Intelligent Method of Identifying Drilling Risk in Complex Formations Based on Drilled Wells Data

Abstract The proposed intelligent method, which is based on the pressure data and the spatial location of the adjacent drilled wells, and the spatial continuity of the targeted region, is able to establish the formation pressure profile with credibility, the safe window with credibility of drilling fluid density, the probability profile of drilling risk and the risk interval of drilling operation. As a result, the proposed method achieves the intelligent identification of drilling risk in complex formations. In this intelligent method, the concept of formation pressure matrix is introduced firstly, and an epitaxial stratigraphic pressure migration algorithm is established to calculate the pressure of the targeted wells in the same tectonic region based on the depth adjustment and the inverse distance weighting algorithm. Secondly, the probability of drilling risk is calculated. Thirdly, the membership function of different degree interval is constructed by fuzzy mathematics theory combining with the accident record of drilled wells; and finally, divide the membership degree of drilling operation risk to achieve the intelligent identification of drilling risk. The intelligent method has been successfully applied in BD gas field in Indonesia. Nine drilling accidents such as kicks, collapses, leakages and stickings have been accurately identified in a certain well. The formation pressure of adjacent drilled wells is transplanted to the targeted wells and the pressure prediction is more accurate. Meanwhile, combined with the accident records of drilled wells, the intelligent identification of drilling risk can be carried out. The two issues mentioned above have been the focus of researchers. Under normal circumstances, the designers use stratigraphic comparison method, in which the formation pressure of the drilled wells is directly applied to the targeted well and does not consider the spatial distribution characteristics of the targeted formation. And the designers cannot intuitively determine the possibility of drilling risk according to the numerical drilling risk probability value. This intelligent method gives the transplantation method calculating the formation pressure of the targeted well in the same tectonic region and divides the drilling risk interval. This method has been used to analyze the risks in BD gas field in Indonesia and the LW oil field in South China Sea. Case studies were carried out to verify the method, in which the predicted formation pressure was compared with that derived from the LWD. The results showed that the maximum relative errors were within 2.76%. Example analysis showed that the identifying result of this method agreed with the actual engineering. The identification results of proposed intelligent method agrees with the engineering practical situation, satisfies the need of drilling engineering and can provide basis for identifying risk in drilling design.