Assessment of uncertainty of borehole resistivity measurements is important to quantify the accuracy of hydrocarbon reserves and production forecasts. We develop an efficient Bayesian inversion method for the quantitative interpretation of general borehole resistivity measurements. It enables the estimation of resistivity properties together with their uncertainties in conjunction with arbitrary sets of
Significant progress has been made with regard to the quantitative integration of geophysical and hydrological data at the local scale. However, extending the corresponding approaches to the regional scale represents a major, and as-of-yet largely unresolved, challenge. To address this problem, we have developed an upscaling procedure based on a Bayesian sequential simulation approach. This method is then applied to the stochastic integration of low-resolution, regional-scale electrical resistivity tomography (ERT) data in combination with high-resolution, local-scale downhole measurements of the hydraulic and electrical conductivities. Finally, the overall viability of this upscaling approach is tested and verified by performing and comparing flow and transport simulation through the original and the upscaled hydraulic conductivity fields. Our results indicate that the proposed procedure does indeed allow for obtaining remarkably faithful estimates of the regional-scale hydraulic conductivity structure and correspondingly reliable predictions of the transport characteristics over relatively long distances.
Gonzalez, Ezequiel F. (Shell International Exploration and Production) | Gesbert, Stephane (Shell International Exploration and Production) | Hofmann, Ronny (Shell International Exploration and Production)
Using inverted seismic data from a turbidite depositional environment, we show that accounting only for rock types sampled at the wells can lead to biased predictions of the reservoir fluids. The seismic data consists of two volumes resulting from a simultaneous (multi-offset) sparse-spike inversion. As is common in an exploration setting, information from a single well (well logs and petrological analysis) was used to define an initial set of discrete “facies” that characterize both rock type and saturating fluid. Based on our geological understanding of the study area, we augmented this initial model with facies expected in the given depositional environment, yet not sampled by the well. Specifically, the new facies account for variations in both mixture type and proportions of shales and sands. The elastic property distributions of the new facies were modelled using appropriate rock physics models. Finally, a geologically consistent, spatially variant, prior probability of facies occurrence was combined with the data likelihood (per facies) to yield a Bayesian estimation of facies probability at every sample of the inverted seismic data. Accounting for the augmented geological prior in this way, we were able to generate a scenario consistent with all available data, which supports further development of the field. In contrast, using the initial, purely data-driven facies model, Bayesian classification leads to downgrading of the field''s prospectivity. We argue that limited well control in Quantitative Interpretation, especially in an exploration setting, needs to be counterweighted by robust geological prior information, in order to unbiasedly risk geological scenarios.
Miranda, T. (University of Minho) | Ribeiroe, Sousa L. (Iowa State University, Department of Civil, Construction & Environmental Engineering, Iowa State University) | Gomes, Correia A. (University of Minho)
The evaluation of geomechanical parameters for rock masses is one of the issues with largest uncertainty degree. This is mostly true in the preliminary stages of design and in works where geotechnical information is scarce. Data Mining techniques have been successfully used in many fields but scarcely in geotechnics. They are advanced techniques which allow analyzing large and complex databases like the ones it is possible to build with geotechnical information. In this work, a large database of geotechnical data produced in the scope of a large underground structure was gathered and these innovative tools were used to analyze it and induce new and useful knowledge. The main goal was to develop new and reliable models to predict geomechanical parameters for rock masses, namely friction angle, cohesion and deformability modulus, when only limited data is available.
The evaluation of geomechanical parameters in underground works corresponding to the preliminary stages of design is normally performed based on scarce and uncertain data. When a small amount of data is available, geomechanical information concerning other works, developed in similar rock masses, can help in defining values for the parameters. The number and type of tests performed in geotechnical site investigation is related to the importance of the work, the inherent risk and budget issues. In geotechnical works where the available geological-geotechnical data is limited, the geomechanical parameters are set based on the available data and conservative engineering judgment. In these cases, great amounts of geotechnical data produced in large projects could help in reducing uncertainties related to the definition of design values for the parameters. Therefore, the advantages of using geotechnical data gathered from several different projects are indubitable. However, this is not a straightforward process. The central question is how vast quantities of data with different types and origins can be managed and explored in order to develop models that can provide a background for future projects. One first step for solving this problem is defining standard ways of collection, organisation, and representation of data. In terms of techniques to analyse these databases and repositories, currently there are automatic tools from the fields of artificial intelligence and pattern recognition, for instance, which allow a deeper understanding of large and complex databases enabling to explore and discover potential embedded knowledge . It is believed that the automated tools of data analysis like Data Mining (DM) can help in developing complex "data-driven" models. The formal analysis process, normally called Knowledge Discovery in Databases (KDD), defines the main procedures for transforming raw data into useful knowledge. DM is just one step in the KDD process concerned with the application of algorithms to the data in order to obtain models. These tools allow a deep analysis of complex data, (i.e. data with a large number of variables, independent determinations and complex and unclear relations with other variables) which would otherwise be very difficult using classical statistics or using only one or even a panel of human experts, who could overlook important details. However, the computational process cannot substitute human experts.
