|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
An innovative approach has been used to model flow through discrete fracture networks in a massive carbonate reservoir in order to understand and predict performance of vertical and horizontal well completions. This approach focuses on completion effectiveness and the influences that fractures have in a three-phase gravity influenced flow system. The model is set up as a dual porosity, dual permeability simulation of a discrete fracture network of high permeability grid blocks capable of modeling three-phase flow. This model reveals the dominant factors controlling well life cycle performance demonstrated in the Yates Field Unit.
Natural fracture networks dominate flow throughout the reservoir with added economic significance to completion efficiency. Therefore 3D discrete fracture network (DFN) models based on connected-fracture orientation from FMI logs and flow surveys have been used as a basis for constructing the 3-phase simulation grid. The differences in mobility between the three phases result in abnormally shaped gas-oil and water-oil contacts as drawdown is applied. As the fracture oil column depletes, oil mobility reduces with the decrease in effective fracture connection to the outlying oil column. This loss of oil mobility through phase dis-connection in flow conduits has not been the focus of prior studies. The simulator has successfully generated production profiles similar to those observed in field performance data. This wellbore simulation has been used to develop a strategy for optimal completion performance and placement.
Many reservoirs throughout the world reside as unconfined oil columns in massive naturally fractured carbonate formations. Field management may encourage the oil column to either move up or down in the reservoir. Engineers have typically designed completions that maximized oil production while minimizing associated fluids under a homogenous reservoir assumption. There are two things that amplify the complexity of this process for the engineer. The first is understanding the homogenous nature of the reservoir and the complex flow connections within the reservoir. The second thing that needs to be understood is how the reservoir's heterogeneous nature effects fluid flow from the reservoir into the wellbore.
Flow feature identification in a massive carbonate reservoir as described by Fitzsimmons
This treatment will incorporate discrete fracture network concepts into the near wellbore simulation. This approach focuses on the completion's effectiveness at recovering oil. Value maximization will be targeted through high oil rate completions with minimum associated fluid cost per barrel of oil. This is achieved by focusing on the DFN impact on the well's production and the local oil column shape.
Abstract When placing two planar hydraulic fractures in close proximity, simultaneously, within a formation, two factors immediately come to mind. One is that the two growing hydraulic fractures must share the injected fracturing fluid, and the other is that one hydraulic fracture will tend to resist the opening of the other, making their hydraulic fracture growth almost impossible and leading to an imminent screenout when proppants are introduced. On the other hand, taking advantage of the stress changes caused by a first hydraulic fracture and placing another fracture very soon thereafter can be effective—provided the fracture extension is in a totally different direction and is substantial. Otherwise, the fractures will later compete for producing hydrocarbons from the same area within the reservoir. Short, narrow hydraulic fractures tend to act in the linear elastic domain of hard-rock formations. Even though such hydraulic fractures cause stress modification in nearby areas, the hydraulic fracture collapses as soon as pressure is released, and modification of stresses will be limited to the displacement from the propped open thickness between the hydraulic fracture faces. The effective hydraulic fracture redirection distance of a later fracture will be small; hence the production improvement contribution of the later fracture will also be small. This paper describes a hydraulic fracturing stimulation method that capitalizes on the heterogeneous qualities of deposited layers of formation rock, which allows them to slip between weaker micro-layers, thus replicating a weak plastic behavior of the rock deformed by the crack opening during hydraulic fracturing. The method emphasizes the importance of the time factor for maximizing the effects of the pseudo-plastic qualities of hard rock for extended distances, and hence maximizing the orthogonal reach of an immediately created second hydraulic fracture for maximum production enhancement. Introduction The alteration of stresses in a formation, whether done for a purpose or as a consequence of other actions, often has been discussed in the industry (Moschovidis et al. 2000; Warpinski and Branagan 1989). Moschovidis et al. (2000) discussed the directional change of stresses by means of production. In this situation, after the well has produced for a while, depleted sectors near the hydraulic fracture face would cause the subsequent fracture direction to shift a few degrees away from the previous direction. And, although some have claimed the analytical ability to compute this angle, the solution is suspect, because formations are generally heterogeneous; meaning it is almost impossible to predict the true depletion scheme accurately. The error in the computation would most probably be larger than the angle itself. When modifying the fracture direction for a purpose, it is generally done to lead the subsequent hydraulic fractures toward much more prolific producing areas in the formation (Warpinski and Branagan 1989; Surjaatmadja 2007a, 2007b, and 2007c). Should this be the case, it must be understood that there are three basic driving forces for the reorientation of the hydraulic fracture direction:Elastic movement of the fracture faces. Pressurization of the matrix around the fracture. Plastic movement of the fracture faces. As will be discussed, it is quite easy to understand that the first two effects will cause very temporary effects to regions near the first hydraulic fracture. On the other hand, pure plastic movement is very slow, and its effects may not be detectable within a reasonable time period. This paper focuses instead on pseudo-plasticity—a feature that is identifiable in subterranean formations and that can be considered as plastic deformation. Practical use of this technology is also presented.
