Laboratory testing has shown that formation mineralogy plays an unrecognized role in determining proppant permeability in hydraulic fractures. Most proppants were found to lose 50–60% permeability before stabilizing, while others lost up to 90% permeability in only a few days.

Fig. 1—Illustration of the pressure solution and compaction mechanism. High stress at the contact points increases the solubility of silica, which diffuses into the pore spaces and precipitates.

The selection of proppant to provide highly conductive pathways in hydraulically generated fractures is typically based on the proppant crush strength, conductivity, availability, and cost. Laboratory-determined conductivity values, obtained using API standardized methods at a variety of simulated well conditions, are available for most proppants. However, well testing after fracture-stimulation indicates that these conductivity values are often too high. In many fields, the productivity of a fracture stimulated reservoir declines rapidly, requiring frequent re-stimulation treatments to remain economically viable.

Conductivity damage mechanisms that have been used to explain this loss of productivity include proppant crushing and embedment, fracturing fluid damage, and fines invasion of the proppant pack. However, this article reports on recent studies and testing that indicate some aluminum-based proppant materials may promote geochemical reactions that can occur at a surprising rate, even at moderate temperatures, resulting in (1) loss of porosity and permeability and (2) creation of fines in the proppant pack.

Fig. 2 – SEM image of a Haynesville core sample showing consolidation of the core surface to the intermediate strength, alumina-based proppant after exposure to diagenetic testing. Notice that (a) the formation and proppant surfaces both become covered with aluminosilicate mineral clusters owing to diagenetic reactions between the formation and the proppant and (b) much of the porosity of the proppant pack becomes filled with these mineral deposits.

In the absence of adequate industry standardized tests, two specialized test methods presented below were developed to study these geochemical reactions and report quantitative changes in permeability, proppant composition, and fluid changes. Surface analysis using scanning electron microscopy (SEM) and energy dispersive X-ray (EDX) provided confirmation of this damage mechanism. Operators, service companies, and proppant suppliers should consider developing their versions of these tests or create tests of their own design to review tendencies of currently marketed synthetic proppants to exhibit early-onset diagenesis on contact with the reservoir (Weaver et al. 2008).

Background

Much effort has gone into standardizing the selection of proppants to optimize stimulation cost against expected stimulation results. Many proppant materials are available, including natural sands, ceramics, sintered minerals, plastics, and composite materials. These are provided in a variety of size distributions and ranked based on their crush strength and conductivity at expected formation closure stress and temperature. Industry standard methods have been implemented to provide performance comparison of these materials. Cost and availability are additional constraints that impact the ultimate proppant selection.

Fig. 3—Load frame equipped with four radial conductivity cells used to determine the effect of long-term stress on the rate of proppant diagenesis.

Standard methods (API RP 56; API RP 60; ISO 13503-2) for determining the crush strength of proppant include putting a carefully sieved and measured dry sample of proppant in a cell of specific dimensions and applying a pressure load for two minutes, then sieving the material to determine the percentage of crushed proppant. This method may not be considered applicable to resin-coated proppants (RCP), or proppants that have been produced back after fracture treatments.

The current API recommended method to measure proppant conductivity values are presented in API RP 61(API RP 61, 1989).

Fig. 4—Schematic showing proppant diagenesis dynamic flow test apparatus capable of performing several sample tests concurrently.

Proppants are loaded at a specified concentration (e.g., 2 lb/sq ft) between two core slabs (e.g., Ohio Sandstone) in an API conductivity cell. The proppant pack conductivity is then measured at a stress level a few times until it stabilizes (normally within only 50 hours) with less than 5% change, for example. However, it is common to observe continuous decline in conductivity, although less than 5% on a daily basis, if the measurement is carried longer than two days.

Laboratory-measured conductivities are often not in agreement with observed data in actual post-frac production tests. In addition, the initial fracture conductivity following fracture stimulation often drops more rapidly than it should based on fractured well reservoir simulations of production values.

The lack of laboratory capability to measure conductivity to match field production test results has led to the study of conductivity damage mechanisms. Considerable understanding about these mechanisms has been achieved over the past few years and numerous papers have been published sharing this understanding. Some of the major mechanisms identified include frac fluid damage, proppant embedment, proppant crushing, fines migration, and formation fines invasion (Nguyen et al. 2008).

Recently, proppant diagenesis (Weaver et al. 2007), a new fracture damage mechanism, was proposed. Classical diagenesis is considered to be a slow process occurring over centuries, but in fact, proppant-diagenesis reactions occur fairly rapidly, requiring only fractions of years at reservoir conditions in a propped hydraulic frature. The centuries of time normally associated with diagenesis is the time required to bury sediment deep enough to reach the pressure and temperature conditions conducive to diagenetic change. In the case of fracture stimulation, generally mature rock formations we fracture have already been at temperature and stress for centuries. During the fracture treatment, a virgin pack of “sediment” (proppant) is placed in the created fracture to prevent the rock faces from closing back to their original position. With the temperature and stress conditions that are already in place, the diagenetic porosity loss of the proppant pack can happen in fractions of a year.

The pressure solution and compaction mechanism as illustrated in Fig. 1 is one driver promoting proppant diagenesis. Consider two proppant grains under a high mechanical load. At the contact points, the load is extremely high. In the presence of water, the solubility of the proppant at the contact points is many times higher than the proppant when it is under no mechanical stress. The mineral at this contact point solubilizes and subsequently diffuses out to the pore space where there is no mechanical stress. The solution becomes supersaturated which results in the precipitation of the minerals in the pores. (Fig. 2) This process removes mineral from the proppant contact points and redistributes it into the porosity. This results in the proppant grains decreasing in diameter in the load-bearing axis while the overall porosity decreases. This process of proppant diagenesis has a major effect on fracture conductivity since it results in loss of fracture width and loss of porosity, i.e. conductivity.

