Stuck pipe has been recognized as one of the most challenging and costly problems in the oil and gas industry. However, this problem can be treated proactively by predicting it before it occurs. The purpose of this study is to implement the two most powerful machine learning methods, Artificial Neural Networks (ANNs) and Support Vector Machines (SVMs), to predict stuck pipe occurrences. Two developed models for ANNs and SVMs with different scenarios were implemented for prediction purposes. The models were designed and constructed by the MATLAB language. The MATLAB built-in functions of ANNs and SVMs, and the MATLAB interface from the library of support vector machines were applied to compare the results. Furthermore, one database that included mud properties, directional characteristics, and drilling parameters has been assembled for training and testing processes. The study involved classifying stuck pipe incidents into two groups - stuck and non-stuck - and also into three subgroups: differentially stuck, mechanically stuck, and non-stuck. This research has also gone through an optimization process which is vital in machine learning techniques to construct the most practical models. This study demonstrated that both ANNs and SVMs are able to predict stuck pipe occurrences with reasonable accuracy, over 85%. The competitive SVM technique is able to generate generally reliable stuck pipe prediction. Besides, it can be found that SVMs are more convenient than ANNs since they need fewer parameters to be optimized. The constructed models generally apply very well in the areas for which they are built, but may not work for other areas. However, they are important especially when it comes to probability measures. Thus, they can be utilized with real-time data and would represent the results on a log viewer.

No abstract

While it is common knowledge that the values of Archie's parameters ‘a’, ‘m’ and ‘n’ may vary for sandstone reservoirs at different conditions ? consolidated/unconsolidated, water-wet/oil-wet, pore geometry, degree of sphericity, clay content and in-situ depositional environment ? the values of a = 1, m = 2 and n = 2 have been linked historically with Archie's equations. Statistically, these values were assumed as the population mean when used in calculations, but the magnitude of errors associated with their use is often neglected. This study shows how a sample of data drawn from experimental and analytical methods determine Archie's parameters. Well logs from sandstone reservoirs are used to draw statistical inferences about the population characteristics of Archie's parameters in sandstone reservoirs. This also shows the magnitude of relative error possible when m = 2 and n = 2 are used to compute water saturation and formation resistivity factors. Introduction The 1942 landmark publication by Gus Archie titled, "The Electrical Resistivity Log as an Aid in Determining Some Reservoir Characteristics "(1) introduced new parameters relevant to describing reservoirs using well logs and set standard parameters for the identification of permeable zones within a reservoir. Basically, Archie's experiments involved measuring the porosity and electrical resistivity of numerous shale-free sandstone cores from the Gulf Coast by saturating them with brine of salinities ranging from 20 to 100,000 ppm of NaCl. Archie's work established the following relationships: Equation (1) (Available In Full Paper) where Ro is the resistivity of the rock fully saturated with brine and Rw is the formation water resistivity. F is termed the formation resistivity factor and is a measure of the effect of formation on the path of electrical current traveling through the electrolytic brine in the rock pore system. The plot of F against porosity (Φ) on log-log scales revealed a linear trend equivalent to: Equation (2) (Available In Full Paper) The ‘m’ parameter represents the trend's negative slope. In deriving the above expression, Archie force-fitted a line to his formation resistivity factor against porosity data such that F = 1.0 at 100% porosity. A replica of Archie's original plot using clean sandstone cores is shown in Figure 1. This was however deemed unnecessary as other research works revealed that when a line is fitted to formation resistivity factor against porosity, the intercept at 100% porosity would not always derive one, but can be greater or less than one. Winsauer et al.(2), for example, duplicated Archie's experiments with sandstone cores from a wide range of reservoirs and arrived at: Equation (3) (Available In Full Paper) Thus, the general form of Archie's formation resistivity factor is expressed as: Equation (4) (Available In Full Paper) Winsauer first referred to Archie's ‘m’ parameter as the cementation factor, while ‘a’ would later be referred to as the tortuosity factor. Archie, again, considered partially saturated hydrocarbon bearing shale-free sandstones and proposed a second factor called the resistivity index (I) that would further increase the rock's resistivity. He expressed this as: Equation (5) (Available In Full Paper) FIGURE 1: Plot of formation resistivity factor against porosity (Available In Full Paper)