The design of a hydraulic fracturing treatment typically requires using a computational model that provides rapid results. One such possibility is to use the so-called classical pseudo-3D (P3D) model with symmetric stress barriers. Unfortunately, the original P3D model is unable to capture effects associated with fracture toughness in the lateral direction due to the fact that the assumption of plane-strain (or local) elasticity is used. On the other hand, a recently developed enhanced P3D model utilizes full elastic interactions and is capable of incorporating either toughness or viscous regimes of propagation by using the corresponding asymptotic solution at the tip element. Since either the viscous or toughness asymptote is used, the intermediate regime is not described accurately. To deal with this problem, this study aims to implement the intermediate asymptotic solution into the enhanced P3D model. To assess the level of accuracy, the results are compared to a reference solution. The latter reference solution is calculated numerically using a fully planar hydraulic fracturing simulator (Implicit Level Set Algorithm (ILSA)), which also incorporates the asymptotic solution for tip elements that captures the transition from viscous to toughness regime.
Hydraulic fracturing (HF) plays a crucial role in the petroleum industry, as it allows one to perform reservoir stimulation and intensify hydrocarbon production . To design a HF treatment, an appropriate HF model needs to be utilized. The simplest model is the onedimensional Khristianovich-Zheltov-Geertsma-De Klerk (KGD) model , in which the fracture propagates in a plane, the elastic interactions are modelled assuming that plane strain conditions prevail, and the coupling between viscous fluid flow and elasticity is included. To represent the fracture geometry more realistically, the Perkins-Kern- Nordgren (PKN) model [3, 4] was developed to predict fracture propagation in a horizontally layered medium. The PKN model assumes that the fracture height is always equal to the thickness of the reservoir layer, the fracture opening in each vertical cross-section is taken to be elliptic, while the fluid pressure is calculated assuming that a plane strain condition holds in each cross-section. Given the fact that the PKN model does not allow for the height growth, the pseudo-3D (P3D) model, which permits height growth, has been developed . Later, with the increase of the computational power, more accurate planar 3D models (PL3D) were developed [7, 8]. As follows from the name, the fracture is contained in one plane, where the fracture geometry within this plane is discretized using a two-dimensional grid. Since the KGD, PKN and P3D are essentially one-dimensional models, while all varieties of PL3D are two-dimensional, the CPU time increases dramatically. The PL3D models improve accuracy and open the possibility of capturing different fracture geometries. Recently, researchers have shifted their effort to investigate the interaction between multiple hydraulic fractures that are growing simultaneously , and to describe non-planar fracture propagation .
Abstract The dimensionless parameter “skin” is commonly used to describe actual well performance with reference to an ideal case, and a number of analytical correlations have been derived over the years to account for the effect on well productivity caused by well geometry, completion design, formation damage and production rate. In a previous publication (Byrne and McPhee, 2012) we outlined the case for reducing the industry's reliance on analytical skin factors for well performance evaluation, particularly since the various “components” of skin, or “pseudo skins,” are often lumped together, and their interaction and interrelationship are often confused. The concept of well design without recourse to a “total” or lumped skin parameter merits further description. For example, whilst hydraulic fractures may present an opportunity for significant enhancement in well performance, the fact remains that the fractured well itself, while an improvement on the unfractured case, may still not be producing at an optimum rate, especially if there is damage in the formation, at the fracture face or in the proppant pack. An overall total “negative skin,” which arises naturally when characterising the fractured well productivity with reference to an ideal vertical, unfractured well, can actually mask the true well potential. There is a tendency to believe that performance has been optimised but this is simply not the case. If the geometry of the wellbore, including the completion and induced fractures, are relatively well understood and can be described then the well productivity can be modelled based on the system architecture without the need to compare it to some hypothetical simpler geometry. The quantity and location of formation damage can then be imposed on the model to determine the impact of damage in the fracture, at the fracture face and deeper within the formation. This paper will illustrate how an indiscriminate use of skin factors can lead to sub-optimal well design and resulting performance prediction, and how this can be improved by representing the real system geometry as rigorously as possible. Decoupling formation damage from its association with other pseudo skins enables a proper description of the impact of damage and its contribution to well performance.