Abstract

There are many areas in western Canada that have shown abnormal pressure responses during hydraulic fracturing which make treatment design and proppant placement very difficult. These unpredicted responses are thought to be due to heterogeneous formation characteristics such as natural fractures, faults, high permeability streaks and abnormal stresses.

The application of computerized data collection and analysis during hydraulic fracturing has made it possible to closely examine many treatments in these areas. It was found that this presence of heterogeneous formation conditions can be identified using treating pressure analysis following specialized minifrac techniques. By examining the pressure responses, it is possible to make specific design modifications that improve proppant placement in such abnormal well conditions.

In this paper some general design guidelines have been developed to deal with heterogeneous formations. Included are case histories of successful applications.

Introduction

Hydraulic fracturing has proven to be a very successful form of stimulation in many oil and gas hearing, formations across western Canada. There have, however, been some areas which have proven to he very difficult to treat using conventional methods. The major problem in these areas has been very unpredictable treating pressure responses that tend to vary significantly on a well-to-well basis. In many of these wells the amount of proppant placed in formation has been small and the production results have generally been disappointing.

Since fracturing still appears to be the most attractive form of stimulation for many of these wells there has been a need to develop specialized design methods to help place more proppant and improve post treatment results. To accomplish this it was necessary to examine the limitation of conventional design procedures and look more closely at the problems encountered in these areas. Many of the computer fracture design models being used today combine fracture mechanics, fluid reservoir engineering and economics to develop an optimal treatment design. For many formations these models have worked very well. There are areas, however, where these methods have consistently failed.

It was felt that formation heterogeneities and rock stresses in these areas presented a condition far too complex to be handled properly with the simulator programs currently available to the industry.

Although there are several fracture geometries available for use in simulation programs, there are several basic assumptions that are common to nearly all cases. Most of these assumptions were made in an attempt to simplify the complex problem of fracture mechanics of a hydraulically induced fracture. Some of these assumptions have been summarized below.

  1. Homogeneous, isotropic and linearly elastic formation rock.

  2. Continuous fracture propagation during pumping.

  3. Unrestricted fracture length propagation.

  4. The created fracture will consist of two identical wings symetrical about the wellbore and follow a known width equation.

  5. Fluid loss is strictly dependent upon matrix leak-off and the surface area of the fracture.

DISCUSSION

Although the process of hydraulic fracturing has been used for many years, we have only recently begun to understand the relationships between treating pressures and fracture development.

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