The Pressure Buildup Well Test Analysis considering Stress Sensitivity Effect for Deepwater Composite Gas Reservoir with High Temperature and Pressure

Some deepwater gas reservoirs with high temperature and pressure have obvious stress sensitivity effect resulting in difficulty in well test interpretations. The influence of stress sensitivity effect on the pressure drawdown well test is discussed in many papers. However, the influence on the pressure buildup well test is barely discussed. For practices in oilfields, the quality of pressure data from the drawdown stage of well test is poor due to the influence of production fluctuation. Thus, the pressure data from the buildup stage is used for well test interpretations in most cases. In order to analyze the influence of stress sensitivity effect on the pressure buildup well test, this paper establishes a composite gas reservoir pressure buildup well test model considering the stress sensitivity effect and the hysteresis effect. Numerical solutions to both pressure drawdown and buildup well test models are obtained by the numerical differentiation method. The numerical solutions are verified by comparing with analytical solutions and the homogeneous gas reservoir well test solution. Then, the differences between pressure drawdown and buildup well test curves considering the stress sensitivity effect are compared. The parameter sensitivity analysis is conducted. Compared with the conventional well test curve, the pressure derivative curve of pressure drawdown well test considering the stress sensitivity effect deviates upward from the 0.5 horizontal line at the inner zone radial flow stage, while it deviates upward from the M/2 (mobility ratio/2) horizontal line at the outer zone radial flow stage. However, for the pressure buildup well test curve considering the stress sensitivity effect, the pressure derivative curve gradually descends to the 0.5 horizontal line at the inner zone radial flow stage, while it descends to the M/2 (mobility ratio/2) horizontal line at the outer zone radial flow stage. The pressure derivative curve of pressure buildup well test considering the hysteresis effect is higher than the curve without considering the hysteresis effect, because the permeability cannot be recovered to its original value in the buildup stage after considering the hysteresis effect. Meanwhile, skin factor and mobility ratio have different effects on pressure drawdown and buildup well test curves. Based on the model, a well test interpretation case from a deepwater gas reservoir with high temperature and pressure is studied. The result indicates that the accuracy of the interpretation is improved after considering the stress sensitivity effect, and the skin factor will be exaggerated without considering the stress sensitivity effect.

Thermodynamic Basis of Brine Density on Pressure, Temperature, and Chemical Composition in Ultrahigh Pressure/High Temperature Environments

Summary Temperature change and the pressure/volume/temperature (PVT) response of wellbore annular fluids are the primary variables that control annular pressure buildup in offshore wells. Though the physics of annular pressure buildup is well understood, there is some ambiguity in the PVT models of brines. While custom tests can be performed to determine the PVT response of brines, they are time-consuming and expensive. In this light, our paper presents a method to determine the density of brines from their chemical composition, as a function of pressure and temperature. It compares theoretical predictions with the results of tests on brines used in our industry and available test data from the oil and gas and other industries. In 1987, Kemp and Thomas used the principles of chemical thermodynamics to develop equations for the density of brines as a function of pressure and temperature and their electrolytic actions. However, their paper contained two (inadvertent, and probably typographical) errors. One of the errors lay in the set of the Debye-Hückel parameters, and the other was contained in the coefficients of the series expansion for the infinite dilution molal volume. Furthermore, they (inadvertently) did not mention the role of a crucial parameter that accounts for the interaction between the ionic constituents of the salt. As a result, nearly a generation of engineers in our industry has been unable to reproduce their valuable results or apply their technically rigorous results to other brine chemistries. In this paper, we return to the basic equations of chemical thermodynamics and the principles of stoichiometry and delineate the inadvertent errors that had crept into the Kemp and Thomas equations. We then present the rectified equations and reproduce their example with the corrected results. Further, we compare the predictions from the original Kemp and Thomas work with results from a leading chemical engineering model. Finally, we compare the results of theoretical models with test measurements from the laboratory and characterize the uncertainty inherent in each model. Thereby, we have rendered the Kemp and Thomas (1987) model useful to the well design community.

