Summary As the exploration of the South China Sea continues into deeper water, the chances of encountering nonhydrocarbon gas (CO2, N2, etc.) reservoirs rise. The question of how to avoid the risks associated with the discovery of nonhydrocarbon gas reservoirs becomes an issue for deepwater (DW) oil and gas exploration. Geomicrobial hydrocarbon detection (GMHD) is a nonseismic hydrocarbon detection technology that is able to predict the hydrocarbon potential of a prospective area at depth. This is accomplished via the detection of specific hydrocarbon-oxidizing bacteria in both onshore soils and offshore sea bottom sediment samples. The effectiveness of this method has been proved repeatedly in DW explorations of the northern South China Sea. It documents a possible solution to nonhydrocarbon gas risk prediction by combining the oil and gas prediction results of geomicrobial hydrocarbon detection with results from geological and geophysical studies to analyze the different microbial responses above nonhydrocarbon gas and hydrocarbon gas reservoirs. This was verified in the DW exploration practices undertaken in the Pearl River Mouth Basin and Qiongdongnan Basin of South China Sea.

Summary Model-based reservoir management workflows rely on the ability to generate predictions for large numbers of model and decision scenarios. When suitable simulators or models are not available or cannot be evaluated in a sufficiently short time frame, surrogate modeling techniques can be used instead. In the first part of this paper, we describe extensions of a recently developed open-source framework for creating and training flow network surrogate models, called FlowNet. In particular, we discuss functionality to reproduce historical well rates for wells with arbitrary trajectories, multiple perforated sections, and changing well type or injection phase, as one may encounter in large and complex fields with a long history. Furthermore, we discuss strategies for the placement of additional network nodes in the presence of flow barriers. Despite their flexibility and speed, the applicability of flow network models is limited to phenomena that can be simulated with available numerical simulators. Prediction of poorly understood physics, such as reservoir souring, may require a more data-driven approach. We discuss an extension of the FlowNet framework with a machine learning (ML) proxy for the purpose of generating predictions of H2S production rates. The combined data-physics proxy is trained on historical liquid volume rates, seawater fractions, and H2S production data from a real North Sea oil and gas field, and is then used to generate predictions of H2S production. Several experiments are presented in which the data source, data type, and length of the history are varied. Results indicate that, given a sufficient number of training data, FlowNet is able to produce reliable predictions of conventional oilfield quantities. An experiment performed with the ML proxy suggests that, at least for some production wells, useful predictions of H2S production can be obtained much faster and at much lower computational cost and complexity than would be possible with high-fidelity models. Finally, we discuss some of the current limitations of the approach and options to address them.

Summary By 2026, USD 5.05 billion will be spent per year on logging while drilling (LWD) according to the market report from Fortune Business Insights (2020). Logging tools and wireline tools are costly services for operators to pay for, and the companies providing the services also have a high cost of service delivery. They are, however, an essential service for drilling wells efficiently. The ability to computationally generate logs in real time using known relationships between the rock formations and drilling parameters, namely, rate of penetration (ROP), rotations per minute (RPM), surface weight on bit (SWOB), surface torque (TQX), standpipe pressure (SPPA), and hookload (HKLD), provides an alternative method to generate formation evaluation information (analysis of the subsurface formation characteristics such as lithology, porosity, permeability, and saturation). This paper describes an approach to creating a digital formation evaluation log generator using a novel physics-informed machine learning (PIML) approach that combines physics-based approaches with machine learning (ML) algorithms. The designed approach consists of blocks that calculate mechanical specific energy (MSE), physical estimates of gamma ray (GR) using physical and empirical models, and formation information. All this information and the drilling parameters are used to build a classification model to predict the formations, followed by formation-based regression models to get the final estimate of GR log. The designed PIML approach learns the relationships between drilling parameters and the GR logs using the data from the offset wells. The decomposition of the model into multiple stages enables the model to learn the relationship between drilling parameters data and formation evaluation data. It makes it easier for the model to generate GR measurements consistent with the rock formations of the subject well being drilled. Because the computationally generated GR by the model is not just dependent on the relationships between drilling parameters and GR logs, this model is also generalizable and capable of being deployed into the application with only retraining on the offset wells and no change in the model structure or complexity. For this paper, the drilling of the horizontal section will not be discussed, as this was done as a separate body of work. Historically collected data from the US Land Permian Basin wells are used as the primary data set for this work. Results from the experiments based on the data collected from five different wells have been presented. Leave-one-out validation for each of the wells was performed. In the leave-one-out validation process, four of the wells represent the set of offset wells and the remaining one becomes the subject well. The same process is repeated for each of the wells as they are in turn defined as a subject well. Results show that the framework can infer and generate logs such as GR logs in real time. The average root-mean-squared error (RMSE) observed from the experiments is 27.25 api, representing about 10% error. This error value is calculated based on the mean estimate and does not consider the predicted confidence interval. Considering the confidence interval helps further reduce the error margin.

