Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Results
ABSTRACT: This paper presents observations of hydraulically-driven fracture initiation (FI) in a cohesionless sand, and develops a Statistical Fracture Mechanics based model to predict the probability of initiating a fracture of a given length and orientation. The FI experiments are done in compacted 2-foot cubic sand packs in a stiff chamber under controlled anisotropic confining stresses. An aluminum pipe (9-mm ID) with 2-mm diameter perforations is used as a wellbore. A preheated low viscosity gel solution containing filter cake building solids is employed as a fracturing fluid, and injection rates are varied over two orders of magnitude. Post-test crosslinking of the gel results in solidification of the invaded zone (IZ), preserving the created fracture morphology and allowing excavation and cross sectioning for detailed analysis. Sectioning of the IZ near-wellbore region reveals a number of primary fractures initiated prior to formation of two main cracks. Tests stopped immediately after break down pressure indicate the number and locations of primary cracks are random and uncorrelated with fracturing fluid solids concentration. However, the lengths tend to follow the magnitude of hoop stress, and the average length decreases inversely proportional to cube root of the injection rate. Based on the stochastic nature of the FI a probabilistic model of HF initiation is formulated which predicts the probability distribution of primary initiation cracks formed around the wellbore as a function of hoop stress using one adjustable parameter. The prediction is in a reasonable agreement with the experimental observations. 1. INTRODUCTION Hydraulic fracturing (HF) is a commonly used technology that enhances oil and gas production [1,2]. HF of consolidated rocks has been well established for a long period of time. In the last few decades, this method has also been applied to unconsolidated reservoirs such as sand beds, like the ones offshore in Gulf of Mexico, Western Canada and Brazil [3]. Accordingly, the laboratory studies of HF in unconsolidated materials are relatively recent, in contrast with HF in solid rocks that have been done over the past six decades. Examples of small-scale laboratory studies of HF can be found in Di Lullo at al. [4], De Pater and Dong [5], Bohloli and De Pater [6], Khodaverdian and McElfresh [7] and Chang [8]. This paper specifically addresses the observation and modeling of fracture initiation. A companion paper [9] presents the effects of varying the stress state, injection rates, and fracturing fluid solids content on the HF characteristics and morphology of the fully developed HF. HF initiation in cohesionless sand (as well as in solid rocks) is the most uncertain stage of the HF process with a very high scatter of fracture initiation (FI) characteristics. In-situ observation of FI is a very difficult task due to the unpredictability of the location and time of the event. There are also conceptual difficulties in defining fracture initiation within conventional continuum mechanics, i.e., fracture of a point of continuum. Indeed, cutting out a point that has zero dimensions does not have any effect on the stress and strain fields.
- North America > United States (1.00)
- North America > Canada (0.87)
ABSTRACT: Observations and characterization of laboratory scale hydraulic fracture (HF) in cohesionless sand are reported. The sand is compacted in a stiff rectangular (2'x2'x2') pressure chamber and subjected to controlled confining stresses px, py and pz. A heated low viscosity gel is employed as a fracturing fluid with variable concentration of filter-cake building solids, which solidifies after injection to preserve the fracture and invaded zone (IZ) morphology. The effect of injection rate, solids concentration and confining stress magnitude and anisotropy on the created HF is reported. Results show that depending on the value of these parameters the outcome spans across matrix flooding, cavity formation, single fracture formation or multiple branching stochastic fracturing. In many cases a fracture (displacement discontinuity) with a clearly defined boundary surrounded by a diffuse filtercake is created. Increasing the injection rate or lowering the filter-cake building solids concentration can have similar effects. Both cases cause a transition from a roughly planer fracture to a more chaotic or diffuse fracture network and possible cavity formation. Increasing the confining stress magnitude and anisotropy also increases the confinement of the fracture and its' associated IZ to the plane perpendicular to the minimum principal stress. The presented results together with complementary data reported in the open literature present a consistent picture of cohesionless sand response to injection of various composition fluids with variable rate. 1. INTRODUCTION Hydraulic fracturing (HF) is a commonly used technology that enhances oil and gas production [1,2]. HF of consolidated rocks has been well established for a long period of time. In the last few decades, this method has also been applied to unconsolidated sand reservoirs such as those offshore in the Gulf of Mexico and Brazil, and onshore Western Canada [3]. Although hard rock fracturing analysis and prediction techniques have been used with some success in these reservoirs, it is also clear that the underlying physics of the HF of these cohesionless materials is still poorly understood. Examples of small-scale HF laboratory studies in unconsolidated materials can be found in Di Lullo at al. [4], De Pater and Dong [5], Bohloli and De Pater [6], Khodaverdian and McElfresh [7] and Chang [8]. These publications describe tests performed in isotropic or biaxial cylindrical or half cylindrical cells, which provides for generally simpler geometry and application of confining stress. However, in general the stress conditions in the field are fully three dimensional, with different magnitudes in all three principal directions. In particular, an isotropic confining stress does not provide a “preferential” direction for crack propagation, which in most cases does not represent field conditions. In order to better quantitatively characterize HF in cohesionless sand, the study presented in this paper was carried out on cubical sand packs with different pressures in the three principal directions. The effects of varying the stress state, injection rates, and fracturing fluid solids content on the HF characteristics and morphology are presented. A companion paper [9] specifically addresses the observation and modeling of fracture initiation in sand.
- North America > United States (1.00)
- North America > Canada (0.87)
- Research Report > New Finding (0.48)
- Research Report > Experimental Study (0.34)
- Well Completion > Hydraulic Fracturing > Fracturing materials (fluids, proppant) (0.73)
- Well Drilling > Drilling Fluids and Materials > Drilling fluid selection and formulation (chemistry, properties) (0.72)
- Reservoir Description and Dynamics > Reservoir Characterization > Faults and fracture characterization (0.66)