Evaluating Precision of Annular Pressure Buildup (APB) Estimation Using Machine-Learning Tools

Summary Annular pressure buildup (APB) is caused by heating of the trapped drilling fluids (during production), which may lead to burst/collapse of the casing or axial ballooning, especially in subsea high-pressure/high-temperature wells. The objective of this paper is to apply machine-learning (ML) tools to increase precision of the APB estimation, and thereby improve the fluid and casing design for APB mitigation in a given well. The APB estimation methods in literature involve theoretical and computational tools that accommodate two separate effects: volumetric expansion [pressure/volume/temperature (PVT) response] of the annulus drilling fluids and circumferential expansion (and corresponding mechanical equilibrium) of the well casings. In the present work, ML algorithms were used to accurately model “fluid density = f(T, P)” based on the experimental PVT data of a given fluid at a range of (T, P) conditions. Sensitivity analysis was performed to demonstrate improvement in precision of APB estimation (for different subsea well scenarios using different fluids) using the ML-basedmodels. This study demonstrates that, in several subsea scenarios, a relatively small error in the experimental fluid PVT data can lead to significant variation in APB estimation. The ML-based models for “density = f(T, P)” for the fluids ensure that the cumulative error during the modeling process is minimized. The use of certain ML-based density models was shown to improve the precision of APB estimation by several hundred psi. This advantage of the ML-based density models could be used to improve the safety factors for APB mitigation, and accordingly, the work may be used to better handle the APB issue in the subsea high-pressure/high-temperature wells.

Machine Learning for Formation Tightness Prediction and Mobility Prediction

From the perspective of wireline formation test (WFT), formation tightness reflects the "speed" of pressure buildup while the pressure test is being conducted. We usually define a pressure test point that has a very low pressure-buildup speed as a tight point. The mobility derived from this kind of pressure point is usually less than 0.01md/cP; otherwise, the pressure points will be defined as valid points with valid formation pressure and mobility. Formation tightness reflects the formation permeability information and can be an indicator to estimate the difficulty of the WFT pumping and sampling operation. Mobility, as compared to permeability, reflects the dynamic supply capacity of the formation. A rapid and good mobility prediction based on petrophysical logging can not only directly provide valid formation productivity but can also evaluate the feasibility of the WFT and doing optimization work in advance. Compared to a time-consuming and costly drillstem test (DST) operation, the WFT is the most efficient and cost-saving method to confirm hydrocarbon presence. However, the success rate of WFT sampling operations in the deep Kuqa formation is less than 50% overall, mostly due to the formation tightness exceeding the capability of the tools. Therefore, a rapid mobility evaluation is necessary to meet WFT feasibility analysis. As companion work to a previous WFT optimization study(SPE-195932-MS), we further studied and discuss the machine learning for mobility prediction. In the previous study, we formed a mobility prediction workflow by doing a statistical analysis of more than 1000 pressure test points with several statistical mathematic methods, such as univariate linear regression (ULR), multivariate linear regression (MLR), neural network regression analysis (NNA), and decision tree classification analysis (DTA) methods. In this paper, the methods and principles of machine learning are expounded. A series of machine learning methods were tested. The algorithms that are appropriate for these specific data set were selected. Includes DTA, discriminant analysis (DA), logistic regression, support vector machine (SVM), K-nearest neighbor (KNN) for formation tightness prediction and linear regression, DTA, SVM, Gaussian process regression SVM, random tree, neural network analysis for mobility prediction. Contrastive analysis reveals that: The SVM classifier has the best result over other methods for formation tightness probability prediction. Based on R squared and RMSE analysis, linear regression, GPR, and NNA delivered relatively good results compared with other mobility prediction methods. An optimized data processing workflow was proposed, and it delivered a better result than the workflow proposed in SPE-195932-MS under the same training and testing dataset condition. The comparison between measured mobility and predicted mobility results reveals that, in most situations, the predicted mobility and measured mobility matched very well with each other. WFT were conducted in newly drilled wells. Sampling success rate also achieved 100% in these wells by optimizing the WFT tool string and sampling stations selection in advance, and NPT is significantly reduced.