|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
Proactive geosteering workflows include a number of steps that are repeated as the drilling operation progresses. First, various measurements such as gamma ray, gravity or electromagnetic (EM) data are collected while drilling. Second, the recorded data samples are processed and used to update the geomodel, including parameters around the well relevant to the steering (e.g., reservoir boundaries, faults, geophysical properties, fluid contacts, etc.). Finally, geosteering decisions are made based on the updated geomodel, other available knowledge and operational constraints while drilling. In many situations, the inversion or interpretation procedure used to update the existing geomodel provides only a single admissible solution, while the uncertainty is not quantified. As a consequence, when put together with other constraints that control the placement of the well, the decision making process might be biased, increasing the risk of taking poor decisions.
An alternative to the conventional deterministic inversion methods is the ensemble-based inversion algorithms (for instance, the ensemble Kalman filter), which have been widely applied in various disciplines such as meteorology, oceanography, hydrology and reservoir engineering in the last decade, and are praised for their satisfactory performance and ability to quantify the uncertainty. In this work, we propose an ensemble-based framework that uses available logging while drilling measurements for continuously updating the geomodel and optimizing the placement of the remaining well path under uncertainty. Deep EM measurements are chosen as observed data for this study because they combine good range and reliability for the look around and are readily available in many drilling operations. Furthermore, a 3D finite difference EM modelling tool, capable of taking into account complex reservoir geometries, is used to solve the forward problem. The proposed framework is tested on both simple and more realistic synthetic cases. The obtained results suggest that the ensemble-based methodology can match the synthetic truth in a probabilistic sense. The subsequent well placement is optimized in a robust way based on these estimations, and achieves good coverages of the reservoir zones.
Abstract In this work we present a systematic geosteering workflow that automatically integrates a priori information and the real-time measurements for updating of geomodel with uncertainties, and uses the latest model predictions in a Decision Support System (DSS). The DSS supports geosteering decisions by evaluating production potential versus drilling and completion risks. In our workflow, the uncertainty in the geological interpretation around the well is represented via multiple realizations of the geology. The realizations are updated using EnKF (Ensemble Kalman Filter) in real-time when new LWD measurements become available, providing a modified prediction of the geology ahead of the bit. For every geosteering decision, the most recent representation of the geological uncertainty is used as input for the DSS. It suggests steering correction or stopping, considering complete well trajectories ahead-of-the-bit against the always updated representation of key uncertainties. The optimized well trajectories and the uncertainties are presented to the users of the DSS via a GUI. This interface enables interactive adjustment of decision criteria and constraints, which are applied in a matter of seconds using advanced dynamic programming algorithms yielding consistently updated decision suggestions. To illustrate the benefits of the DSS, we consider synthetic cases for which we demonstrate the model updating and the decision recommendations. The DSS is particularly advantageous for unbiased high-quality decision making when navigating in complex reservoirs with several potential targets and significant interpretation uncertainty. The initial results demonstrate statistically optimal landing and navigating of the well in such a complex reservoir. Furthermore, the capability to adjust and re-weight the objectives provides the geosteering team with the ability to change the selected trade-offs between the objectives as they drill. Under challenging conditions, model-based results as input to a decision process that is traditionally much based on human intuition and judgement is expected to yield superior decisions. The novel DSS offers a new paradigm for geosteering where the geosteering experts control the input to the DSS by choosing decision criteria. At the same time, the DSS identifies the optimal decisions through multi-objective optimization under uncertainty. It bridges the gap between developments in formation evaluation and reservoir mapping on one side, and automation of the drilling process on the other. Hence, the approach creates value based on the existing instrumentation and technology.
