Torque and drag [T&D] are critical elements in drilling deep vertical wells with unnoted tortuosity along the borehole and in extended-reach and deviated wells with high and repeated tortuosity. Current T&D models generate unreliable predictions due to the assumption that borehole trajectories are composed of constant curvature arcs between survey stations [the minimum curvature calculation method]. This assumption causes the bending parameter in the T&D equilibrium balance equations to be nil. Today's T&D models are either based on a continuous drillstring to wellbore contact [the soft-string model] or intermittent contact due to drillstring stiffness [the stiff-string model]. In both cases, the wellbore trajectory is based on the minimum curvature method. The T&D model proposed is a phase II initiative of the non-constant curvature trajectory model: the advanced spline curves [ASC] borehole trajectory model published at the 2016 IADC/SPE Drilling Conference and Exhibition (Abughaban el al. 2016). This modified three-dimensional [3D] T&D model [the ASC 3D T&D model] is a stiff-string model that includes geometric torsion, wellbore curvature, change in the rate of wellbore curvature and drillstring bending stiffness in the T&D equilibrium balance equations. The model has been validated using field cases with real-time forces that define T&D measured at the surface. The calculated outputs from these wells provide more accurate view of the drilling conditions downhole, including the downhole weight on bit and torque on bottom. The novelty of the proposed model is the ability to estimate a more realistic bending effects, accurately predict the contact forces between the drillstring and the wellbore, and solve T&D parameters from surface to total depth in reasonable time using standard engineering computer. Accurately estimating these parameters will allow drilling engineers to update the driller with surface weight on bit and torque parameters to improve the drilling performance without taking undue risks with the drilling system such as stuck pipe, casing and drillpipe wear and drillstring fatigue. Thus, the ASC 3D T&D model is not only an alternative approach to accurately model downhole T&D parameters to be used in real-time operation centers [RTOC]; it can also serve as a step toward drilling automation.
As development of the Barents Sea continues with new plays such as the Castberg, accurate specification of the local magnetic field is important to reliably infer the orientation of the bottomhole assembly (BHA) in horizontal drilling. Since magnetic fields at high latitudes vary spatially and temporally, one requires both spatial models and a way to capture temporal changes. Large temporal changes in the magnetic field can severly distort measured azimuths and therefore must be corrected for. This study, based on a report written for Petroleumstilsynet (Maus et al., 2017), shows that in regions of the Barents Sea within 50 km of a magnetic observatory, either the nearest observatory, interpolated infield referencing (IIFR), or the disturbance function (DF) method may be used for corrections in wellbore surveying to meet accuracy requirements. IIFR and DF will give better error reduction but are slightly more complicated to implement. At distances between 50 km and 250 km, the disturbance field (DF) method best meets accuracy requirements. In remote regions beyond 250 km, a local observatory must be deployed to meet the highest accuracy specifications, but the DF will still far outperform the other interpolated methods at such large distances from an existing observatory. Despite having focused on the Barents Sea region, this comparison of the accuracy of different spatial and temporal magnetic field mitigation methods for wellbore surveying is applicable to high latitude northern and southern regions across the globe.
Geosteering solves for a drill bit's stratigraphic location to optimally guide a wellbore through a target formation. Geosteering solutions focus on correlating a subject well's real-time measurements to a type log representative of the stratigraphic column. Traditionally, this is done by matching localized sections of each log with a combination of shifts and stretches applied. This is a single-solution-at-a-time approach where only the best correlation is represented, as determined subjectively by a human or via algorithmic minimization of differences between measurements (Maus et al 2020). A method that considers the full space of possible stratigraphic interpretations and assigns a complexity-related correctness likelihood to each would give geosteerers greater confidence in selecting the correct interpretation. This would be a prohibitively large space to explore via traditional optimization and inversion methods, but it is possible through application of a Viterbi algorithm to a Bayesian state space matrix. A set of 1440 synthetic geosteering trials was generated by producing a geologically realistic layer cake, passing a well trajectory through it, and simulating realistically corrupted gamma measurements (reflecting sampling rate, calibration error, and measurement noise). This gives a true solution for accuracy comparison, and a realistic log that can be interpreted as follows: a Bayesian state space matrix is constructed which captures the likelihood of correlation between subject well and type log measurements. Prior knowledge is used to inform a state-transition probability matrix. The Viterbi algorithm is then applied to the state space matrix and state-transition probability matrix to determine the highest likelihood interpretation. The trial data was split 80/20 into training and test sets. For the training data, three metrics were used to tune the algorithm: mean distance from true solution; misfit ratio against true solution, and algorithm runtime. On the remaining test data, highest-likelihood paths were compared to the interpretations generated by an existing residual-minimizing automated geosteering algorithm (Gee et al 2019). Performance was analyzed separately in the vertical, curve, and lateral, and solutions were spot-checked for reasonable behavior. Compared to the existing automated method, the Bayesian method produced interpretations with comparable performance in 59% of laterals, and significantly improved performance in 34% of laterals. It also returned results 30 times faster. These results held over several sets of tuning parameters suggesting robustness. A well-tuned Bayesian algorithm has been shown to outperform existing automated methods on performance and accuracy, signifying a potential step change in the space of automated geosteering. Viterbi is an established algorithm with many applications, but the splitting of stratigraphic mappings into a Bayesian state space and application of Viterbi is novel and allows for efficient, probabilistic solution-finding. The whole space of possible solutions can be considered, and implicitly gives solution likelihoods. The technique also accounts for the complexity of produced solutions.
