Horizontal hydraulic fractures in shales: are they real?

In ResFrac, we are always challenging ourselves—what should we be doing better? What new capabilities should we add to the simulator?

One of our newest projects is adding horizontal fracture propagation. Under most conditions, hydraulic fractures form vertically, not laterally. However, in specific circumstances, horizontal fractures develop. Sometimes, they form in addition to vertical fractures, and sometimes, they form exclusively without any vertical fractures. Horizontal fracture propagation has not conventionally been included in commercial hydraulic fracturing simulators, but we think this is a capability well-worth developing.

Please let us know if you would like to be involved in guiding and developing these new capabilities. We are assembling a group of operators who want to collaborate with us on this topic.

Unconventional reservoir rocks are naturally highly laminated [1], see Fig.1. What is the effect of these layers on hydraulic fracture propagation? Many authors have addressed this question, but mostly from the point of view of arresting a vertically growing hydraulic fracture [1-3], asking, “How does rock layering affect vertical hydraulic fracture height growth?” An equally important issue is whether these weak interfaces can be reactivated and mechanically opened. This process leads to the formation of horizontal hydraulic fractures, which can significantly change the overall behavior of the fractured rock formation.

Figure 1: Laminated shale formations, adopted from [1].
It is commonly assumed that horizontal fractures occur at shallow depths (less than roughly 1000-2000 ft) [4] because the vertical stress is smaller than horizontal stresses for such conditions. In fact, shallow horizontal hydraulic fractures have been proposed for energy storage [5] as well as for preconditioning of rock for mining purposes [6]. Can they occur deeper? Hydraulic fractures do not ‘sense’ depth, they ‘sense’ the state of stress. If the state of stress is such that somewhere, even locally, the vertical stress is the minimum principal stress, then the answer is yes, horizontal hydraulic fractures can form.

An early paper that discussed deep horizontal fractures is [4]. The authors provide tiltmeter data that demonstrates the formation of horizontal hydraulic fractures during conventional petroleum operations. A case study from the North Shafter Field confirms the formation of horizontal fractures at 7500 ft depth. Stress reversal occurs due to abnormally high horizontal stress. Another example is presented for the Belridge oil field, in which EOR changed the state of stress such that the vertical stress became smaller than its horizontal counterpart, which led to the formation of horizontal fractures. These case studies demonstrate that the formation of horizontal hydraulic fractures is possible and observable. In some cases, the vertical stress is the minimum principal stress, and horizontal fractures form at the onset of stimulation. In other cases, the initial state of stress causes vertical hydraulic fractures, but operational procedures lead to stress reversal and, consequently, horizontal fractures form.

A relatively recent paper by [7] discusses the formation of horizontal hydraulic fractures in the Marcellus Shale. The difference between the initial minimum horizontal and vertical stress is estimated to be small. The authors present microseismic results for several wells. Results indicate that vertical hydraulic fractures are formed at the heel of the well. Towards the toe, the fracture height growth becomes substantially confined. The authors interpret that this occurred because of the formation of horizontal fractures. In this case, the horizontal fracturing and height confinement probably occur because of exposure to an overlying high-stress formation, where the vertical stress becomes less than the minimum horizontal stress. In addition, the observed ISIP values were greater than the vertical stress at the zone with strong height containment. This seems to be a good indicator for the formation of horizontal fractures in the absence of other diagnostic data. However, care must be taken when interpreting ISIPs because they can be affected by complicating factors, such as near-wellbore tortuosity.

Yet another example is presented in [8]. Fig. 2 shows microseismic results from the Eagle Ford formation that indicate typical hydraulic fracture height growth and do not show any obvious signs of horizontal fractures. The pad did not produce to expectations, so a vertical well was drilled and cored a year later. The core indicated the presence of multiple horizontal fractures and one propped (!) horizontal fracture, as shown in Fig 3. The location of the vertical well was somewhat away from the main microseismic cloud, which indicates that the opening of the horizontal fracture was aseismic. One possible explanation for the formation of the horizontal fractures, in this case, is that accumulating stress shadow from the toe stages of the pad reverted the state of stress locally around weak interfaces and caused the horizontal fractures. This example demonstrates that we can have a situation in which both vertical and horizontal hydraulic fractures co-exist and that there can be no obvious signs of horizontal fracture formation.

Figure 2: Pad configuration and observed microseismic events [8]. A vertical core well is shown by the black circular marker. Figure courtesy of Roberto Suarez-Rivera.

Figure 3: Propped horizontal fracture observed in the core from a vertical well [8]. Figure courtesy of Roberto Suarez-Rivera.
So far, we have discussed evidence that horizontal hydraulic fractures exist, that they can often occur at shallow depths, that they sometimes occur at larger depths in conventional reservoirs, and also, they occur in unconventional reservoirs. They can also be ‘induced’, i.e. caused by field operations if the initial conditions are favorable. The next question is, why do we care about horizontal hydraulic fractures? And why do we want to model them? For the same reason we do everything at ResFrac… to maximize operators’ rate of return. Below, we list five specific reasons why horizontal fractures can affect economic performance.