ABSTRACT Borehole measurements (resistivity, nuclear properties, and acoustic slowness) are affected by several factors specific to the design of logging instruments. These effects can be corrected using fast numerical simulations of well logs and inversion algorithms, thereby improving the estimation of mineral/fluid concentrations. Gradient-based inversion yields acceptable results; however, the calculation of derivatives is difficult without explicit information of the tool/instrument properties. Bayesian methods can also be useful, but they are computationally expensive, requiring ∼10,000 times of simulations, which translates into hours if not days of CPU time. We develop an efficient Bayesian method for generalized well-log inversion (petrophysical and compositional). First, we estimate layer-by-layer properties combining a quasi-Newton method with Markov chain Monte Carlo (QNMCMC) sampling. Derivatives are approximated by the difference between the target field log and its numerical simulation. Next, uncertainty of inversion results is determined using Gaussian distributions. Estimated layer-by-layer properties and their uncertainties are used to estimate compositional properties using MCMC sampling accelerated by a surrogate model constructed with radial basis functions (RBF). We verify the applicability of the proposed method with both synthetic and field data. Results from the synthetic example indicate that the proposed method is reliable and efficient under extreme conditions (large property contrasts and thin layers). The application to field measurements yields 75% of core data within the 95% credible interval of interpreted results. Comparison to traditional random-walk MCMC (RWMCMC) indicates that the computational time for separate and joint inversion is reduced by a factor of 30 and 100, respectively. In addition to its speed, robustness, and reliability, the Bayesian inversion method provides great flexibility to include user-defined correlations among solid components and mineral groups. It can also perform rock-class-dependent detection based on Gaussian Mixture Models (GMM) with minimum user intervention, thereby successfully competing with commercial solvers. INTRODUCTION In reservoirs with complex lithology, successful assessment of porosity (ϕ) and water saturation (Sw) requires the accurate estimation of mineral concentrations. Petrophysical interpretation via commercial software uses linear joint inversion to calculate volumetric concentrations of solid and fluid components in the formation (Quirein et al., 1986). However, there are several disadvantages to the latter method. First, it ignores the nonlinear mixing behavior of neutron porosity and resistivity. Second, it assumes that well logs represent true formation properties. These assumptions lead to inaccurate petrophysical interpretation in thinly-laminated formations because of shoulder-bed effects. Third, it is difficult to determine the reliability of inversion results without properly quantifying uncertainty.
Abstract At the well-planning stage target selection usually accounts for drillability. However, during geosteering operations the drilling constraints are not updated and some fixed limits in terms of maximal inclination, dogleg severity, etc., are used instead. We demonstrate a methodology that uses fast physical models of the drilling hydraulics to calculate constraints and costs for geosteering dynamically during an operation. In field development, many companies have adopted workflows that use ensemble-based methods for decision support. A real-time variation of such a decision support system (DSS) has been recently proposed for geosteering. The DSS is capable of optimization full well trajectories across all realizations of the earth model and can consider multiple objectives and constraints simultaneously. We present a method that makes steady-state hydraulic computations for all possible trajectories ahead-of-bit simultaneously at a low added cost. The output of the computation can provide more precise constraints (geo-pressure margins and cuttings transport) and cost estimates for the DSS. In this paper we focus on verification and testing of the proposed multi-trajectory hydraulic model (MTHM). Discretization of the model acts as a trade-off between the preciseness of the computation and the computational speed. On our benchmark cases, a simulation that computes the hydraulic parameters for all trajectories with acceptable errors is fast enough for real-time geo-steering applications. Furthermore, we present a case based on data from the Norwegian Continental Shelf for which we demonstrate how hydraulic computations would influence the decisions of steering and stopping. Applying the DSS with the MTHM allows to precisely update the allowed steering interval, thus achieving safe operation while maximizing the expected well profit. We emphasize that integration of the drilling processes modelling as part of the decision support for the geosteering operation enables better decisions. This is facilitated by the digitalization of the oil industry, but still requires development of new approximate models of the drilling processes. This paper demonstrates the MTHM as an initial step towards integration of drilling and geosteering modelling.
Abstract Innovative interpretation methods based on efficient measurements to map structure and fluids around high angle and horizontal wells while drilling are critical for future success in a marginal, but increasingly strategical, business on the Norwegian Continental Shelf (NCS). A high number of infill wells are expected to prolong the operational life of assets for the coming decades. Success of horizontal infill wells targeting bypassed zones are challenged by uncertainties in the subsurface description and by fluid content variation generated by differences in sweep efficiency over time. Optimizing well placement of such producer wells has a direct impact on cost and recovery. This could also potentially unlock targets not accessible today with currently used methods and technologies. Driven by the advanced integration of subsurface information, the capacity of 3D mapping and characterization of the reservoir features has reached an unprecedented understanding level. Through the incorporation of data from real time LWD measurements, 1D and 2D ultra-deep azimuthal resistivity inversions, all the way to 3D and 4D surface seismic, it is now possible to describe in detail not just the structural and stratigraphic components but also the reservoir properties and fluid distribution along horizontal wells. The key challenge is to establish more efficient workflows for advanced integration of subsurface information through new digital solutions transforming future geosteering and well placement solutions with 3D mapping and characterization real-time. Key elements of the integrated workflows and potential benefits in all phases of a well placement job will be discussed in this paper with examples from real cases from NCS.