Shale Tech: Provocative concepts in fracturing

ALI DANESHY, Shale Technology Editor

Background. Provocation is an essential part of the creativity process. Faced with a mental or practical observation that contradicts and challenges the accepted norm, our mind is provoked to seek alternative explanations. Provocation can be deliberate, or compulsory, and forced by necessity. In either case, the process opens new avenues for learning and innovation.

In initial stages, provocative thinking does not need to be rational or satisfy traditionally accepted boundary conditions. Sooner or later, it needs to establish its own new rationale and satisfy old and new boundary conditions that conform to the existing observations. In this process, the novelty of an idea is not by itself sufficient proof of its validity. It needs verification and proof by observation.

Discussion. The widespread use of horizontal well fracturing has led to many observations that contradict and challenge the established theories of fracturing that are based on solid mechanics, laboratory experiments and our experience with vertical wells. In this article, we will review some of these concepts and offer alternatives that are intended to provoke our thinking in new directions. This is intentional provocation. All of these explanations, or suggestions, are intended to be provocative, different than the accepted norm and, hopefully, pushing us outside of our “comfort zone”.

Fracture growth pattern. In the early 1950s, the initial theory of fracturing postulated that industrial hydraulic fractures are horizontal and created by lifting of the overburden. This was not based on any sound theory, but “reasonable” speculation based on existing knowledge of geomechanics. The theory was soon contradicted by field observations that failed to show the expected communication between nearby fractured vertical wells. The next generation of theories were based on solid mechanics, supported by laboratory experiments. The laboratory experiments were often designed specifically to minimize deviations from theory (including those conducted by this author). The new theory postulated single fractures created by tensile stresses that propagate in a plane perpendicular to the least in-situ principal stress. This led to extensive laboratory experimentation and development of other new and supporting theories that articulated aspects of fracturing, such as the role of fluid mechanics and proppant transport inside the fracture, measurement of fracture conductivity, computation of production increase, computation of fracture geometry (length, width and height), fluid leak-off, etc.

Gradually, these theories established themselves as the accepted mechanisms describing each aspect of a fracturing treatment. The large volume of laboratory experiments and resulting technical publications, all based on and in support of the single-fracture theory, gave it an air of invincibility and accuracy. The practical outcome of the new theories was development of many new tools, techniques and practices, with the net effect of substantially improving the productivity of our reservoirs and the growth of the fracturing industry. Perhaps the most telling impact of the new fracturing theories is the fact that in the past 50 years, more technical papers have been written about hydraulic fracturing than any other single technology or practice in the oil and gas industry. Thus, the new theories not only met our scientific objectives, but also created useful means of increasing the productivity of our reservoirs and profitability of the companies that supported and funded their development.

In the past few years, and especially in connection with multiple fractures created in horizontal wells, the past theories of fracture propagation have proven inadequate to explain the contradictory observations made during day-to-day practices. These observations have led to the development of new theories of fracture growth, called “complex” and/or “off-balance”, which have basically replaced the single-fracture theory. Both off-balance and complex growth theories use the formation heterogeneity as their foundation and the explanation for their validity, Fig. 1.

Fig. 1. Fracture extension models

The complex fracture growth pattern has more degrees of freedom and can allow the fracture to grow in many directions, including a crisscrossing pattern. Its primary justification is presence of natural fractures in the formation that divert the hydraulic fracture from its natural path (perpendicular to the least in-situ principal stress). Another justification is the relatively high production rates generated from ultra-low-permeability shale reservoirs. The argument is that without a complex network of fractures, these formations could not produce at the observed high rates. This argument, while reasonable, does not provide a direct and rigorous proof for the existence of a crisscrossing “complex” fracture pattern. Indeed, such fracture pattern contradicts the fundamental rules of solid mechanics. As a way of overcoming this serious objection, some authors have postulated a truly provocative explanation that the two horizontal principal stresses may be equal. This claim is not supported by accepted direct and independent stress measurements. Nevertheless, because of its high potential impact, the concept of complex fracture propagation pattern needs more study before it can be ruled out. Furthermore, if we reject the complex fracturing theory, then what do we offer in its place to explain the productions achieved in shale gas reservoirs.

I have a fundamental objection to the use of the term “complex”. The term indirectly implies a problem that defies our comprehension and is inherently beyond our technology reach and, therefore, it should be accepted because of its “complexity”. This implies a technical resignation and defeat that does not foster active pursuit of an explanation, or shows a path for building new technical knowledge. For this reason, I have chosen to avoid using the term “complex”. Setting the wording aside, the proposition that the two horizontal principal stresses may be equal is truly provocative and can have significant consequences in our daily oil and gas production activities. With stress difference removed as a variable, formation mechanical heterogeneity takes the center stage as the dominant parameter in controlling formation behavior. While this author has not yet seen a single case overwhelmingly suggesting equal horizontal principal stresses, one needs to keep this as an open option deserving of more consideration and review.

