Formation pressure is the most fundamental data in oil and gas drilling and production; it has an important position in the entire cycle of oil and gas extraction. However, most current prediction methods are limited to parametric methods with fixed models; such that the accuracy does not meet requirements. This is especially true for deeper layers of marine sedimentary basins where the safety density window is extremely narrow. In this study, we propose a novel method to predict pore pressure using machine learning techniques. For the first time, the effective stress (direct output variable) was accurately predicted by a combination of four input variables (2900 sets of data, of which 90% is the training subset and 10% is the testing subset), including longitudinal velocity, porosity, mud content, and density. As such, an accurate prediction of the formation pressure was achieved based on the effective stress theorem. The performance of machine learning techniques was verified by comparing and analyzing the prediction results with traditional parametric single and multivariate models; whereby the best algorithm was chosen by structural optimization and comparative analysis of five algorithms (multilayer perceptron neural network, radial basis neural network, support vector machine, random forest, and gradient boosting machine). Compared with the methods based on parametric one‐dimensional and multivariate models, the machine learning‐based method was determined to possess high accuracy, adequate self‐adaptation, and high fault tolerance (D2 = 0.9981, RMSE = 0.00718 g/cm3). Moreover, the multilayer perceptual neural network algorithm outperformed other machine learning algorithms in terms of goodness of fit, generalization, and prediction accuracy, with D2 = 0.9981 and RMSE = 0.00709 g/cm3. The formation pressure prediction model developed in this study is not affected by the mechanical depositional environment and is applicable to sandy mudstone formations, such that it can be a useful and highly accurate alternative to the traditional formation pressure prediction methods with fixed parameter forms.
The Yinggehai Basin (YGH) in the South China Sea is a Cenozoic, hightemperature, and high-pressure basin. Exploration and development of natural gas resources in the YGH show that owing to the diapir structures make the formation pressure system is complex, overpressure characteristics are obvious, prediction is difficult, and spatial distribution is irregular, which severely restricts the development of natural gas. To evaluate the influence of diapir structures on the formation and distribution of overpressure, we compared and analyzed the ancient and modern geothermal temperatures, organic matter maturity, clay mineral evolution, burial depth, and pressure coefficient in the diapir and nondiapir zones of the YGH. The results showed that of the diapir structures in the YGH occurred in the following order: Uplift-spiking-piercing-depositing. The Dongfang area is dominated by low-amplitude turtle-back diapir structures with weak energy, whereas the Ledong area is dominated by high-amplitude spiked diapir structures with strong energy, both of which pressurize the upper formation. The Changnan area is dominated by high-amplitude pierced/ depositing diapirs with strong energy, which can relieve pressure. The influence of the diapir and its associated structures on the formation of overpressure can be attributed to the transport of the diapir body and the associated fault-fracture system and the local thermal effect caused by the upwelling of hot fluids. Under the influence of the diapir structure, the overpressure top surface of the basin was shallower. According to the degree of influence and intensity of the diapir structure, variations in formation pressure coefficient with depth can be divided into four evolution modes: constant pressure, slow pressurization, "Z" pressurization, and "S" pressurization types. This study provides some guidance and references for the identification and accurate prediction of the overpressure formation mechanism in basins containing diapirs, and helps in formulating efficient drilling and production plans.
Compared with conventional drilling methods, the method of drilling horizontal wells with a slim hole (HW‐SH) has advantages, such as high productivity, environment‐friendliness, and low costs. However, during the drilling process, the narrow annulus may become blocked, which may lead to an increase in the annulus pressure if the wellbore is not clean. This often leads to serious drilling accidents, such as sticking and leaking formations. Many models have been used to study the cutting transport law in an annulus, but the influence of drill‐pipe rotation on cutting transport is often ignored and weakened in these models. None of the existing models can effectively simulate the cutting transport law of an entire well section. Therefore, using these models to predict the cutting transport law in the annulus of the HW‐SH will cause significant errors. In this study, the authors first establish a pipeline model of the same size as the on‐site annulus using computational fluid dynamics (CFD). Subsequently, a one‐dimensional (1D) two‐layer model that uses three correction factors to express the influence of drill‐pipe rotation on cutting transport and annulus pressure loss is developed. Finally, the parameters in the 1D model are calibrated using the CFD simulation results to accurately predict the cutting transport behavior and annular pressure loss. From the calculation results of the annulus cutting quality, annulus section‐cutting concentration, and annulus pressure drop, it was observed that the two algorithms proposed in this study have a high degree of coincidence, thereby confirming the reliability of the model. The 1D model provides close to three‐dimensional (3D) accuracy at a much shorter central processing unit time than the 3D CFD models. In addition, the accuracy of the developed model was verified using the results of an indoor pipeline two‐phase flow experiment conducted at the University of Tulsa, without considering the rotation of the drill pipe.
The Yinggehai Basin has become the main battlefield for natural gas exploration in recent years in China, but the formation pressure system in this area has the characteristics of poor horizontal distribution regularity and severe vertical distribution changes. This paper proposes a new method for predicting the horizontal and vertical distribution trend of formation pressure for the situation of limited offshore drilling data. The method is based on the topological triangulation algorithm, combined with the data set to fit the multivariate interpolation function, and conveniently realizes the prediction and visualization of the three-dimensional horizontal and vertical distribution trend of regional formation pressure.On this basis, through the analysis of the relevant data of the complex accident points that have been drilled, the characteristics of the single well safety density window and the regional safety density window are clarified. The study found that there is no high pressure in the shallow layer of the block, and the formation pressure spreads smoothly in the lateral direction; the high pressure top surface is at 3500m, and the formation pressure gradually increases with the increase in longitude and decrease in latitude in the lateral direction. The formation pressure rises extremely rapidly with the increase of depth in the longitudinal direction, which has the characteristic of "broken line pressurization". The safety density window of a single well is in the shape of a funnel with the characteristic of turning back, and the turning section is between 3500m and 4000m. The safety density window feature of the block can be divided into three stages according to the depth., and the window appears extremely narrow after 3500m (<0.3g/cm3).According to the method proposed in this paper, it is possible to efficiently and conveniently predict and visualize the three-dimensional horizontal and vertical distribution trend of regional formation pressure under the condition of limited formation data during deep-water drilling.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.