Fixing a Hole and Filling the Cracks (Part 2)

The discussion of the near-wellbore frac geometry in shale plays is now totally dominated by equal distribution across perforation clusters.?Every piece of shale deserves equality in a frac design effort to leave no clusters behind.

In episode 2 of this write-up, I want to focus on the here and now instead of the past.?We have mostly fixed near-wellbore problems at the surface, but presentations at a near-wellbore workshop organized by SPE showed that equally distributing fluid and proppant downhole remains a challenge.?As I demonstrated in my garden last year, distribution of proppant and fluid is about as difficult as designing a sprinkler system.?

I was asked to participate in the final panel by Eric Marshall and join industry icons like Dave Cramer, Karen Olson and Mary van Domelen.?Our task was to review presentations and share lessons learned.?I should have realized there was a heavy price to pay – paying attention throughout the conference.?

I wanted to provide a shoutout to a few presentations I thought helped us gain a better understanding of what we are achieving downhole – first on improved diagnostics, math and modeling tools.

Som Mondal with Shell discussed some of his work on improved stepdown test interpretations.?This is a topic close to my heart after helping develop a practical tool for field use 25 years ago.?With the added luxury of bottomhole gauges, it becomes possible not only to determine the difference between perforation and near-wellbore friction, but also to determine the power-law coefficient beta for near-wellbore friction’s relationship to rate.?In the extremes, this allows and evaluation of a pressure-dependent near-wellbore opening (power law exponent beta at 1/4) or a fixed opening (power law exponent beta at 1).?100s of stepdown tests in various fields show a consistent change from a beta of about 0.5 at the beginning of most jobs toward a beta of 1 near the end of frac jobs.?Erosion is expected to be the main cause of change.?Ultimately, an improved method to measure tortuosity and perforation friction helps to better identify the number of open perforations, which is important for the identification of partial screen-outs and the distribution of fractures from a clustered well.?See SPE paper 204147 for more details.

Kim Wu at Texas A&M discussed progress on micro-strain inversion and our ability to count how many fractures reach an offset well with a fiber-optic wireline.?With information about perforation cluster efficiency at the frac’d wellbore, this can provide us with information about how many of these fractures make it to the next wellbore, allowing more accurate models to tie fracture length to volume-of-first response.

In a revival of water-hammer analysis, Dan Moos of Seismos discussed how the measured reflectivity from a sharp pulse might tell us more about the near-wellbore connectivity.?Stepdown tests have been observed to sometimes change the number of propagating fractures from clusters as rates temporarily drop to zero and some fractures do not come back.?If it becomes possible to relate the measured water-hammer reflectivity to a near-wellbore friction measurement from a stepdown test, quick access to this measurement may be critical to understand injectivity and possibly – stage productivity.

Proppant does not end up where fluid ends up.?It is a big deal for cluster frac’ing.?Proppant inertia has been a focus of attention at various companies and universities.?Mukul Sharma at UT Austin presented a proppant inertia model and reviewed various sensitivities.?Proppant and fluid basically have different preferences that need to be balanced in a clustered stage design: fluid preferentially goes heel-side of the stage as it encounters more pipe friction further downhole; proppant prefers to travel straight while fluid turns the corner into a perforation, thus preferring the toe-side of the stage…unless so much proppant goes there that toe-side clusters screen out. In addition, lowside 0-degree perforating is the best insurance against inertia bias and help a more evenly distribution of proppant with fluid.?Below is a simple example from a fluid dynamics discrete element method for three perforation clusters of 4 perforations each – while it is assumed that fluid distribution is uniform, heel-side perforations receive a lower-concentration slurry and toe-side perforation receive a higher-than-average proppant concentration.

Craig Cipolla with Hess brought some unique direct measurements on another topic that is very close to my heart – net pressure.?In a project where he combined fiber optics with multiple downhole pressure measurements, a combination of in-casing pressure measurement with an outside-liner measurement in the treatment well and another outside liner measurement in an offset well, we arrive at a perfect setting to measure perforation friction, near-wellbore friction and net pressure at the well, AND the net pressure over a distance of the frac.?This is mouthwatering stuff that I cannot share, and hopefully Craig will present it at some point in the future.

Jackson Haffener with Devon discussed fracture model calibration with Sealed Wellbore Pressure Monitoring, SWPM.?Measuring fracture half-length in offset sealed wellbores, as described in Figure 17 of SPE paper 199731, provides detailed volume-to-first-response information to calibrate the relationship between fracture half-length and volume pumped.

Taking this a step further, SPE paper 204140, Figure 12, shows how much Volume-to-First Response (VFR) was needed to reach various offset wells, allowing for detailed fracture half-length and heigh model calibration.

Then there was a paper (of many – the conference was filled with lots of great presentations!) that discussed experimentation with perforation cluster and stage length design choices. Design choices (Paul Huckabee, Shell) for stage lengths, cluster spacing and eXtreme Limited Entry (XLE) perforating are intimately tied.?It appears XLE is not generally necessary, as it causes unstable erosion in subsets of perforations, which than take most of the flow; instead, a modest limited entry perforating pressure drop between 500 – 1800 psi was observed to reach 100% cluster efficiencies in various basins around North America where stage spacing of 250 – 280 ft were considered.?Below is one plot of variations in cluster efficiency in one basin where a 10x change in stage length was investigated, and where a modest 1,100 psi perforation friction successfully ensures a long stage interval is covered effectively.

Finally, in my opinion the best paper was by Ohm Lorwongngan from Hess about the impact on the of all these choices on the bottom line. He discussed production impact of increasing clusters, using XLE to cater fluid and proppant to all of them in several North Dakota test wells, as partially detailed in this URTeC paper. Production was boosted through this strategy while the number of separate stages were reduced, saving as much as 10% on total completion cost – similar to what we have seen in Delaware cluster changes.

I am afraid to say I threw Mary under the bus a few times.?Reminiscent of better times when SNL hosts could still be politically incorrect, it was her turn to throw me under the bus immediately after the broadcast, with her opening argument: “Leen, you ignorant slut, ….”?