  • The vertical permeability of shales is known to be much smaller than that in the horizontal direction due to fine lamination. Therefore, the creation of surface area by horizontal fractures will likely contribute less to production than vertical fractures.
  • Horizontal fractures limit height growth, which reduces exposure to the desired formation.
  • Proppant transport in horizontal fractures is expected to differ from vertical fractures. It is unlikely that proppant will flow as far in horizontal fractures because of the close proximity of the lower fracture wall and the possibility of duning. At the same time, as shown in the experimental study [9], flow instabilities and fingering can enhance proppant transport. Overall, proppant transport in horizontal fractures is not well understood, and this topic deserves further study.
  • Horizontal hydraulic fractures may cause hydraulic connections between neighboring stages on the same well or different wells. The strength of these connections and how they affect production are largely unknown and require further investigation.
  • Horizontal fractures can cause casing deformation. It is shown in [10] that type III casing deformation (slippage along a fault or bedding interface) is the most common type of casing failure. While the majority of such failures are due to faults, bedding interfaces or horizontal fractures contribute to a significant portion of the statistics. In particular, according to data presented in [10], 21% of all casing deformation cases in China (Sichuan and Qaidam Basins), 37% in Canada (Montney and Duvernay), 11% in the US (Anadarko, Appalachian, Midland, Powder River, and Western Gulf Basins), and 21% in Argentina (Neuquen Basin) are due to activation of bedding planes. Therefore, understating conditions of formation of horizontal fractures is vital for avoiding undesired casing deformation and failure. Horizontal fractures can cause casing failure in the vertical part of the well, which can be catastrophic for the whole well or, if several wells are affected, for the whole pad! This is worse than casing failure near the toe of the well, in which case only a fraction of the well is affected.

To recap, and to answer the question raised in the title of this blog post – yes, horizontal hydraulic fractures are real. Are they important? Absolutely. Are they widespread? Probably not. This phenomenon is likely only significant in specific geologic settings. However, they appear to play a meaningful role in a not-insignificant number of field datasets.

Fiber optics can be used to confirm the presence of horizontal fractures. In particular, if a vertical monitoring well is used, then a horizontal fracture would have a very strong signature that is similar to a vertical fracture crossing a horizontal observation well, i.e. a heart-shaped signal. At the same time, if a horizontal observation well is used, then the signal should be similar to a vertical observation well sensing a vertical fracture, i.e. it should outline the fracture front location. Although, this signal will probably be much smaller in magnitude compared to vertical fractures crossing the horizontal fiber. Usually, we do not see these sorts of observations in the fiber data.

How can we learn more about horizontal fractures? How do we mitigate potential negative effects or even exploit positive attributes of horizontal fractures? We will be addressing these questions with ResFrac. Fluid selection, landing zone, cluster spacing, pump rate, and more are all controllable design parameters that will impact the frequency and magnitude of horizontal fracture propagation. Just as there is no one optimal well spacing, the best design to mitigate or exploit horizontal fractures will vary across and between basins, acreage positions, etc.

If you are interested in participating in these new developments, getting early access to model functionalities, and collaborating with others working on this topic, contact us at [email protected]. We can provide more information.



[1] R. Suarez-Rivera, W.D. Von Gonten, J. Graham, S. Ali, J. Degenhardt, and A. Jegadeesan. Optimizing Lateral Landing Depth for Improved Well Production. Paper presented at the Unconventional Resources Technology Conference held in San Antonio, Texas, USA. 10.15530-urtec-2016-2460515, 2016.

[2] Ghazal Izadi,Daniel Moos, Leonardo Cruz, Michael Gaither, Laura Chiaramonte, Scott Johnson. Fully Coupled 3-D Hydraulic Fracture Growth in the Presence of Weak Horizontal Interfaces. Paper presented at the SPE Hydraulic Fracturing Technology Conference and Exhibition, The Woodlands, Texas, USA. SPE-184854-MS, 2017.

[3] D. Chuprakov, O. Melchaeva, and R. Prioul, 2015. Hydraulic Fracture Propagation Across a Weak Discontinuity Controlled by Fluid Injection in “Effective and Sustainable Hydraulic Fracturing”, edited by Andrew P. Bunger, John McLennan and Rob Jeffrey, ISBN 978-953-51-1137-5, Published: May 17, 2013.

[4] C. A. Wright, E. J. Davis, L. Weijers, W. A. Minner, C. M. Hennigan, G. M. Golich. Horizontal Hydraulic Fractures: Oddball Occurrences or Practical Engineering Concern? Paper presented at the SPE Western Regional Meeting, Long Beach, CA. SPE-38324-MS, 1997.

[5] A. P. Bunger, H. Lau, S.  Wright, H. Schmidt. Mechanical model for geomechanical pumped storage in horizontal fluid-filled lenses. Int J Numer Anal Methods Geomech.,1349–1372, 2023.

[6] R.G. Jeffrey, K.W. Mills. Hydraulic fracturing applied to inducing longwall coal mine goaf falls. Paper presented at Pacific Rocks 2000, Balkema, Rotterdam, 2000.

[7] A. A. Alalli and M. D. Zoback. Microseismic evidence for horizontal hydraulic fractures in the Marcellus Shale, southeastern West Virginia. The Leading Edge, 2018.

[8]  R. Suarez-Rivera, W.D. Von Gonten, J. Graham ,S. Ali , J. Degenhardt, and A. Jegadeesan. Analyzing the dependence of well production on lateral landing depths. Invited presentation, University of Texas, Austin TX, 2016.

[9] D.K. Lowe, J.L. Huitt. Propping Agent Transport in Horizontal Fractures. J Pet Technol. 18 (06): 753–764, SPE-1285-PA, 1966.

[10] J. A. Uribe-Patino, A. Casero, D. Dall’Acqua, E. Davis, G. E. King, H. Singh, M. Rylance, R. Chalaturnyk and G. Zambrano-Narvaez. A Comprehensive Review of Casing Deformation During Multi-Stage Hydraulic Fracturing in Unconventional Plays: Characterization, Diagnosis, Controlling Factors, Mitigation and Recovery Strategies. Paper presented at the SPE Hydraulic Fracturing Technology Conference and Exhibition, The Woodlands, Texas, USA, SPE-217822-MS, 10.2118/217822-MS, 2024.

Learn why both independents and supermajors alike trust ResFrac