Duvernay Keeps Pushing Boundaries Of Conventional Geomechanics

September 9, 2017 admin 2 comments
Figure 1. Detailed studies on the Duvernay show that the conventional problem-solving approaches have serious challenges explaining geomechanical behaviour of this play.

At least in some regions, I like to think of the Duvernay Formation as a half-asleep mountain lion that wildly roars back with a steaming breath and trembles the earth beneath when awakened. The aggressive/sensitive behaviour of this formation during drilling and fracturing is because it is extremely overpressured, critically stressed and highly fractured and faulted. This fierce temper, however, has different levels of severity in different regions of the Duvernay making this play a curious case for geomechanical experts. The reality is (similar to some other unconventional plays), with its special characteristics, the Duvernay play has been challenging the status quo of some of the very common problem-solving approaches in petroleum geomechanics. In the following, very briefly, I will discuss some of these observations at a very high level. The provided references contain details for the interested readers*.

Figure 2. Map of Woodbend/Leduc reef complexes, platforms and basins showing the location of the Duvernay play (Source: Switzer et al., 1994).
Figure 3. Schematic cross-section of reef, shelf/platform and basin-fill within Woodbend intervals, west-central Alberta (Source: Switzer et al., 1994).

Where high pressure rules geomechanics

The extremely high pore pressure (PP) (close to twice as large as hydrostatic PP in some regions) in some areas of this play has influenced almost every geomechanical characteristics of this play. For instance, as a result, stresses are in a highly sensitive equilibrium state and can be easily disturbed by underground operations. Likewise rock properties and fracture fabrics in this play are strongly affected by the excessive pore pressure in the formation.


Even minimum stress measurement can be difficult

Pressurized fracturing tests such as DFIT are supposed to provide closure pressure (CP) which is usually considered as the best proxy for minimum stress (read more about pressurized tests in this article on my blog). As its name suggests, CP is the measured pressure when the created fracture is closing. Fracture closure in highly overpressured formations, if ever happens, will take a very long time far more than the patience of the test operators. This is probably the reason that, in several interpretations of CP for the Duvernay, this parameter does not show a significant difference with PP. If valid, this observation (i.e., small difference between PP and CP) can challenge the applicability of common theories of stress characterization such as frictional equilibrium state or poroelasticity. The specific relations between PP and stresses can be related to the fact that in highly overpressured and low permeability rocks, tectonic stresses are mostly borne by the fluid rather than the rock matrix, a topic that I will discuss in a coming post soon.


Seismically at the edge both in-zone and out-of-zone

The proximity to the tectonic front and high PP in some areas such as Fox Creek region have brought the half-dormant faults in the Duvernay on the brink of reactivation and made induced earthquakes with magnitudes as high as 4.4 possible. Surprisingly, several of the recorded induced seismic events are interpreted to occur in the zones other than the targeted play (similar observations exist for some other formations), sometimes as far as several meters deeper. Although some believe this issue could be related to the level of accuracy in interpretation of event locations, some others, accepting the validity of locations, have tried to investigate the potential reasons behind these observations. In these presentation and article, more details on this controversial observation have been provided.

 

Figure 4. A profile of seismicity induced by hydraulic fracturing showing that the events are not confined to the targeted fracturing zone in the Duvernay Formation and occur far deeper. (Source: Bao and Eaton, 2016).

Rock properties are influenced by high pressure

High PP is expected to influence different mechanical characteristics of rock matrix/mass such as stiffness/compliance and strength. Nevertheless, almost all the measurements of rock properties in the lab are performed on dry samples and do not show the effects of ambient pore pressure. On the other hand, dynamic measurements of elastic properties in the field (i.e., using sonic or seismic tools) are highly influenced by both PP and fluid type as shown for Poisson’s ratio in Figure 4a. Now, look at Figure 4b and you will see an abrupt shift in Poisson’s ratio within the Duvernay zones comparing to the adjacent zones. This shift is likely to be caused by three different effects: presence of gas , PP effect on the matrix and PP effect on the fluid (as shown in Figure 4a). To separate these effects and find the static rock matrix/mass responses (which are required for geomechanical studies) more rigorous rock physical models than the commonly used ones (that simply replace gas) are required. See more on the Duvernay’s rock properties in this presentation.

Figure 5. (a) Variation of Poisson’s ratio and P-Impedance with fluid type and pore pressure (b) Stratigraphic variation of log-based dynamic elastic properties in the Ireton and Duvernay units (Sources: Dvorkin; Soltanzadeh et al., 2015).

Alternative brittleness indices

Brittleness Indices have become popular parameters these days especially among geophysicists and petrophysicists who use them for evaluating mechanical characteristics of rocks based on their seismic and logs data (my articles/posts on brittleness try to explore the controversies around this subject). A study on the Duvernay shows that either of the commonly used elastic or mineralogical brittleness indices can not properly explain the mechanical nature of the Duvernay rock and alternative indices might work more efficiently probably because of different reasons including what we discussed about rock properties along with lithological and mineralogical characteristics (e.g., influence of high clay content) of this formation. For more details see this presentation and its corresponding paper.

Figure 6. Mechanical and mineralogical parameters calculated for several wells for the Duvernay and Ireton formations: (a) Rickman’s Brittleness Index, (b) plane-strain Young’s Modulus, (c) mineralogical brittleness index based on dolomite and quartz content, and (d) mineralogical brittleness index based on clay content (Source: Soltanzadeh et al., 2015).

