THE WOODLANDS, Texas— High school math teachers were right—geometry does end up being needed later on in life. At least, it does for oil and gas operators.  

Fracture geometry plays a significant role in the evolution of flow paths and is important to fracture execution and subsequent post-treatment well performance. At SPE’s Hydraulic Fracturing Technology Conference on Feb. 7, industry executives looked for new ways to crack the code in evaluating frac geometry. But before the problem can be solved, some concessions have to be made.

“We're never going to perfectly describe fracture growth in the subsurface,” Craig Cipolla, Hess’ principal engineering advisor, told the audience.

Even so, operators are doing their best to describe a well’s fracture geometry through the copious amounts of data they gather.

“You can't model anything if you don't have some measurement of the thing you’re trying to model,” Kyle Haustveit, manager of Devon Energy Ventures, added.

Some of the diagnostic data gathered includes offset pressure, microseismic logs, geochemical data and PVT (pressure, volume and temperature) data. Using this data, operators have created models to achieve the “optimum frac design,” Michael Mayerhofer, director of technology at Liberty Energy, told the audience.

However, problems pop up when trying to measure frac geometry. Even with measurements and data, effective propped geometry is rarely measured nowadays. Because of this, a big question that came up for Mayerhofer during his research was whether the proppant’s effectively produced geometry will scale with the measure of created fracture length.

“It seemed like they had a fairly constant factor or ratio of that effectively producing length or geometry versus the created one,” he said. “So that would be very important, some kind of a reference catalog that we can hopefully generate as an industry in each basin and get an idea of what those ratios could be, because it’s much more cost effective to do diagnostics that just measure the hydraulic geometry and then you can just scale the effectively producing one with that ratio.”

Cipolla believes it isn’t “unreasonable” to at least get close because of available measurements such as lateral length, volume curves, height and microseismic data.

“We’ve got this suite of measurements—a pretty large toolbox—but we missed something that is trying to help us understand what’s really producing, what's the conductivity, and even more important than that, what is the conductivity profile along this producing fracture,” Cipolla said.

Mapping proppant has been an industry-wide goal since proppants were first implemented in the late 1940s. Operators attempting to map fracture geometry have used fracture diagnostics, which involves analyzing data before, during and after a hydraulic fracture treatment to determine the shape and dimensions of the created and propped fracture.

“Calibrated fracture and reservoir models are required to generate more significant value from frac diagnostics,” Mayerhofer said. “An understanding of how models perform when conducting sensitivity studies is critical.”

Cipolla noted the need to upgrade the proppant transport model.

“I think that advancing proppant transport models and tying in some of these relatively gross but valuable field measurements that we’re seeing more and more might be the low hanging fruit that we’re missing,” he said. “We’ve talked about it and we’ve been working on it off and on for a long time, and is very, very difficult. But now at least we’re understanding what the range looks like.”

Karen Olson, SPE’s completions technical director, agreed, describing a future where sensors make that idea commonplace.

“I keep envisioning a sensor that we can pump that’s small enough and it says, ‘I’m here, and by the way, I’ve got this much pressure on me and it’s this hot and it gives you an X, Y, Z location.’ It’d be great if we had little sensor balls that we can throw in the sand and send it back and tell us where things are.”

Dreams such as these don’t feel that far from reality to Olson. The industry has the necessary tools to make it happen.

“I don’t think it really would take that much effort because we have so much computing power,” she said.