Summary The development and operation of geothermal plants play a crucial role in the transition to sustainable and low-carbon energy systems. In this paper, we have presented a seamless and flexible pore-to-process digital solution for the design and assessment of geothermal systems, encompassing the geothermal reservoir, gathering network, and geothermal power plant. Our primary focus in this study centers on the geothermal power plant with a detailed analysis of the functionality and performance of two commonly used configurations—a single-flash power plant and a double-flash geothermal power plant. Our work highlights that overall exergy efficiency of the studied geothermal power plants declines over time, primarily due to a decrease in the quality of the geothermal reservoir. Additionally, our analysis demonstrated that variations in the inlet separator pressure have a notable impact on the overall behavior of the power plant. Parametric studies also reveal that increasing the inlet separator pressure leads to decreased overall exergy efficiency and turbine power, resulting from less efficient conversion of available exergy into useful work. Our studies showed that a substantial portion of the available exergy in the geothermal fluid is being dissipated in the condenser. Consequently, optimizing the design and operation of the condenser emerges as a crucial factor in enhancing the overall efficiency of geothermal power plants.

Summary United Arab Emirates (UAE) is seeking to become self-sufficient in gas supply by 2030. This has led the country to initiate several exploratory and appraisal projects to achieve this goal. This study covers one such pilot project targeting production from tight gas reservoirs in three wells through a coiled-tubing (CT) underbalanced drilling (UBD) project in ADNOC Onshore. CT pressure control equipment (PCE) was rigged up on top of production trees with wells already completed and cemented. A CT tower was used to accommodate the drilling bottomhole assembly (BHA) and eliminate risks related to its deployment. CT strings were designed to reach target intervals with sufficient weight on bit (WOB), suitable for sour environments, and able to withstand high pumping rates with mild circulating pressures. To address the hazards of H2S handling at the surface, a custom-fit closed-loop system was deployed. The recovered water was treated on the surface and reused for drilling to decrease the water consumption throughout the operations. The plan was to drill 3 3/4-in. horizontal laterals in all candidate wells. Each well was completed with a combination of a 4 1/2-in. and a 5 1/2-in. tubing and a 7-in. liner. Five laterals were drilled across the three candidate wells targeting carbonate reservoirs with each lateral having an average length of ~4,000 ft. The achieved rates of penetration varied significantly from 15 ft/min to 30 ft/min while drilling through the various formations. Over the course of the pilot project, several challenges had to be addressed, such as material accretion on the CT string during wiper trips, treatment of return fluids having high H2S content and rock cuttings, and ensuring the integrity of the CT pipe while operating in severe downhole environments. Solutions and lessons learned from each well were implemented subsequently in the campaign, such as the use of increased concentrations of H2S inhibitor to coat the CT string, the use of nitrified fluids based on changing well parameters to maintain underbalance, thorough pipe management through real-time CT inspection, and adding a fixed quantity of fresh water to the drilling system every day to avoid chemical reactions between the drilling fluid additives and hydrocarbons. The wells completed with this method exceeded production expectations by 35–50% across the project while reconfirming the value of the technology. The use of CT for UBD is still considered a challenging intervention worldwide. Such cases in high H2S environments are rare. This study outlines best practices for a CT UBD and a setup that can be replicated in other locations to implement this methodology with high H2S and when rig sourcing is a concern.

Summary This paper presents an inflow model for partially completed (PC) horizontal wells that is validated against a simulation benchmark with performance evaluated in terms of a normalized dimensionless productivity index (PI). The model is based on a pressure-averaged uniform flux line-source solution that builds upon the model proposed in 1991 by Goode and Wilkinson (GW) and is validated against a numerical simulation-based benchmark. The model is then subjected to a series of tests where specific inputs are allowed to diverge from the assumptions used to derive it, effectively a form of “stress test.” The rationale was to determine model accuracy when applied to wells that do not conform to the idealized well used to develop the model. Inputs “stress tested” include well inclination, asymmetric completion placement, anisotropy, and a significantly larger number of (shorter) completions, up to around 100 (compared to just three in the original GW model). In all cases, the model performed reasonably well, with divergence from the numerical benchmark quantified. However, for all tests, it was found that model accuracy was dependent on the open fraction, with accuracy increasing as the open fraction increased while model reliability became suspect when the open fraction was ≲ 0.3. Practical usage guidelines and limitations are presented along with associated input ranges and estimated divergence from benchmark. The model is thought suitable for inclined and/or undulating wells with a large number of completions—both long and short—with or without asymmetric placement in presence of anisotropy.