Wang, Grace S. (Department of Construction Engineering Chaoyang University of Technology) | Huang, Chien-Lin. (Department of Construction Engineering Chaoyang University of Technology) | Huang, Fu-Kuo (Department of Water Resources and Environmental Engineering Tamkang University)
The decline of production of conventional reservoirs and increase in demand of hydrocarbon fuels bring the tight gas into the forefront of the energy future. This unconventional energy source is a fast-growing market and is recognised as having huge future potential for production worldwide. Low permeability of tight gas shale requires conducting effective hydraulic fracturing and applying horizontal drilling technologies to produce at commercial level. Therefore, successful production from such complex reservoirs is heavily dependent on the selection of an appropriate completion technology requiring sufficient knowledge of borehole shape. Therefore, understanding the rock properties and the earth's stresses (Geomechanical Model) is the first critical step in reservoir evaluation and ultimately development of these resources. In order to determine in-situ stress condition and calibrate geomechanical models, having information of borehole breakouts existence plays a key role.
Caliper and image logs, the regularly used methods to identify breakout zones, have several limitations such as not having a full coverage of the borehole wall, being affected by different parameters, pad width limitation, low resolution, and complicated processing procedure. In addition, good quality image logs are not usually available for shaly formations because of the requirement of using oil-based mud.
This paper presents a new multi-variant approach to identify borehole breakouts in tight gas shale using some petrophysical logs. This approach employs number of data processing techniques, including Bayesian classification, wavelet decomposition and data fusion, to determine borehole intervals with maximum likelihood of enlargement. The results approved the applicability and the generalization capability of the approach with a significant accuracy. Case study on Burnett shale is presented in this paper.
Current downhole measuring technologies have provided a means of acquiring downhole measurements of pressure, temperature, and sometimes flow-rate data. Jointly interpreting all three measurements provides a way to overcome data limitations that are associated with interpreting only two measurements--pressure and flow-rate data--as is currently done in pressure-transient analysis. This work shows how temperature measurements can be used to improve estimations in situations where lack of sufficient pressure or flow-rate data makes parameter estimation difficult or impossible.
The model that describes the temperature distribution in the reservoir lends itself to quasilinear approximations. This makes the model a candidate for Bayesian inversion. The model that describes the pressure distribution for a multirate flow system is also linear and a candidate for Bayesian inversion. These two conditions were exploited in this work to present a way to cointerpret pressure and temperature signals from a reservoir.
Specifically, the Bayesian methods were applied to the deconvolution of both pressure and temperature measurements. The deconvolution of the temperature measurements yielded a vector that is linearly related to the average flow-rate from the reservoir and, hence, could be used for flow-rate estimation, especially in situations in which flow-rate measurements are unavailable or unreliable. This flow rate was shown to be sufficient for a first estimation of the pressure kernel in the pressure-deconvolution problem.
When the appropriate regularization parameters are chosen, the Bayesian methods can be used to suppress fluctuations and noise in measurements while maintaining sufficient resolution of the estimates. This is the point of the application of the method to data denoising. In addition, because Bayesian statistics represent a state of knowledge, it is easier to incorporate certain information, such as breakpoints, that may help improve the structure of the estimates. The methods also lend themselves to formulations that make possible the estimation of initial properties, such as initial pressures.
Autonomous underwater vehicles (AUVs) have a role to play in several phases of offshore exploration and production in the Arctic. They have the potential to enable year-round data gathering under sea ice. However, this potential can only be realized if their reliability is sufficiently high and sufficiently well established. This paper describes a Risk Management Process-AUV that has been developed to assess (a) the probability of losing an AUV and (b) the availability of an AUV, that is, "How likely is it that, when required, the AUV is ready to begin its operations??? Assessing these probabilities requires bringing together knowledge of vehicle faults and incidents, a body of knowledge on the operating environment, and how the vehicle and environment interact. The tools necessary to make these assessments are described. Examples are given of how they can be used by owners and operators to provide a clear and traceable derivation of how the risks have been estimated. These examples draw upon risk assessments for the Autosub3 AUV in Polar Regions.