling rigs, thus impeding technological progress in this domain. This paper introduces two new datasets to support researchers in developing intelligent algorithms to enhance oil/gas well drilling research. The datasets include data samples for formation pressure prediction and kick detection with 28 drilling variables and more than 2000 data samples. Principal component regression is employed to forecast formation pressure, while principal component analysis is utilized to identify kicks for the dataset's technical validation. Notably, the R2 and Residual Predictive Deviation scores for principal component regression are 0.78 and 0.922, respectively.

Figures (5)

• Figure 1: Overview of the inferential measurement system (IMS) architecture for drilling rigs. The sensors and actuators are connected to the drilling rigs to measure different drilling variables. After preprocessing, the secondary variables related to the primary variable are chosen. The inference model is then used to predict the primary variable, such as formation pressure from the secondary variables. The inference model can also take feedback from the predicted data and update its parameters.
• Figure 2: OTR simulator in the experimental room. The main simulation computer is connected to a workstation PC, HMI, and an instructor laptop. The workstation PC hosts the API for different rig packages and is used to collect data from the simulator. The instructor's laptop is used to set drilling scenarios based on geological data.
• Figure 3: Patterns of different layers of the designed formation for Scenarios 1 and 2 (Figures are taken from the simulator)
• Figure 4: (a) Mutual information scores between the Formation pressure and other significant secondary variables. (b) Pearson correlation coefficient matrix between the Formation pressure and other highly correlated secondary variables
• Figure 5: (a) Predicted vs. exprected regression line for principal component regression, (b) shows the expected variance for individual principal components and the cumulative variance, (c) reconstruction error for the training set, (d) reconstruction error for test set