Accuracy and Correction of Hook Load Measurements During Drilling Operations

Abstract

Hook load is a very important parameter used during drilling operations to control the weight on bit and to assess possible deteriorations of the downhole conditions such as poor hole cleaning or excessive tortuosity. However, the quality of the hook load measurement is often regarded as quite poor. This paper investigates the sources of uncertainties associated with the hook load measurement and proposes correction methods in order to obtain better data quality. In practice, the hook load is normally measured indirectly, either in the travelling equipment or as a tension in the dead-line. This apparent hook load is subject to load-generating forces between the measurement location and the top of the string, including the weight of the mud hose attached to the top drive, imperfect tension transmission across sheaves and gravitational and inertia forces associated with weight and rotation of the drill line, respectively. The load contribution from these forces can amount to several metric tons and this must be accounted for whenever the true hook load is to be derived. In this paper, a mathematical model describing the forces affecting the hook load measurement is developed. The model predicts true hook load as a function of block position, velocity and other conditions that can influence the measurement like the mud weight or whether the dolly is retracted or not. It is a generalization of the industry-standard hook load estimation technique and successfully replicates a number of trends observed in the field. Extensive model evaluation is performed by comparing model results to field measurements of simultaneous travelling equipment and dead-line tension derived hook loads. The model is also compared with data from a laboratory scale test rig equipped with distributed load cells connected to a data acquisition system. The hook load correction model works exclusively at the software level and represents a potentially cost-efficient alternative to a direct hook load sensor installed in the internal BOP of some top drives. Assessment of model uncertainty and its effect on hook load predictions is given particular attention as this ultimately determines whether the correction method can replace direct hook load measurements in practice.