In matrix-acidizing long intervals, diversion is essential to obtain good placement of the reactive fluids. Over the years, several diversion techniques have been applied, resulting in improved zonal coverage. These diversion methods can be divided in mechanical methods such as the use of balls to seal the perforations and chemical methods such as gelled fluids.
In the design of matrix acid treatments, placement models that predict the zonal coverage of the fluids are of great help. These models improve the understanding of the complex processes and will help to improve the design of matrix-acid treatments. For studying diversion effects, correct modeling of the diversion methods in a placement simulator is essential. We have developed such a fluid placement simulator (FPS) that contains models for the different diversion methods.
In this paper, we will give an overview of the different diversion methods and their application. Further, we will discuss the implementation of the models in a comprehensive FPS. We will show how this simulator can be used to optimize placement and diversion. Validation of the models will be presented based on the analysis of two case histories.
In an acid treatment, the fluid diversion design is often based on guidelines, rules-of-thumb, and an intuitive idea on how diversion "works." Simulators are not used, usually because they are not available. However, the use of a diversion simulator will show that many of the guidelines and intuitive ideas are wrong, or at least incomplete; this will be illustrated with example calculations.
Optimum fluid placement and complete zonal coverage is essential in a successful matrix-acid treatment. This is especially true for long intervals with a high degree of permeability heterogeneity. Without effective fluid diversion, the injected fluids will follow the path of least resistance, and will only stimulate the zones with the highest permeability or the least damage.
To place the injected fluids more uniformly, the injection profile needs to be changed by diverting the flow from one zone to another. Fig. 1 shows an illustration of a successful diversion treatment. An FPS was used to simulate the flow profile in two zones with different permeabilities: 200mD and 50mD. Initially, most of the fluid enters the 200mD interval, but after diversion the flow is divided equally over both zones.
Various diversion methods, such as viscous fluids, foams, ball sealers, particulates and the maximized pressure differential and injection rates technique (MAPDIR), have been used and are described in the literature.1–3 These methods will be briefly discussed in this paper together with a description of how these methods are modeled in the FPS. Besides these diversion methods, mechanical isolation devices such as straddle packers or bridge plugs, to completely isolate an interval, are applied. Further, coiled tubing is frequently used to place the fluid at a desired location in the wellbore.
During the design phase of a matrix treatment, it is essential to understand if diversion is required and which diversion method is effective. Currently, selection and design of a diversion method is often based on general guidelines and rules-of-thumb. Simulators are often not used because they are not available. However, in most wells with a heterogeneous permeability profile, flow conditions are complex and fluid distributions cannot be predicted without using a numerical simulator. In this paper several examples will be presented with simulator results that are counter intuitive or at least show the limitations of generally accepted rules of thumb.
Therefore, validated simulators are essential tools to help in the design optimization process. Various diversion methods can be simulated, the diverter volumes can be varied, and predicted results can be compared to obtain an optimum treatment design.
Prerequisite for using these simulators for the design of a diversion treatment is trust in its reliability and correctness of the input data. Even with well-established simulators the "garbage in garbage out" statement holds. Therefore problem identification, well and reservoir data collection, and sensitivities analysis on this collected data to cover for uncertainties in the data are important.
The FPS used for this investigation is part of an integrated software package for candidate selection, treatment design and evaluation.4 The FPS is based on the model described by Jones and Davies.5 An outline of the model is provided below. For additional details we refer to the original paper.