Session: 07-17-03: Machine Learning and Artificial Intelligence in Dynamics, Vibrations and Control

Paper Number: 146060

146060 - Real-Time Bayesian Estimation of Drilling Tool Lateral Motion by Fusing Physics With Sensor Data 

The performance and health of logging while drilling (LWD) tools can be significantly impacted by the drilling dynamics of the bottomhole assembly (BHA). With drilling dynamics becoming increasingly harsh, BHA whirl is one of the undesired motions the tool can face during drilling operations, where the BHA follows an eccentric rotation about a point along the wellbore other than the geometric center. For example, parasitic motion that causes a plethora of distortions to the echo train, at times manifesting as erratic noise or a complete loss of cohesion has been a measurement quality concern for LWD tools. Advanced motion sensors can be installed only at discrete and limited locations of various LWD tools. If we simply take the sensor measurements as the dynamic states at locations away from the sensors, risks associated with tremendous uncertainty can be present in decision making. The farther away from the sensors, the more uncertain the dynamic states will become. Therefore, robust inference of the motion states at the LWD measurement sensor of interest with motion sensors at discrete locations is critical for risk flagging and quality assurance of LWD answer products.

 

In this paper, we present a Bayesian data assimilation method that fuses physics with sensor data for inferring the dynamic states at points of interest on the BHA with proper uncertainty quantification. A 4.75 inch-LWD tool has been used as a use case, where the dynamic states at the LWD measurement sensor can be predicted in real time with the measurements at the motion sensor as the required inputs. This was achieved with a developed transfer function that utilizes unscented Kalman filtering technique. The robustness of the transfer function was evaluated with synthetic data obtained from drilling dynamics and finite element analysis (FEA) simulations for various BHA configurations and drilling conditions. It was found that the prediction by the transfer function agrees favorably well with the true states of motion at the LWD measurement sensor. The developed transfer function method was further assessed with experimental roll test data, which is considered as close to drilling conditions. The prediction by the transfer function was found consistently close to the ground truth at various rotation speeds.

Presenting Author: Ke Li SLB

Presenting Author Biography: Ke Li is currently a technology program manager with SLB.

Authors:

Fei Song SLB
Kevin Shi formerly with SLB
Ke Li SLB
Amine Mahjoub SLB
Sepand Ossia SLB
Ives Loretz SLB
Robson Serafim SLB