Lateral spacing in unconventional plays can have a significant impact in the economics of field development (Bharali et al., 2014; Lalehrokh & Bouma, 2014). This spacing is most often verified using magnetic measurement while drilling (MWD) instruments. In well spacing studies, the distance between two laterals is typically assumed to be precise, however, MWD may have large uncertainties associated with their bottom hole locations (Williamson, 2000; Grindrod et al., 2016). Standard error models were built primarily using data from major service providers' offshore operations and assume a level of accuracy that may not reflect current practices for onshore drilling in North America (Love, 2019). This study better quantifies the positional uncertainties of MWD surveys in long laterals and uses those uncertainties to estimate well spacing uncertainty. More than 35,000 MWD bit runs across more than 9,000 laterals wellbores in major basins across North America have been analyzed for survey errors. The observed errors are then tested in magnitude and distribution against assumptions of industry standard models. An empirical MWD error model for North America is generated, and additional area-specific analysis is performed for the Bakken, Denver-Julesburg, Eagle Ford, Marcellus/Utica, Permian/Delaware and Western Canada. The wellbore trajectories were analyzed to produce a prototypical wellplan for each basin studied, as well as an overall "North America" wellplan. Positional uncertainty calculations were performed for each wellplan using several industry standard positional uncertainty models as well as the empirical models based on real MWD data. Using the combined covariance method of wellbore separation, the uncertainty in lateral spacing is estimated for each wellplan at landing point, mid-lateral, and the toe. A pair of typical North America long laterals (10,000ft step out, drilled parallel) were modeled using the empirically derived MWD error model. After accounting for geomagnetic correlations, there remained an estimate 350 ft of uncertainty in the separations (at 2-sigma) between the wellbores are the toe. This is significantly larger than the 160ft of spacing uncertainty that would be predicted using an industry standard MWD model or the 100ft theoretically achievable with standard survey management practices. A similar analysis was performed at a basin level, and in all cases the modeled uncertainty from common practice was larger than predicted by industry standards. The dominant error source that is impacting lateral spacing is magnetic drillstring interference (DSI). The 68th, 95th, and 99.7th percentiles for observed DSI magnitudes were 520, 1400 and 4200 nanotesla respectively. Creating a 95% confidence equivalent error model (2-sigma) requires and error magnitude more than 3 times greater than the industry standard. For companies that use greater than a 2-sigma limit when well planning, deviations even more extreme should be expected. Further discussion of the probability distribution and its impact on lateral spacing is included. Additional analysis compares how these uncertainties change for parallel wells drilled in opposite directions (anti-parallel) as opposed to drilled in the same direction. MWD error models are routinely used as the basis for both safety critical and economically significant workflows in onshore North America operations, to date, there has been a lack of data collected regarding their suitability for this purpose. For drilling programs where lateral spacing will have an impact on economic performance, proper estimation of the spacing uncertainty will lead to better asset modeling. Further, it will enable better estimation for the marginal value provided by improvements in survey accuracy.
Wellbore trajectories are a fundamental piece of data used for decisions throughout the oilfield. Trajectories are typically mapped through measurement-while-drilling (MWD) survey stations collected at 95ft intervals. Previous work suggests that this sparse sampling interval masks short segments of high curvature, negatively impacting workflows that consume this data (Stockhausen & Lesso, 2003; Baumgartner, et. al., 2019). This can come in the form of poorly estimating the true vertical depth of a well, poorly mapping geologic structure, and poorly quantifying the tortuosity of the wellpath. Several methods have previously been proposed to improve trajectory mapping by incorporating additional data collected between stationary surveys (Stockhausen & Lesso, 2003; Gutiérrez Carrilero, et al., 2018). Two sources of such data are continuous survey measurements and slide/rotate behaviors captured in slide sheets. Two methods of improving the wellbore trajectory mapping were compared in several extended reach lateral wellbores. The impact of the new trajectories on landing point selection, dip estimation, and wellbore tortuosity analysis was determined. One method took continuous inclination data and mapped directional changes between stationary surveys. The second used bit projections generated through automated-slide-sheet-analysis from real-time tool face data, estimating the location and direction of curvature produced by slide/rotate operations. These curvature estimations were used to predict wellbore shape between surveys. As a final check, in the curve sections of the wellbores, stationary surveys were collected at more frequent intervals (e.g., 31ft) to provide validation on the high-resolution trajectories and to understand the cost-benefit of simply surveying more frequently. Both methods of high-resolution trajectories imply that errors present in a 95ft course length survey are enough to impact decisions made when drilling an extended reach lateral. Landing point estimations were shifted in several cases by over 10ft, the approximate thickness of the target formation. Similar discrepancies in true vertical depth were observed along the length of the laterals. Both methods showed strong agreement through the curve sections of the wellbore, however this agreement weakened during the lateral where short slides and geological effects on rotary tendency reduced the accuracy of the automated-slide-sheet method. A discussion of the discrepancies between the two methods in laterals is included. Dogleg severity analysis confirmed that short sections of high curvature exist that are masked by traditional 95ft survey course lengths. Surveying at 31ft intervals improves the mapping of this tortuosity but still does not capture the full effects seen on continuous survey data. Previous work has suggested that typical wellbore trajectory mapping may be unsuitable for accurate analysis of things like geological structure and wellbore tortuosity analysis. Two methods are evaluated here that support those claims, suggesting that in the future high-resolution trajectories may be a necessity for accurate decision-making.
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.