It is generally understood that the types of completion designs utilized in early shale wells (e.g. wide cluster spacing, high viscosity fluid systems, poor perforating charge selection, etc) led to EURs that are significantly less than what is achievable today using modern techniques. In such wells it is clear that reserves were "stranded" downhole, unable to be produced. Recognizing this, operators are increasingly choosing to restimulate ("refrac") their older producing wellbores in hopes of liberating those stranded reserves from their subterranean bondage. However, refrac operations are complex and expensive, and attempting one involves not only the risk of a production uplift that is uneconomic, but it places the existing parent reserves at risk as well. The high stakes involved merit careful prior consideration, and an essential part of such planning is preparing an estimate of the volume of additional hydrocarbons that might potentially be recovered. This is not always an easy task, however, as properly executed, mechanically isolated refracs within a reasonable offsetting distance are sometimes scarce. To enable economic evaluations in such situations, Barba and Villarreal (URTEC 2622 2020) and Barba (2015) proposed that the added recoverable hydrocarbons from a refracturing operation on a producing well can be estimated from the difference between its current cumulative production and what can be expected with a modern tight cluster spacing high proppant loading completion. Often times, however, there are an insufficient number of modern offset wells from which to calculate an average EUR, and in these cases Barba and Villarreal proposed that a reasonable estimate of that expected EUR could instead be made based on the product of OOIP in the expected drainage area and the expected Recovery Factor with a modern tight cluster spacing high proppant volume completion. Simply put, this difference (between the current cumulative production and the expected refrac EUR) should in theory be the "size of the prize" that one might hope to recover if the refrac was successful. For instance, if a Gen 1 well has a cumulative production of 150 MBO, and the average EUR of the modern offsets are 500 MBO, then it seems reasonable that the most one could hope to add to the current cumulative production in a refrac would be 350MBO. Or put another way, one would have no reason to expect a liner refrac, with a limited ID for much of the measured depth, tremendous variations in closure stress behind pipe, two strings of pipe to shoot through, and a number of other real operational headaches to outperform a best-in-class new well that suffers from none of these conditions. In fact, we know that no process in the real world is 100% efficient and therefore even if a refrac were designed and executed flawlessly, the best one could hope for would be to approximate a new well's EUR less whatever losses would be caused by various inefficiencies.
This paper presents results from a recent refrac campaign in the Eagle Ford that tests this "size of the prize" framework for predicting production uplift in a properly executed refrac. The post refrac recovery factor far exceeded the average of offsetting recovery factors in modern completions – designs that utilized high intensity slickwater completions, tight cluster spacing, optimal perforation charge selection and orientation, etc. In fact, looking back these refracs would easily compete for limited capital with new drill economics. We present five refraced wells that averaged 72.6 BO/ft EUR with a 13.2% total recovery factor, along with seven offsetting new drills that averaged only 44.9 BO/ft EUR and an 7.4% recovery factor. The new drills averaged 3458 lb/ft and 54.4 barrels of fluid per foot with a 16.9 ft average cluster spacing. The refracs averaged 2184 lb/ft, 44.8 bbl/ft, and 19.6 ft cluster spacing. This raises an important question: where are these extra reserves coming from? How can a refraced well exceed the results of the best in case offsetting new wells with higher loadings and closer cluster spacings? It seems that there must be an additional mechanism at work in a refraced well that is not present in a new well. We propose that one such possible mechanism could be fracture reorientation from pore pressure reduction in the near wellbore region due to the removal of reservoir fluids during the interim years of production (Elbel and Mack 1993). If the pore pressure is materially reduced, especially in reservoirs with a relatively small difference between the magnitude of the maximum and minimum principal stresses, reorientation is likely. In such situations, even though the azimuth of the horizontal wellbore was likely selected to create transverse fractures during stimulation, it is possible that if restimulated, fractures will reorient longitudinally (to varying degrees) in the in areas of the near-wellbore region where the pore pressure has been reduced (such as in between perf clusters from the 1st treatment). Such reorientation would, in theory, significantly increase the total Stimulated Rock Volume. The increase in recovery factors observed in these five refracs serve as powerful evidence that something else is going on besides simply accessing stranded reserves, and the authors propose that increased SRV as a direct result of fracture reorientation is likely the root cause.