رسائل ماجيستير
English
2019
Estimation Of Hole Cleaning Condition In Real-time While Drilling (Operational Point Of View)
Abdulazeez Abdulraheem , Salaheldin Elkatatny , محمود نادر محمود الزناري
كلية هندسة البترول وعلوم الأرض-جامعة الملك فهد للبترول والمعادن · السعودية
Hole cleaning is the main parameter while considering the quality and efficiency of drilling deviated and horizontal wells. Hole cleaning is being affected with many factors while drilling such as: weight on bit (WOB), rate of penetration (ROP), rock geomechanics, drilling fluid properties, and rig hydraulics.Hole Cleaning has a great impact on drilling efficiency, as the hole cleaning getting better the drilling overall efficiency will increase. Bad hole cleaning in many cases leads to non-productive time (NPT). Stuck pipe, slow ROP and drilling bit damage are common problems related to inefficient hole cleaning.To measure the hole cleaning while drilling, field, and experimental measurements will need to be conducted with the complicated and high-cost process. For example, while drilling the rig crew will need to handle a long process of drilling parameter optimization (e.g. weight on bit, torque limit, flow rate, rate of penetration …. etc.) to achieve the best hole cleaning scenario for a specific section. This process will be costly as it will not be counted as productive time.A lot of researches were conducted to evaluate the hole cleaning and related cutting transport efficiency while drilling. The main gaps in these researches come from the fact that, these researches contain mainly experimental and empirical models which most of the time will not reflect all factors affecting the hole cleaning even it may also be not applicable in the field from the operational side of view.With the new technology called Artificial Intelligence (AI) and its related applications such as; support vector machine (SVM), adaptive neuro-fuzzy interference system (ANFIS), and artificial neural network (ANN), downhole parameters affecting the hole cleaning process will be predicted with high accuracy and hence, with the right model the hole cleaning condition will be measured.This thesis proposes new means of predicting hole cleaning in both vertical and highly deviated wells using a new model which includes two artificial intelligence models. The first artificial intelligence (AI) model was built to predict the drilling fluid rheology parameters (yield point YP and plastic viscosity PV). The second artificial intelligence (AI) model was built to predict the equivalent of circulation density (ECD) while drilling.These two AI models will be part of a new approach called hole cleaning index (HCI) to estimate the hole cleaning condition in real-time bases while drilling in vertical to highly deviated wells with all ranges and through well different drilling sections.