This study achieves two main goals. First, it develops a method that uses the Composite Layering Effect (CLE) Equation to predict the behavior of potential fractures in conventional and unconventional reservoirs from core samples. The second goal of this study is to determine how different mineralogical and elemental components affect the behavior of fractures predicted using the CLE equation. After the samples are fractured, X-ray Powder Diffraction (XRD) and X-ray Fluorescence (XRF) techniques are executed to measure the mineralogical and elemental compositions of the core samples respectively. In this method, core samples are first obtained from the formation. Next, X-Ray Computed Tomography (CT) is used to determine if core samples have preexisting fractures. The samples are then fractured slightly using Uniaxial Compressive Strength (UCS), in which a compressive-strength machine initiates fractures by applying uniaxial load and stopping automatically upon reaching a predetermined load. CT then confirms the existence of the new fractures, and Image J interprets the height-length ratio of each fracture. These results are used in calculating the CLE. The results of these experiments revealed the relationship between the mineral and elemental compositions of the rocks and the crack dimensions. It was seen that the presence of quartz and clay minerals had the strongest influence on the CLE value due to the brittle behavior of the quartz and ductile behavior of the clay minerals (nacrite). The highest CLE value was recorded for the shale sample that had a preexisting fracture. The fracture patterns developed in the shale samples were mainly parallel to one another. In contrast, the fracture patterns developed in sandstones started out parallel and later merged together to form a connected fracture network.

Mature fields lack the sparkle of a new play, and an operator typically will not invest capital for waterflooding, much less EOR. But prevailing higher oil prices can turn such a mature oil field more profitable by employing innovative production enhancement techniques. We propose the use of a simple, tapered tubing string completion (using larger internal diameter (ID) tubing pipes in the upper sections) that can be customized for specific reservoirs. Historically there are few instances of tapered ID tubing completion, which were basically necessitated by technical constraints (liner, workover, etc.). But our approach is focused on enhancing economic performance. We have employed nodal analysis technique to develop an equivalent tubing diameter (ETD) concept. The ETD allows for comparing the well performance for single-ID tubing completion. The procedure also seeks an optimum length for the larger tubing ID in the upper section. Using reservoir simulation for full life cycle, and oil prices projected over time until abandonment, the economic performance is evaluated using NPV and other economic parameters. The proposed production enhancement method is suitable for wells with moderate to high open flow potentials (AOFP). It is especially suited for low GOR wells with high future water-cut that will eventually require an ESP system, and also remote oilfields, where reservoir pressure maintenance and EOR is not viable. The use of larger tubing ID section entails only a marginal increase in CAPEX. However, the tapered completion gives increased production rate sustained over a long time, which results in significant economic gain. The economic benefits accrue from the prevailing high oil price, yielding a quick payout and many returns on investment. The proposed completion approach does not involve complexity, and the innovative application of nodal analysis coupled with high oil prices show how to make mature oil fields onshore and offshore, more profitable. Introduction Nodal analysis was performed in the sixties and seventies by hand calculations, using vertical pressure traverse graphs generated in-house by big oil companies. Smaller operators, if they at all used nodal analysis, relied on Brown's (1963) famous pressure traverse graphs. The workflow was tedious at best, discouraging engineers to explore for horizons that nodal analysis could lead to. However, with the advent of affordable PC software (e.g., Fekete's FAST™, IHS's PERFORM™, etc.), and even MS-Excel™ based spreadsheet programs (e.g., Guo et al. 2007), the power of nodal analysis now can be unleashed even in a classroom setting. One such unexplored horizon is tapered tubing string design - with gradually larger internal diameter (ID) in the upper sections of tubing string. Conventional tubing string design entails selecting a constant internal diameter for all the tubing sections—from bottom to top. The upper sections of the string, however, have a greater wall thickness to support the load of the string below. Thus conventional tubing strings are tapered in terms of outer diameter, which is necessitated by mechanical loading requirements.