Failure Characterization During Oil Well Drilling

Abstract

Accessing remaining oil and gas reserves around the world is getting more and more complicated and wells have to be longer, deeper and more complex in order to compensate. With an increase in well complexity comes new complex drilling related wellbore problems. Drilling operations are extremely costly and all time spent being non-productive is an unnecessary company cost. Wellbore failures can happen at any time during the drilling operation and can happen to any part of the drilling equipment or wellbore. Early detection of any failures can potentially allow for early counter-measures to be taken on the surface in order to minimize damage and costs. This is especially vital for companies in the industry, to date, as the oil price is so low and costs must be kept to a minimum.

Ontology engineering will be applied to known downhole failures. The most common failures during drilling operations, from the basis of a study conducted on 427 wells offshore Gulf of Mexico, will be discussed in detail in order to understand when and where the different failures occur and for what reasons. Top-down ontology models will be used for mapping out failures in the theory sections and bottom-up methodology will be utilized for wellbore failures found in real time drilling data.

Three failure cases were found in analysis of real time drilling data. The first failure case was induced lost circulation, the second was lost circulation due to naturally fractured formations and the third was a motor stall failure case. Evaluation of the cases revealed that lost circulation is undetectable (prior to time of occurrence) through analysis of real time drilling data. Motor stall is a detectable failure and initial indications of motor stall were observed up to 18 hours prior to POOH.