Mark G. Mack, PhD, Oxane Materials, Inc., reviews the importance of proppant selection for shale reservoirs and looks at the keys to optimising conductivity, proppant transport and cost.
In low-permeability reservoirs, hydraulic fracture design focuses more on creating fracture length than conductivity because hydraulic fracture surface area strongly influences production. In ultra-low permeability reservoirs such as shales, hydraulic fractures are designed to create complexity because the surface area in a network of created and natural fractures is much larger than in a simple planar fracture. Thin fluids are used to promote complexity and reduce cost. These fluids severely limit proppant transport to the fracture tips because conventional proppant forms an immovable bed at the bottom of the fracture.
Three assumptions are often made regarding required conductivity, proppant transport mechanisms and cost trade-offs in unconventional reservoirs, namely:
Proppant pack conductivity is unimportant because hydraulic fractures in ultra-low permeability formations are effectively infinitely conductive.
In thin fluids such as slickwater, proppant settles much faster than Stokes Law implies, and it is difficult to transport proppant far into the fracture, unless later proppant rides over a bank of previously-settled proppant.
Ceramic or advanced ceramic proppants (ACPs) do not improve production enough to justify their cost.
These assumptions have a significant impact on well performance, and are often incorrect.
To avoid limiting production due to converging flow in a fracture connected to a horizontal wellbore, a much higher near-wellbore conductivity is required than a comparable fracture connected to a vertical well. Conductivity deep in the fracture can be achieved by using small-diameter advanced ceramic proppants (50/60 and 60/70).
The three mechanisms of proppant transport in thin fluids (suspension, saltation and reptation) are all enhanced by using proppants with a high strength/weight ratio. The high coefficient of restitution (‘bounciness’) of ACPs increases the development and transport of the dynamic layer above the proppant bank, increasing proppant transport due to saltation, and the low friction coefficient increases the ability of the proppant to roll when fluid velocity is low.
Field trials of an ACP have demonstrated that wells treated with advanced ceramic proppant increase oil and gas production by 20% and 38% respectively over wells treated with conventional ceramic proppant. An economic model based on a simple type curve and costs of a typical Duvernay well shows that advanced ceramic proppant has the potential to increase net present value by 35%, reduce payout time by 20%, and increase the internal rate of return of the well from 43% to 54%.
Conductivity in complex hydraulic fractures around a horizontal wellbore
The dimensionless fracture conductivity (FCD) is the resistance to flow in the reservoir divided by the resistance in the fracture. Prats (1961) showed that increasing the dimensionless conductivity beyond approximately 10 or 20 would not significantly increase production, leading to the common rule of thumb that a fracture with a dimensionless conductivity greater than 10 can be assumed to have ‘infinite conductivity’. If the formation permeability is very low, even low conductivity fractures theoretically exhibit infinite dimensionless conductivity. For example, in a 100 nD reservoir, 40/70 sand would appear to provide infinite conductivity even at a stress of 10 000 psi. This type of analysis has led to the widespread misperception that almost any proppant is sufficient in ultra-low permeability formations such as shales because all fractures will effectively exhibit infinite conductivity.
Complex hydraulic fracture networks around horizontal wells introduce two factors that challenge the classical theory:
The flow in transverse fractures connected to horizontal wells converges to the wellbore, with radius much smaller than the fracture height. This can increase the near-wellbore pressure drop in the fracture by a factor of 10 or 20 relative to a fractured vertical well.
The complex fracture network may have many branches. As a result, even if the stimulated region extends only a few hundred feet from the wellbore, the equivalent bi-wing fracture length (fracture surface area divided by average fracture height) can be at least an order of magnitude larger, i.e., several thousand feet.
Converging flow around a horizontal wellbore
In a vertical hydraulic fracture, the fracture cross-sectional area available for flow is the product of fracture height and propped width. Near a horizontal wellbore, the flow in the fracture converges from linear flow to radial flow as the fluid in the fracture approaches the well. The same amount of gas thus flows through a much smaller cross-sectional area, i.e., the velocity and pressure gradient both increase by a factor of H/(pr) relative to a vertical well, analogous to the pressure gradient in radial flow in a reservoir. The effective dimensionless conductivity decreases by this same factor. In addition, the pressure loss due to non-Darcy effects is increased by a factor of (H/(pr)2 over the pressure loss in a fracture associated with a vertical well. As a result, the pressure drop in the fracture within a few feet of a horizontal wellbore can be 10 to 20 times that in a vertical fracture connected to a vertical wellbore.
The large pressure drop in the near-wellbore region implies that the pressure further into the fracture approaches reservoir pressure, and that there is no production from the part of the fracture, which extends deeper into the formation, i.e., the effective fracture length is very short, especially if sand is used as a proppant. In contrast, the near-wellbore pressure drop through conventional ceramic proppant leaves 70% of the total pressure drive available for production, while using Advanced Ceramic Proppant would increase production a further 16% over conventional ceramic. An increase in production of this magnitude, achieved with a 200 000 to 300 000 lb tail-in in each stage would have a payback period of less than a month in many unconventional reservoirs. This example shows the critical importance of even small improvements in near-wellbore proppant conductivity.
