Think You Know Where Proppant Goes During Fracturing? Based on Recent Testing: Don’t Bet on It

Abstract

Completion engineers have figured out some rules of thumb about how proppant flows based on decades of experience. That knowledge was put to the test when GEODynamics created a fracturing surface test that offered a full-size, full-pressure recreation of a stage design created by its initial backer, PDC Energy. Before the oilfield service company’s first test was pumped. those involved put their money into a pool with the winner based on who most accurately predicted how much sand flowed out of each cluster. “The one that won the bet was the CFO who had no idea how fracs were supposed to work. The worst were the ones that thought they knew fracturing,” said Phil Snider, a consultant on the project who played a key role in the design of the test. Like the pool, the test results diverged from the widespread assumption that the fluid and the sand flow out in roughly equal proportions at each cluster. The results suggested “proppant and fluid does not move as uniformly as many believe,” said Steve Baumgartner, senior engineering technical advisor at GEODynamics, while describing the testing at the recent SPE Hydraulic Fracturing Technology Conference and Exhibition (HFTC). Some of the results were consistent with earlier studies using computer modeling and less-realistic flow tests showing that many fast-flowing sand grains slip past the first few clusters in a stage. GEODynamics found that medium-sized proppant (40–70 mesh) is likely to slip past early stages, resulting in reduced outflow in early clusters and more flowing out later in the stage. But if the grains are smaller (100 mesh), the distribution is more even. A second round of tests showed that a change in the fracture design aimed at achieving more even slurry distribution from cluster to cluster further reduced the differences among clusters, but bigger grains still tended to slip past early clusters. What GEODynamics made public is a first look at an ingenious bit of engineering used for a series of tests that ended in 2019, before COVID-19 hit (SPE 209141). The idea for the test goes back to questions raised by past fracturing jobs. For example, when a well where data gathered during fracturing indicated all clusters were effectively stimulated, but later analysis indicated about half of them failed to produce. Why? “This nonuniformity can be attributed in part to formation variability and stress shadowing from adjacent fracture stages, but nonuniform flow of proppant in the casing can also play an important role,” said a second paper on creating a model for completion engineering. The notion that sand grains and fluid do not move in lockstep does not seem surprising because sand grains are likely to behave differently from a mix of water and friction reducer. The hard question for any engineer who wants to begin designing completions based on the assumption that fluid and sand flows are not similar is how to quantify that difference.