The “off-balance” growth pattern has less freedom to wander around. The term, first used and defined by this author, describes a fracture that grows in a clear direction, along a band perpendicular to the least in-situ principal stress, but with varying degrees of associated branching and secondary shear fractures. Off-balance fractures have been observed to exist in the field, and their presence has been explained by existing fracturing and solid mechanics theories. In some papers, the term “complex fracture” is used without any clear definition of what it means, although their drawings and schematics depict an off-balance fracture. Other graphic representations for complex fractures show a large crisscrossing network of fractures (as shown in Fig. 1) that are far different than what is defined as off-balance in this article.

A logical outcome of both off-balance and complex growth patterns is the expectation that actual hydraulic fractures should be shorter than computed based on simple fracture theories. Comparatively, the off-balance fracture is expected to be longer than the complex fracture, mainly because it has less freedom to wander around. Intersection of fractures created in offset horizontal wells provides a strong clue to the length of the created fractures. I am aware of several cases of communication between horizontal wells separated by over 1,500 ft (and occasionally even farther). In fact, the distance between some of these wells along the fracture direction can be in excess of 2,000 ft (because of the oblique angle of the fractures with respect to the wellbore). This implies fracture lengths greater than 1,000 ft. When reviewed together with the magnitude of injected volumes, these lengths indicate low levels of scatter along the fracture, which is more supportive of the off-balance growth pattern.

Regardless of which idea is correct, the subject has tremendous impact on our effective and efficient use of hydraulic fracturing. For this reason, further investigation of the topic is highly recommended, together with an open mind to accept the observed results, even if contrary to one’s own initial beliefs.

Fracture closure. One direct outcome of the simple fracturing theory is that fractures close when the fluid pressure creating them falls below the magnitude of the least in-situ principal stress. Pursuit of this point has led to many technical publications that attempt to determine the exact time of fracture closure, and also the pressure at which the fracture closes. Further extension of these calculations has been development of methods to compute other fracturing parameters; such as fracturing fluid leak-off coefficient, fracture length, and even formation permeability. In reality, the closure of hydraulic fractures is not a single event and is shown to be gradual, partial, and incomplete, as evidenced by the partial recovery of the fracturing fluid itself, and the fact that one can get production enhancement even without the use of any proppant. There is even strong anecdotal evidence suggesting that low-recovery factors sometimes yield better production results. Both complex and off-balance fracture growth patterns allow and can cause incomplete fracture closure, while also providing good production results. There is even strong anecdotal evidence suggesting that low recovery factors sometimes yield better production results. Both complex and off-balance fracture growth patterns allow and can cause incomplete fracture closure, while also providing good production results. The provocative thinking here is the idea that fractures do not completely close, they self prop and we do not even need to use a propping agent! It is a fact that good production results have been achieved without using a proppant. The question is “does the use of proppant improve our production results?”. Consider fracturing without the worry of screen-out! Would this author then advocate elimination of proppant from the fracturing operations? This is a provocative question which this author is not prepared to endorse at this time. The reason is the fear that with time, gradually and slowly, the shear fractures may close and thus the fracture loses its conductive conduits. Furthermore, the author has field evidence showing total closure of parts of the fracture shortly after pumping stops. If so, the proppant has the positive role of keeping more of the fracture open.

Proppant transport. If fractures do not completely close, then what is the role of proppant in hydraulic fracturing? Is proppant responsible for directly providing a permeable path for flow of reservoir fluid, or does it serve the secondary role of keeping the fracture open so that reservoir fluid can flow mainly through highly permeable channels kept open by proppant packs (Fig. 2)? Many years ago my colleague, the late John Tinsley, came up with the patented idea of “pillar frac”. In a pillar frac, the proppant served as “pillars” inside the fracture that kept it open so that reservoir fluid could flow through the “rooms” kept open by the pillars. His method of creating a pillar frac was to pump alternate slugs of proppant-laden slurry and clear fluid so that the proppant in the slurry created the pillars and clear fluid created the rooms. In fact, I had developed the design computations for pillar frac. After several field applications of the process, the production output of wells fractured with pillar frac were not measurably better than conventional fracturing. Was this because the real fracture growth pattern itself actually created pillar fracs without any need for changes in pumping design?

