Fracture treatment design
The most important data for designing a fracture treatment are the in-situ stress profile, formation permeability, fluid-loss characteristics, total fluid volume pumped, propping agent type and amount, pad volume, fracture-fluid viscosity, injection rate, and formation modulus. It is very important to quantify the in-situ stress profile and the permeability profile of the zone to be stimulated, plus the layers of rock above and below the target zone that will influence fracture height growth.
There is a structured method that should be followed to design, optimize, execute, evaluate, and reoptimize the fracture treatments in any reservoir. The first step is always the construction of a complete and accurate data set. Table 1 lists the sources for the data required to run fracture propagation and reservoir models. The design engineer must be capable of analyzing logs, cores, production data, and well-test data and be capable of digging through well files to obtain all the information needed to design and evaluate the well that is to be hydraulically fracture treated.
To design the optimum treatment, the effect of fracture length and fracture conductivity on the productivity and the ultimate recovery from the well must be determined. As in all engineering problems, sensitivity runs need to be made to evaluate uncertainties, such as estimates of formation permeability and drainage area. The production data obtained from the reservoir model should be used in an economics model to determine the optimum fracture length and conductivity. Then a fracture treatment must be designed with a fracture propagation model to achieve the desired length and conductivity at minimum cost. The most important concept is to design a fracture with the appropriate data and models that will result in the optimum economic benefit to the well operator, as Fig. 1 shows.
Fig. 1—Fracture treatment optimization process.
A hydraulic fracture propagation model should be run to determine what needs to be mixed and pumped into the well to achieve the optimum values of propped fracture length and fracture conductivity. The base data set should be used to make a base case run. The engineer then determines which variables are the most uncertain. The values of in-situ stress, Young ’ s modulus, permeability, and fluid-loss coefficient often are not known with certainty and must be estimated. The design should acknowledge these uncertainties and make sensitivity runs with the fracture-propagation model to determine the effect of these uncertainties on the design process. As databases are developed, the number and magnitude of the uncertainties will diminish. In effect, the design engineer should fracture treat the well many times on his or her computer. Sensitivity runs lead to a better design and educate the design engineer on how certain variables affect the values of both the created and propped fracture dimensions.
Fracturing fluid selection
The selection of the fracture fluid for the treatment is a critical decision. Economides et al. developed a flow chart that can be used to select the category of fracture fluid on the basis of factors such as:
- Reservoir temperature
- Reservoir pressure
- The expected value of fracture half-length
- Water sensitivity
Fig. 2 presents the fluid-selection flow chart for a gas well. The information in Fig. 2 is compatible with the information in Table 2.
To use Fig. 2, one must follow a path that depends on formation temperature, reservoir pressure, and an intangible variable called water sensitivity. For a low-temperature, high-pressure reservoir, the desired fracture conductivity and the desired fracture length must be considered. Economides et al. suggest that Fig. 2 can also be used to select a fluid to fracture treat an oil reservoir that is not water sensitive.
The definition of what comprises a water-sensitive reservoir and what causes the damage is not always clear. Most reservoirs contain water, and most oil reservoirs can be waterflooded successfully. Thus, most fracture treatments should be pumped with suitable water-base fracture fluids. Acid-base fluids can be used in carbonates; however, many deep carbonate reservoirs have been stimulated successfully with water-base fluids containing propping agents. Oil-base fluids should be used only in oil reservoirs when water-base fluids have proved conclusively to not work. Pumping oil-base fluids is more dangerous than pumping water-base fluids, and special care should be taken in the field.
Fig. 3 presents a flow chart created by Economides and Nolte for selecting propping agents. To use Fig. 3, the maximum effective stress on the propping agent must be determined. The effective stress is defined in Fig. 4. The maximum effective stress depends on the minimum value of flowing bottomhole pressure expected during the life of the well. If the maximum effective stress is less than 6,000 psi, then Fig. 3 recommends that sand be used as the propping agent. If the maximum effective stress is between 6,000 and 12,000 psi, then either RCS or intermediate-strength proppant should be used, depending on the temperature. For cases in which the maximum effective stress is greater than 12,000 psi, high-strength bauxite should be used as the propping agent.
Fig. 3—Proppant selection based on closure pressure (after Economides and Nolte).
Fig. 3 should be used only as a guide, because there will be exceptions. For example, even if the maximum effective stress is less than 6,000 psi, the designer may choose to use RCS or other additives to "lock" the proppant in place when proppant flowback becomes an issue. In high-flow-rate gas wells, non-Darcy pressure drops can lead to the use of ceramic proppants to maximize fracture conductivity.
For fracture treatments in countries that do not mine sand for fracturing, the largest cost of the proppant is often the shipping charges. If the propping agent must be imported, intermediate-strength proppants may be selected, even for relatively shallow wells, because the cost differential between the intermediate strength proppants and sand is not much of a factor.