Intuitively, lower fracture conductivity is adequate near the tip of the fracture because this part of the fracture carries much less of the total flow. However, the conventional fracture models that guide this intuition are not valid in unconventional reservoirs. For example, Cipolla et al (2012) used a complex fracture model to show the impact of the conductivity of unpropped fractures on production. In one example, they showed that increasing far-field conductivity from 0.03 md-ft to 1 md-ft led to a 90% increase in EUR. These analyses showed that even in extremely low-permeability reservoirs, a minimal conductivity of 5 to 10 md-ft is required in the far-field.
It is difficult to transport conventional ceramic proppants deep into the fracture network with thin fluids because they settle out of suspension. Exotic materials such as ultra-lightweight and thermoplastic alloys can be carried deeper into the formation, but traditional proppant packs of these materials are not very conductive because they are extremely weak or highly deformable under stress and temperature.
A more robust solution has been developed, namely relatively strong but light ACPs in very small size ranges.
Conductivity of small-diameter ACPs
ACPs are engineered to exhibit tight strength and size distributions, high strength/weight ratios, and exceptionally uniform shapes. The small number of low-strength grains reduces the amount of crushed fines blocking porosity, and tight size distributions can significantly enhance proppant pack permeability, because the particle size distribution can have a greater effect than absolute particle size on porosity.
In sparsely packed areas of the fracture network, the load borne by individual proppant grains may not be uniform. For example, if the grains are sparsely packed, the grains which are not surrounded by other grains bear a larger load than those which share the load with their neighbours. The two major effects of this stress amplification are that individual grains carrying a larger load are more prone to break and the added pressure at the contact points increases proppant embedment. Both of these effects can significantly impact pack conductivity.
The stress experienced by the proppant grains is the effective stress ,(seff), i.e., the closure stress minus the pressure in the fracture. As the fracture flows back and cleans up, the fluid pressure in the fracture decreases, increasing the effective stress. The maximum ,?-eff. occurs where the pressure is lowest, i.e., at the wellbore. In a high-conductivity fracture, with minimal pressure drop due to flow in the fracture, the effective stress remains high deep into the fracture. At the other extreme, even in a fully cleaned up very low-conductivity fracture, the fracture pressure near the tip is the reservoir pressure, and the effective stress is lowest, i.e., the closure stress minus the reservoir pressure.
At typical Duvernay stresses, the effective stress ranges from 8000 psi in a highly conductive fracture to 3000 psi in a zero-conductivity fracture. Near the wellbore, the effective stress in a well-packed fracture would be 8000 psi, with some potential stress amplification due to less than perfect packing. Near the fracture tips, stress amplification would result in an effective stress of at least 6000 psi, even in a low-conductivity fracture, and as much as 16 000 psi or higher in a highly conductive fracture.
Although the conductivity of packs of small beads (e.g. 50/60 or 60/70) is lower than packs of larger size ranges (e.g. 30/40 or 40/50) at low stress, Figure 1 shows that the conductivity is more sustained than other proppants at high stresses. Above 7000 psi, 50/60 ACP is more conductive than 40/70 or 40/80 ISP, and at even higher stresses (e.g., 14 000 psi), even much larger proppants (30/50 and 30/60) only provide 20% more conductivity. However, it is difficult to place large proppants, especially in thin fluids such as slickwater.
Figure 1. Production and presure: Top left – infinite-conductivity fracture; Top right – moderate conductivity fracture; Bottom left – low conductivity fracture; Bottom right – production and pressure drop depend on conductivity.
Proppant transport in slickwater
In conventional fracturing fluids, suspension dominates proppant transport. Proppant slowly settles out of the fluid to form a bank, and most of the proppant remains in suspension until the fracture closes. In thin fluids, proppant settles faster, and two other mechanisms are important namely saltation – in which proppant grains bounce off the surface back into the flowstream, and reptation (creep) – in which proppant grains roll or slide along the surface of the settled proppant bank. Density and particle diameter affect all three transport mechanisms, the coefficient of restitution influences saltation, and friction determines the rate of reptation.
Suspension: In conventional fracturing fluids, the proppant settling velocity is proportional to the square of the proppant diameter, and in slickwater it is proportional to diameter to the power of 1.3. Figure 2 shows the settling rates for a broad range of proppants, as well as the critical fluid velocities for saltation and reptation discussed in the next two sections.
Reptation: If the flowrate over a settled proppant bank gradually increases, the particles will first reptate or creep and then saltate. Camp (1946) showed that the minimum ‘scouring velocity’ (i.e., the velocity at which particles move across a surface) is proportional to the square root of the friction coefficient. ACPs typically exhibit a friction coefficient roughly 20% lower than conventional ceramic proppants, and 40% lower than sand. ACP particles can thus be mobilised at fluid velocities 20% lower than sand and 10% lower than conventional ceramic particles.
Figure 2. Proppant settling velocity in slickwater.