Fig. 2. Proppant deposition inside the hydraulic fracture

Existing theories of proppant transport, based on simple fracture extension theory, postulate very organized and sequential proppant movement inside the fracture. Each slug of slurry moves behind the previous slug, and ahead of the next one. Accepting the orderly proppant movement has been the foundation and reasoning behind our present slurry injection practices; starting with low concentration of smaller-size proppant, and gradually increasing size and concentration to accommodate larger production flowrates expected nearer the wellbore, and even use of resin-coated tail-in proppant to prevent proppant flowback. There are plenty of field observations that contradict this mode of proppant movement. Among them, flowback of non-resin coated proppant, and flowback of mixture of proppant sizes used in different stages of pumping, including the very first. Radioactive tracer sand logs also show near-wellbore presence of mixture of various types of sand pumped at different stages of fracturing.

Perhaps the most compelling contradictory evidence for orderly proppant transport inside the fracture is the occurrence of screen-out. If proppant were to move inside the fracture as postulated by theory, then the only way it can screen out is if it reaches very high concentrations at the fracture tip. But this causes two contradictions of its own. First, in many actual fracturing treatments, the rapid pressure rise seen at the time of screen-out indicates a near-wellbore event. Second, screen-out at the tip will expose the entire fracture length to high screen-out pressures. No formation rock can sustain such high pressures without further failure. In fact, if one considers proppant deposition/blockage inside the fracture in the form of pillars, then one can easily explain the composition of returned proppant, as well as near-wellbore screen-out.

Accepting the concept that proppant can get blocked inside the fracture and form pillars has significant operational consequences, some of which can create treatment simplification and cost savings. For example, if reservoir fluid is flowing mostly around the proppant packs (instead of through them) then the objective of fracturing becomes formation of proppant packs, not proppant bed permeability. If so, the need for proppant bed permeability loses much of its significance and appeal. Another consequence of this is the reduced need for high-strength proppant, since even crushed proppant can serve the role of pillar and keep the fracture open. The urge to use different size proppant also becomes much less significant. The net result will be substantial reduction in cost of fracturing operations. It should be noted that, as provocative as these statements may be (as they are intended to be), some of these are already being practiced while fracturing horizontal wells. I am aware of very unorthodox proppant pumping schedules (size as well as concentration), use of very low-strength proppant in deep formations, and pumping whatever proppant is available to get the job away (in the face of material shortage) without an outcry of measureable negative production effects.

One consequence of fracture closure and proppant transport theories has been the push to avoid over-displacing the proppant inside the fracture. The fear is that without proppant to keep the fracture open, it will close near the wellbore, thus choking well production. As a matter of fact, over-displacement is a standard practice in horizontal well fracturing and has not been shown to harm the well’s production. The fact is that rocks are solid materials and do not close like “rubber”. A fracture supported by pillars can stay open tens of feet without any support. Presence of pillars also provides the open channels that are required if proppant is to flow inside the fracture and into the wellbore! Another idea challenged by this provocative view of fracturing is practice of forced fracture closure. If fractures are held open by proppant pillars, would rapid recovery of fracturing fluid be helpful?

Fracturing fluid. Simple tensile fracturing theories led the industry towards development of higher viscosity fluids to carry the proppant deeper inside the fracture. In the 80s and 90s, high viscosity fluids were the choice fracturing materials. Introduction of horizontal well fracturing in very low-permeability formations forced reconsideration of this practice, initially driven by the high cost of these fluids. The substitute, commonly called “slick water” (a mixture of water plus friction reducer), made it possible to pump very large fluid volumes at affordable costs, especially in ultra low-permeability shale gas reservoirs. Slick water had been developed much earlier and was commonly used in marginal gas reservoirs. However, horizontal well fracturing of shale reservoirs changed the significance of slick water. In these formations, the production results of slick water are often better than treatments performed with more expensive viscous “gels”, so much so that today the dominant fracturing fluid in shale gas production is slick water.

The inability of water to carry the proppant deep inside the fracture can very quickly be demonstrated by a simple laboratory test. If so, what causes the unexpectedly high production rates from many of our shale reservoirs? More important, what provides fracture conductivity when fracturing is performed with pure nitrogen, without any proppant at all? The answer may simply lie in the fracture propagation pattern. The shear fractures and some of the branches created in off-balance growth pattern do not close after the end of pumping. If so, then the remaining open fractures can easily transmit the produced fluid through the fracture and into the wellbore.

Conclusions. It is important to point out, once again, that the intent of provocative ideas presented in this paper is to push the reader out of his/her comfort zone and force re-examination of what we have come to believe as “truths” in hydraulic fracturing. Whatever new concepts and ideas emerge as a result of this re-thinking and re-evaluation will eventually have to pass the test of conformity with established fundamentals of geo- and solid mechanics and field observations before they can gain technical legitimacy.

Note. This paper is based on the works of many others who have published on the general topic of fracturing. To take best advantage of the available space, and because of the large number of articles that deserved to be cited, I have intentionally avoided the use of references.