Economides, Michael J., SPE, Dowell Schlumberger


There are many causes of abnormally low well production rates, and a good engineering approach should identify them. They include low permeability, low reservoir pressure, high bottomhole pressure (BHP), high reservoir fluid viscosity, and high skin. Well stimulation is any action to remedy or to counteract any of the causes of low well production rates. However, one usually deals with low permeability or high skin values (i.e., damaged zones near the wellbore). Matrix stimulation, whose aim is to remove the formation damage associated with high skin values and to restore or improve the permeability near the wellbore, is not considered here. Hydraulic fracturing is usually the most effective stimulation treatment in low-permeability reservoirs. While hydraulic fracturing does not change the reservoir permeability, it superimposes a new, high-permeability structure within the reservoir that is in communication with the well. Several authors have introduced the concept of effective wellbore radius, implying that the reservoir "looks" at the fracture (especially at late time) as a wellbore enlargement. Tight formations (usually less than 1-md permeability) are often candidates for hydraulic fracturing. Propped fractures are indicated for sandstone reservoirs. These fractures are of "finite conductivity" (a measure of permeability contrast between the fracture and the surrounding reservoir), and their performance is a function of this conductivity and the fracture length. Acid fractures are often indicated for formations that can react with an acid (i.e., limestones and dolomites reacting with HCl). These stimulation treatments should not be confused with matrix acidizing. Acid fracturing is done with the BHP above the fracturing pressure of the formation. Etching of the created fracture face is then accomplished, and when the treating pressure subsides, an open channel remains. This fracture has a high conductivity, provided that the strength of the rock is sufficient to prevent collapse. Depending on rock and proppant properties, it is possible for an acid fracture to have a lower fracture conductivity than a propped one. Fracturing in higher-permeability formations is usually not indicated because the incremental benefits diminish rapidly when the reservoir permeability exceeds 10 md. The remainder of this paper considers the elements that can contribute to a successful hydraulic-fracture-treating program. Not all the elements need to be done in every case. However, they should be considered carefully, because they might affect both the success and the interpretation of the fracture-stimulation treatment. Inadequate evaluation could fail to identify changes and improvements to include in subsequent treatments. The following sections cover the pretreatment, execution, and evaluation phases of hydraulic fracturing. As can be appreciated, the key phase in fracture design is the pretreatment analysis.

Pretreatment Analysis and Design

An appropriate hydraulic-fracture design depends on the use of representative data. A pretreatment well test is an invaluable tool to supply permeability, skin. and initial or average reservoir pressure. Permeability is the most important parameter affecting fracture-design parameters and makes posttreatment evaluation possible. In the absence of flow test capability (very tight or very low-pressure formations), an injection/falloff test is recommended to estimate permeability. For many wells, the mere displacement (at subfracturing pressures) of the wellbore fluid and evaluation of the accompanying pressure decline could result in enough data for the estimation of the permeability and skin. Other useful sources of data include a good suite of logs (for porosity, saturations, and alternative indications for permeability), cores and petrographic analysis (to decide whether an acid fracture or propped fracture may be indicated), and geologic information (for estimating reservoir extent). The expected fracture geometry, orientation, and azimuth should be estimated. Except for very shallow wells (e.g., less than about 1,500 ft [460 m]), fracture orientation will generally be vertical. Fracture azimuth has major implications on the reservoir exploitation strategy and will be in the direction of maximum horizontal in-situ stress. Fracture-height growth, a major design consideration, is a function of the vertical in-situ stress profile and rock properties. In-situ stresses are preferably obtained by carefully controlled small volume breakdown/flowback testing. These in-situ stresses may then be used in simple linear fracture mechanics models to offer qualitative indications of fracture growth. Generally, a fracture height is estimated and either a Perkins and Kerns or Khristianovich and Zheltov model is used to complete the design. Rock properties (Poisson's ratio and Young's modulus) influence these fracture geometry calculations and also, together with fracture fluid viscosities and injection rates, allow estimates of expected fracture-treatment pressures to be calculated. The fracturing fluid and its proppant transport abilities should be thoroughly examined. The fluid should transport the proppant during fracture creation and continue to suspend the proppant until the fracture closes. The fluid should then flow back easily through the proppant bed as the well is produced. However, recent studies have shown that permanent fracture-permeability impairment is observed with certain fluids, proppant, and formation combinations. Such "damage" affects the "finite conductivity" of the fracture. A "minifrac" injection test should be done to define the fracture leakoff characteristics into the formation. This is an integral part of pretreatment data and should be incorporated into the fracture design. Nolte has developed a technique to calculate the "effective" leakoff coefficient from the slope of the BHP decline following an injection test. Other tests such as the "pump in/flowback test" previously noted and the step-rate test allow the calculation of the minimum in-situ stress from the inflection points in the observed pressure profiles before and after fracture closure. The reservoir features, fracture modeling results, forecasts of expected well performance, and economic considerations can be coupled to provide an "optimum" design. The word is in quotation marks because optimization criteria and models used will vary. Each job should be justified and optimized by a comparison of the economic impact of production forecasts of expected performance, such as those illustrated in Fig. I for a gas well. Shown are the cumulative production curves for the unstimulated and stimulated well (the results of matrix stimulation are included for contrast).


P. 1343^

This content is only available via PDF.
You can access this article if you purchase or spend a download.