A PTTC workshop in Denver, conducted by Mike Vincent with Insight Consulting, helped participants understand how to put the hydraulic fracturing pieces together in a real-world environment. Treatment design should never be considered a “cookie cutter” process. Key initial questions include:

• How wide does the fracture need to be?
• How much oil and gas needs to be accommodated?
• How good is the reservoir that feeds the fracture?

This approach is in stark contrast to traditional “rules of thumb.”

There are some basic truths about hydraulic fracturing that must be accepted. Modeling fractures as simple, planar features of constant width will drastically overestimate flow capacity (as much as 100-fold). Proppant selection and fracture width do matter, even in low-flow wells. Despite the myriad complicating factors, field studies demonstrate excellent potential to increase production from most wells, primarily from improved conductivity and connectivity (vertically crossing reservoir laminations, and laterally past induced damage and heterogeneities) to increase the ultimate recovery.

Actual pressure losses in proppant packs greatly exceed estimates. Contrasts between proppants can be greatly magnified under realistic conditions. Other complicating factors include:

• Bad assumptions about fracture geometry;
• Proppant behavior in narrow fracs;
• Gel damage;
• Long-term proppant degradation;
• Flow convergence to perforations;
• Nonuniform proppant distribution;
• Emulsions, foams and froths; and
• Asphaltenes, wax, scale fouling, and fines plugging.

The Denver workshop examined how proppants performed under cyclic stress loading, high temperatures, and with long duration testing. Improved proppant performance can be obtained with superior roundness, sphericity, uniform grain size and strength, but this extra performance must be balanced against the extra cost. Both sand and ceramic proppants can be resin coated. Pre-cured or dual-coat resin coated proppants may be used to improve distribution of stresses and encapsulate fines generated on crushing. Curable resins are used to consolidate the pack and reduce proppant flow back.

Crushing and embedding can be more severe with large proppants, since with smaller proppants, stress will be distributed over more contact points and there will be more layers. Participants gained insight into balancing these considerations. Sands, ceramics and resin-coated proppants crush differently. Although crush data are a useful quality control measure, the results were shown to not directly translate into conductivity.

Although routine API conductivity tests are helpful, other factors must be considered to realistically estimate proppant conductivity. These factors include non-Darcy flow, reduced proppant concentration, multiphase flow, gel damage, fines migration, cyclic stress, etc. Regarding multiphase flow, there can be a 60-80 percent reduction for even modest liquid rates. No wonder fractures designed with reference conductivity data are poorly optimized!

Gel damage may result from three mechanisms: distributed gel damage, loss of effective width resulting from filter cake buildup, and loss of effective length because of a static gel plug in the fracture tip. Gel residue tends to accumulate in the narrowest pore throats, typically affecting Beta (inertial flow coefficient, reflective of tortuosity) much more than permeability. Higher stresses reduce pore throat aperture and reduce cleanup.

To initiate gel cleanup, a “yield stress” or threshold pressure must be exceeded to initiate non-Newtonian gel flow. Flow is more easily initiated with uniform spherical proppants that contain large, uniform pore throats. Thus, tightly sieved ceramics achieve superior cleanup with less plugging by “clumpy” broken gel.

Filter cakes of concentrated gel having the consistency of rubber sheet will be deposited at the fracture face. Gel damage has proven to be durable and likely reflects permanent conductivity loss. The seminar included an excellent discussion of slick-water fracturing to help participants understand proppant transport and where water frac techniques have been successful.

Field results confirm that pressure losses within fracs are high. Reduced pressure losses within a high-conductivity frac can increase cleanup and provide longer effective frac lengths. Well tests and production analyses frequently indicate short effective fracture lengths and low conductivity. SPE 77675 summarizes 80 field studies in which fracture redesign resulted in improved production rates. This study has been expanded to 140 tests accessible through an interactive map on Carbo Ceramics’ Web site (www.carboceramics.com).

Overall, these field results demonstrate that increased frac conductivity is more important than has been recognized. Wider fracs and better proppants provide substantially higher production–often providing more improvement than can be explained with conventional correlations. Numerous field studies confirm fracture complexity, emphasizing that the key point in a realistic optimization strategy is to analyze field results and not rely solely on simplistic models/correlations.

Natural fractures/faulting affect hydraulic fracturing. To gain insights on their interplay with hydraulic fractures, readers are encouraged to consider AAPG’s Hedberg Research Conference in July in Casper, Wy. (www.aapg.org/education/hedberg/casper/index.cfm), titled “The Geologic Occurrence and Hydraulic Significance of Fractures in Reservoirs.”