To confirm exactly which type of propping agent should be used during a specific fracture treatment, the designer should factor in the estimated values of formation permeability and optimum fracture half-length. Cinco-Ley published an equation that can be used to determine the optimum fracture conductivity. The dimensionless fracture conductivity is defined as
To minimize the pressure drop down the fracture, the value of CfD should be approximately 10 or greater. The required fracture conductivity can be computed as
where k = the formation permeability (md) and Lf = the fracture half-length (ft). For example, if the formation permeability is 25 md and the optimum fracture half-length is 50 ft, then the optimum fracture conductivity would be 3,927 md-ft. The treatment must be designed to create a fracture wide enough, and pump proppants at concentrations high enough, to achieve the conductivity required to optimize the treatment. However, in many low-permeability reservoirs, the dimensionless fracture conductivity, CfD , must be 50 to 100 for the fracture fluid to clean up after the treatment. As such, the "optimum" value of CfD = 10 is considered a minimum value, and CfD should be even larger than 10 when fracture fluid cleanup issues are a problem. In high-permeability formations, CfD values of 10 or greater are often not feasible.
Some tend to compromise fracture length and conductivity in an often unsuccessful attempt to prevent damage to the formation around the fracture. Holditch showed that substantial damage to the formation around the fracture can be tolerated as long as the optimum fracture length and conductivity are achieved. However, damage to the fracture or the propping agents can be very detrimental to the productivity of the fractured well. Ideally, the optimum fracture length and conductivity can be created while minimizing damage to the formation. If the opposite occurs—that is, the formation is not damaged, but the fracture is not long enough or conductive enough—then the well performance usually will be disappointing.
Evaluating risks in the design
The well operator always should evaluate risks such as:
- Mechanical risks
- Product price risks
- Geologic risks
Uncertainties in the data can be evaluated by making sensitivity runs with both reservoir models and fracture propagation models. One of the main risks in hydraulic fracturing is that the entire treatment will be pumped and/or paid for (i.e., the money is spent), but the well does not produce at the desired flow rates nor achieve the expected cumulative recovery. In some cases, mechanical problems with the well or the surface equipment cause the treatment to fail. Other times, the reservoir does not respond as expected.
To evaluate the risk of mechanical or reservoir problems, 100% of the costs and only a fraction of the revenue can be used in the economic analyses. For example, one in every five fracture treatments in a certain formation is not successful; therefore, 80% of the expected revenue and 100% of the expected costs can be used to determine the optimum fracture length. Fig. 4 illustrates how such an analysis can alter the desired fracture length.
Finally, after the optimum, risk-adjusted fracture treatment has been designed, it is extremely important to be certain the optimum design is pumped correctly into the well. For this to occur, the operator and the service company should work together to provide quality control before, during, and after the treatment is pumped. The best engineers spend sufficient time in the office designing the treatment correctly, and then go to the field to help supervise the field operations or provide on-site advice to the supervisor.
|Cf||=||fracture conductivity, md-ft|
|CfD||=||dimensionless fracture conductivity|
|k||=||formation permeability, L2, md|
|Lf||=||fracture half-length, L, ft|
- Veatch Jr., R.W. and Moschovidis, Z.A. 1986. An Overview of Recent Advances in Hydraulic Fracturing Technology. Presented at the International Meeting on Petroleum Engineering, Beijing, China, 17-20 March. SPE-14085-MS. http://dx.doi.org/10.2118/14085-MS.
- Economides, M.J. and Nolte, K.G. 2000. Reservoir Stimulation, third edition. New York: John Wiley & Sons.
- Cinco-Ley, H., Samaniego-V., F., and A.N., D. 1978. Transient Pressure Behavior for a Well With a Finite-Conductivity Vertical Fracture. SPE J. 18 (4): 253–264. SPE-6014-PA. http://dx.doi.org/10.2118/6014-PA.
- Holditch, S.A. 1979. Factors Affecting Water Blocking and Gas Flow From Hydraulically Fractured Gas Wells. J Pet Technol 31 (12): 1515–1524. SPE-7561-PA. http://dx.doi.org/10.2118/7561-PA.
Noteworthy papers in OnePetro
Britt, Larry K. 2013. Application of Low Viscosity Fluids To Hydraulic. https://webevents.spe.org/products/application-of-low-viscosity-fluids-to-hydraulic-fracturing-spe-distinguished-lecturer
Burnett, David. 2012. New Options for Produced Water Treatment and Re-use in Gas/Oil Shale Fracturing. https://webevents.spe.org/products/new-options-for-produced-water-treatment-and-re-use-in-gasoil-shale-fracturing
Palmgren, Tor. 2013. Treatment Options for Reuse of Frac Flowback and Produced Water from Shales. https://webevents.spe.org/products/treatment-options-for-reuse-of-frac-flowback-and-produced-water-from-shales
Xiong, Hongjie. 2017. "Optimizing Cluster or Fracture Spacing: An Overview." The Way Ahead. Society of Petroleum Engineers. https://www.spe.org/en/twa/twa-article-detail/?art=3007