The message from a single chart’s data from the first full-scale hydraulic fracturing surface test is simple: Far less proppant flows out of the first clusters passed in a stage than the last ones.
The likely explanation drawn from a surface test created by GEODynamics is that the momentum of those relatively large sand grains prevents them from making that turn early on, leaving a lot of sand for later stages where the slowing flow will make the turn easier.
“The flow is going 45 miles an hour; sand particles have to turn in three-eighths of an inch,” said Jack Kolle, senior technical advisor at a sister company, Oil States Energy Services. He made the comment during a presentation about modeling fracturing test data to create an engineering model of proppant flows during at the recent SPE Hydraulic Fracturing Technology Conference and Exhibition (SPE 209178).
When he first heard about the test results, Kolle thought he could create a model by using basic fluid mechanics concepts. After he started working on a model, though, he realized that proppant flow was more complex than expected. “It quickly became apparent we could not explain it without CFD modeling,” he said.
He was referring to computational flow dynamics (CFD) which requires massive amounts of computing power to model complex flows such as the flow of air around an aircraft wing. In the past it has been used in studies that concluded that the fast-moving flow of water and sand during fracturing resulted in uneven distribution of water and sand.
The model he created based on data from GEODynamic’s unique surface testing setup and subsurface fracturing analysis evolved into the company’s fracturing-flow advisor program, StageCoach.
Based on a quick look at four charts in a paper about the testing, it appears that larger‑grained proppant is far more likely to slip past early clusters than smaller grains, which tend to be distributed evenly among clusters. And fracturing designs that more evenly distribute the slurry among the clusters can further flatten the distribution (SPE 209141).
The results favor some established trends. The industry has embraced 100 mesh proppant for fracturing and limited-entry designs which, to varying degrees, ensure more-even distribution.
What constitutes limited entry has evolved. A 2019 paper on the first two rounds of testing preceded current stage designs using clusters with only one perforation per cluster, often at the top of the hole.
The test work supported the rule of thumb that shooting perforations at the bottom of the casing is a bad idea. The thinking has been that a perf gun lying on the bottom of the casing and shooting at point-blank range will create a larger hole than shots made from a distance. When fracturing begins, bottom holes will take in far more of the fluid and proppant, causing rapid wear which is magnified by the force of gravity.
It wasn’t something they were trying to test. They shot down to ensure all the fluid and sand would flow into a tank below. During the process they found it drained the larger-sized proppant slurries so effectively in the heel-side clusters that there was little sand left at the end of the stage in the toe-side clusters.
Kolle said that when it comes to perforations at the top of the pipe, gravity might limit the volume of sand going into those holes. The outflow might be optimized if the perforations were in the middle, a bit below the three or nine o’clock positions.
The paper can be read as a series of thoughts on fluid and sand flow based on the proppant-transport surface tests and downhole fracturing analysis by the oil companies that partnered with GEODynamics.
The method used for the analysis—Eulerian multiphase computational fluid dynamics (EMP‑CFD)—was chosen because it is able to account for the differences in the flow of sand relative to water.
- It is observed that proppant placement within each stage can be highly nonuniform.
- Nonuniform flow of proppant in the casing can be as important as formation variability and stress shadowing.
- Fine sand is distributed relatively uniformly throughout the length of a perforated completion while coarser sand tends to slip past the heel perforations and concentrate on the bottom toward the toe of the completion.
- At high axial-flow velocities, the slurry exiting the perforation is drawn from a relatively small, semicircular region of the flow—the ingestion area.
- The ingestion area is proportional to the ratio of flow through the perforation to total flow in the casing.
- Sand particles are observed to settle toward the bottom of tubing during water-slurry flow at velocities comparable to those used for proppant placement
- Modeling turbulent multiphase flows of particles in viscoelastic fluids, such as those containing concentrations of friction reducer, is beyond the capabilities of current multiphase CFD codes.
- For particles negotiating the turn into a perforation, the inertial forces are orders of magnitude larger than they are for gravitational settling. Hence, friction-reducing (FR) polymers may reduce the slip of proppant past perforations but to a lesser degree than the reduction in gravitational settling.The optimum selection of FR loading for uniform proppant placement remains unanswered and will only be resolved by further testing on the scale of the proppant-transport surface tests.
For Further Reading
SPE 209178 Modeling Proppant Transport in Casing and Perforations Based on Proppant Transport Surface Tests by Jack Kolle, Oil States Energy Services; Alan Mueller, ACMS; and Steve Baumgartner and David Cuthill, GEODynamics.
SPE 209141 Execution and Learnings From the First Two Surface Tests Replicating Unconventional Fracturing and Proppant Transport by Phil Snider and Steve Baumgartner, GEODynamics; Mike Mayerhofer, Liberty Oilfield Services; and Matt Woltz, PDC Energy.