Fig. 5—SEM images of sintered bauxite. Image on the left is before testing with increasing magnification from top to bottom. Image on the right is the same proppant after 7 days of flowing synthetic formation water at 550°F. Magnifications are the same for both images.

Quad-Cell Test Method

The rate at which geochemical degradation of proppants occurs is, among other things, a function of closure stress on the proppant and the temperature to which the proppant is exposed. Information required to allow calculation of reaction rates with respect to closure stress and temperature was acquired using the apparatus in Fig. 3.

This test fixture consists of a load frame and four modified radial conductivity cells. A common closure stress (up to 10,000 psi) can be applied to each cell and four temperature controllers allow four different temperatures to be evaluated. Proppant is loaded (at 1.5 lb/sq ft) into the cell with formation wafers both above and below the proppant. The four cells are filled with the test fluid, sealed, and set at four different temperatures for the duration of the tests. Dimensional changes in the proppant packs are measured and recorded. Micro samples of the interstitial fluids can be taken for elemental analysis during the testing session if required. Testing times range from two to six weeks.

Post-process analyses are performed to identify any changes in the formation samples, the proppant, and the fluid. Dimensional changes of the proppant pack provide insight as to the rate of compaction and embedment with respect to time and temperature. In some tests, polished thin-section analyses are performed. Brunauer, Emmett, Teller (BET) surface analysis measurements provide surface area and porosity data showing that proppant materials change significantly during the exposure time.

Temperature-Promoted Diagenesis Test Method

Fig. 6—This chart shows significant loss of conductivity for all three sintered bauxites after exposure to temperature-promoted diagenesis testing at 550°F for 7 days, using synthetic formation water. Conductivity is shown as percentage of reported API baseline conductivity for 20/40 mesh at 2 lb/ft2 with 2% KCl at 300°F, between Ohio sandstone wafers. These proppants are rated “equal” in conductivity provided.

This test method enables study of the chemical interactions between formation and proppants with respect to temperature; the method allows multiple proppants to be evaluated simultaneously. Fig. 4 shows a schematic diagram of this test apparatus. The process begins with a supply of deionized (DI) water supplied to the system at high pressure, which is maintained by the use of backpressure regulators that keep the water in the liquid phase at test temperatures. Testing is performed at temperatures ranging from 200 to 550°F (93 to 287°C).

The deionized water is first heated to the test temperature by passing it through a heat exchange coil located in the same convection oven used to house the proppant-pack cells. Since it is not possible to use actual formation water, synthetic formation water is created by flowing the water through multiple packs of crushed formation material arranged in series. By experiment, the number of formation packs required to generate equilibrated formation water is determined. Crushed formation material is screened and only the 8/35-mesh fraction is used. These additional specific steps complete the test:

1. Create equilibrated formation water by flowing DI water in series through formation packs.
2. Direct flow stream to flow in parallel through multiple proppant packs.
3. Determine compositional difference by ICP (inductively coupled plasma emission spectrometry) with samples taken several times during the test. (Proppants vary greatly in their response to flow conditions.)
4. Calculate solubility data at the designed test temperature by ICP analysis of the interstitial fluid. These steps provide information regarding the solubility product of the proppant minerals at elevated temperatures.
a. Fill the system with DI water.
b. Shut the system in.
c. Increase temperature to test level and hold for 24 hours.
5. Run the test for 10–30 days, flowing about 1 mL/min through each pack.

Results of tests conducted following the above steps are summarized:
• Some proppants lost mass and some gained mass, dependent on the proppant and formation compositions.
• Permeability losses were dramatic with the accelerated diagenesis studies. Such test results correlate directly to field observations and fractured well productivity.
• Surface changes were also drastic, as illustrated in Fig. 5.
• Conductivity (Fig. 6) measured after flow at 550°F (287°C) for seven days differed significantly from reported API baseline conductivity. The graph shows the percentage of conductivity compared to the baseline. These three proppants were classified as sintered bauxite, and were (otherwise) rated equal in their capability to provide fracture conductivity.

Bibliography
ASTM C 1239, Standard practice for reporting uniaxial strength data and estimating Weibull distribution parameters for advanced ceramics. 2007.

ISO 13503-2, Measurement of properties of proppants used in hydraulic fracturing and gravel-packing operations. 2006.

Nguyen, P., Weaver, J. and Rickman, R. 2008. Prevention of Geochemical Scaling in Hydraulically Created Fractures – Laboratory and Field Studies. Paper SPE 118175 presented at the SPE Northeastern Regional Meeting, Pittsburgh, PA, 11–15 October.

RP 56, Recommended Practices for Testing Sand Used in Hydraulic Fracturing Operations. 1995. Washington, DC: API.

RP 60, Recommended Practices for Testing High-Strength Proppants Used in Hydraulic Fracturing Operations, 1995. Washington, DC: API.

RP 61, Recommended Practices for Evaluating Short Term Proppant Pack Conductivity, 1989. Washington, DC: API.

Weaver, J., Parker, M., van Batenburg, D., and Nguyen, P. 2007. Fracture-Related Diagenesis May Impact Conductivity. SPEJ September: 272.

Weaver, J., Rickman, R, and Hongyu, L. 2008. Fracture Conductivity Loss Due to Geochemical Interactions Between Man-Made Products and Formations. Paper SPE 118174 presented at the SPE Northeastern Regional Meeting, Pittsburgh, PA, 11–15 October.