|dc.description.abstract||Poor hole-cleaning and wellbore instability are well-known problems during drilling operation. If they are not diagnosed in their early stage of development, they result in non-productive time (NPT). In order to reduce NPT, potential problems due to poor hole-cleaning or wellbore instability need to be revealed and preferably distinguished between them before they occur and take the correct measures to rectify the situation.
Hook load (HKL) is one of many parameters of the real time drilling data (RTDD) measured at the surface during drilling operations. In this work the RTDD were investigated closely to find the deviation of HKL from normal, and snapshots for each detected case were presented.
From 12 well studied, 26 downhole restriction cases were found. The cases were distinguished based on HKL deviation due stability of the wellbore (Abnormal HKL type 1), HKL deviation due to cleaning of the wellbore (Abnormal HKL type 2), and cases of formation related. Cases (1 to 9) were of Abnormal HKL type, cases (10 to 16) were of Abnormal HKL type 2, and cases (17 to 26) were of formation related. The ontologies for both Abnormal HKL type 1 and Abnormal HKL type 2 were also presented along with the average probabilities of the main cause.
Distinguishing the downhole restriction into restriction due to wellbore and wellbore wall is a challenging task. However, snapshots of HKL response are useful in creating data agents to detect downhole problems before it occurs or become serious||