Saltation: Even when the proppant bed is moving, much of the load travels close to the bed (Wilcock, 2004), and particles are continuously moving between the bed and the flowstream. Saltation is the tendency of particles in the fluid to bounce off the bed and for particles on the bed surface to be kicked up into the flowstream and then travel in an arc before hitting the bed and continuing the process. The coefficient of restitution (CoR) is the ratio of the velocity at which an object leaves a collision to the velocity at which it enters the collision, measuring how much it bounces. ACP has been demonstrated to have a higher CoR than conventional ceramic proppant or sand, and will thus saltate more because it has both a lower density and higher CoR than conventional ceramic proppants.
In summary, all three mechanisms of proppant transport are improved for light, small particles. The two mechanisms related to banks or dunes of proppant are also affected by other properties of the solid particles, specifically the coefficient of restitution and the static friction coefficient. The material properties of ACPs are favourable to proppant transport under all three mechanisms. In particular, small particles such as 50/60 and 60/70 can remain in suspension and be transported far from the wellbore before forming a proppant bank.
Field trial of advanced ceramic proppant
A comprehensive 20-well field trial of an ACP was conducted in the Midland basin, to assess the benefit of ACP versus Intermediate Strength Proppant (ISP). The ACP was pumped in 10 wells, and conventional ceramic proppant in the other 10 offset wells. Each ACP well was paired with a direct-offset ISP well, to minimise the impact of variations in reservoir quality, and stimulated at similar times to minimise the impact of interference or overlapping drainage areas.
The completed interval ranged from approximately 8000 to 11 000 ft in depth, including the Wolfcamp shale and several formations above and below the Wolfcamp, with a typical frac gradient of 0.7 psi/ft. Approximately 800 000 lbs of proppant were placed in each well, of which 300 000 lbs were ceramic (either ACP or ISP), using a low-viscosity fluid pumped at a high rate. In the offset wells, 40/70 ISP was used in the deeper zones of the ISP well, and 30/50 ISP in the shallower zones, whereas most of the ACP wells were treated with 40/50 ACP in all zones.
Advanced Ceramic Proppant increased cumulative 12 month oil production by at least 20% relative to conventional ceramic proppant. Significantly, the operator had previously observed a similar increase in production when ISP was substituted for sand. In other words, ACP improves production by 40 percent relative to sand. An ‘equivalent production’ analysis can be undertaken by subtracting all effects other than the contribution of proppant from the actual production from the ACP well. This represents the production, which would have been achieved in the ACP well if it had been located and treated identically to the ISP well. Figure 3 shows the results of the analysis. The contribution of proppant varies from a low of 6% in Well Pair 8 to a high of 43% in Well Pair 4.
Figure 3. Left - Conductivity of advanced ceramic and intermediate strength proppant. Right - Conductivity of small-diameter proppant.
Economics of proppant selection
A longer, more conductive hydraulic fracture or fracture network will deliver more production than a small, low-conductivity fracture. However, fracture treatment design is a trade-off between the increasing unit cost of adding length or conductivity and the diminishing increase in production derived from the incremental length or conductivity. Most of the approaches used for economic optimisation of fracture design vary parameters such as fluid and proppant volumes, concentrations and flowrates, but seldom consider changes in proppant type in any detail.
The key decision criterion in selecting a specific new technology is whether the incremental production warrants the incremental cost. In many cases, the additional cost of a better proppant is small relative to the significant investment, which has already been made to drill and complete the well. Similarly, even a small percentage increase in production from a good well (with a high NPV) is sufficient to justify the additional cost of superior proppant.
Investor information published by several operators was used to evaluate the economic value of selecting ACP over conventional ceramic proppant (ISP) in the Canadian Duvernay play. The data used in the economic model is shown in Table 1. ACP is estimated to increase the well cost by 7%, which would require a production increase of 4% to be NPV-neutral. Assuming a 20% production increase vs. ISP, consistent with the results reported from using ACP in the Permian, the NPV would increase by 35%, payout time would reduce from 24 months to 20% and the IRR of the well would increase from 43% to 54%. The IRR of the investment in ACP would exceed 100%.
Summary and conclusions
Increasing near-wellbore conductivity can have a significant positive impact on the production from horizontal wells, even in ultra-low permeability formations. Advanced ceramic proppants are especially beneficial due to their low resistance to non-Darcy flow.
Fracture conductivity deep in the fracture can be reliably enhanced using strong, small-diameter proppants which transport well and provide good conductivity in narrow fractures. ACPs are well-suited for this purpose due to their tight size distribution, high sphericity and high retained conductivity at high stress.
Three proppant transport mechanisms have been identified. Suspension is enhanced by reduced density and particle size, reputation (creep) is enhanced by reduced friction, and saltation is enhanced by an increased coefficient of restitution. In slickwater, the majority of proppant is often transported by saltation.
In a field trial, 12 month oil and gas production were increased by 20% and 38% respectively when ACP was substituted for conventional ceramic proppant.
A model has been developed to estimate the economic benefit of the increased production achieved by replacing sand or conventional ceramic proppant with ACP. In a representative Duvernay well, this substitution increased the well cost by 7%, and increased the well net present value by 35%.
Edited for web by Cecilia Rehn