Regional Variation in drilling Experience

If we divide the actively targeted regions of the play to: (i) behind-the-reefs and (ii) other locations which are not obstructed by the reefs, the general drilling patterns in these areas show interesting differences. While the wells behind the reefs can be drilled with lower mud weight and without serious drilling issues, the wells in other regions experience serious problems (e.g., tight spots, kicks, etc.) even when drilled with high mud weights. This difference becomes more interesting if we remember that, generally, the pressure behind the reefs is larger than the other non-obstructed areas. Unfortunately, the common borehole stability models based on poroelasticity can hardly explain this difference and it seems that different modeling approaches are necessary. This paper provides a detailed account of the story of drilling in the Duvernay.

Figure 7. Differences between different regions of the Duvernay showing the influence of reefs presence on the geomechanical character of the play.

Fractures that vanish behind the reefs

Duvernay is naturally fractured almost in all the targeted areas. The reasons are simply the high pore pressure and proximity to the tectonic front/belt. Many of these fractures are sub-vertical fractures that are consistent with the present-day stress regime. Nevertheless, a detailed study on the fracture fabrics (see this presentation and paper) shows that among different fracture fabrics,there are some sub-horizontal polished ones that only exist in areas which are not obstructed by the reefs. These fractures are probably formed by the influence of high maximum horizontal stress (SHmax) magnitudes. The disappearance of these fractures may bring us to the conclusion that SHmax is probably smaller behind the reefs than other non-obstructed regions as will be discussed in the following.

Figure 8. Different fracture fabrics in the Duvernay Formation. Figures (e) and (f) show the polished fractures discussed here. The formation of these fractures is speculated to be related to the high SHmax values influencing zones with high bitumen. (Source: Soltanzadeh et al., 2015)

In-situ stresses seem to be affected by the reefs

Every geomechanics expert knows that finding SHmax is one of the most hassling and uncertain exercises in geomechanics. In absence of practical direct measurement methods, applying traditional approaches (i.e., poroelastic modeling, reverse borehole stability or frictional equilibrium) with all their limitations can provide an idea of the ranges of this important stress component. A study based on these methods in the Duvernay shows that the behind-the-reefs regions have lower SHmax gradients and lower stress anisotropy than other areas. It seems that the stress state behind the reefs is influenced by two different factors: higher pore pressure and the role of the stiff reefs as shields that have probably relaxed the effect of tectonic forces on the less stiff Duvernay. These results are consistent with the drilling experience and fracture distribution in the Duvernay, as discussed above.

Figure 9. Parameters influencing different geomechanical characteristics of the Duvernay Formation such as in-situ stresses, mechanical properties and fracture fabrics.

Much more to learn and explore

It seems that the untamed plays like Duvernay with their complex natures and unexpected behaviours can provide great opportunities to enhance and evolve our existing practices in classical petroleum geomechanics, a discipline that has been developed mainly around conventional reservoirs/operations and burrowed its materials from other disciplines with different objectives and materials such as geotechnical engineering, mining and soil mechanics. In addition, what discussed above shows the value of regional studies with integrated multi-source geomechanics in providing a broader perspective of our characterization methods for unconventional plays.

Figure 10. Multi-source Integrated Geomechanical Characterization in regional scale can provide a much broader perspective for our understanding of a target play.

The good news is, during the last few years, attention to geomechanics has significantly increased as we are seeing it with the increasing number of tests and logs performed by the operators. For instance, in a Duvernay’s regional geomechanical characterization conducted in 2016-17, we found that the number of geomechanical data had exponentially increased compared to to another regional study from 2013-14*. Though this was partly due to the growing number of explorations and developments in the play, the high number of wells chosen for geomechanical data acquisition is clearly showing that geomechanics importance and value for field development is acknowledged more than ever. This special attention from the industry is a great opportunity for petroleum geomechanics to review, criticize and improve its conventional methods and theories. Fortunately, many outstanding and highly knowledgeable experts in the academia and industry are advancing the frontiers of petroleum geomechanics on an everyday basis although more effort, especially in revising the fundamental assumptions and theories used in regular practices, is still necessary.

Going back to my discussion of the Duvernay, as a step forward, PetroGem Inc. is planning to implement its experience of the Duvernay in developing an integrated geological/geochemical/geomechanical numerical model in a basinal scale. This basinal model will use advanced numerical techniques to focus on depositional, tectonic and hydrocarbon generation history of the play, very different from what has been done so far. With detailed calibration with the actual data, this study is expected to provide more clarification on the complex behaviour of the Duvernay and its adjacent formations. PetroGem welcomes all the companies and organizations who might be interested in joining, helping, supporting or collaboration with this study (email: info@petrogeminc.com).


* I must acknowledge that several of these findings and observations are based on my experience in leading geomechanical modules of regional Duvernay studies conducted by Canadian Discovery Ltd., a company that has shown a great respect for the role of science in its studies.

 

 

 

2 Comments on “Duvernay Keeps Pushing Boundaries Of Conventional Geomechanics

  1. Thanks for sharing your great findings. Since you have mentioned dynamic logs have been affected by fluid, pore pressure and etc. (which can not be detected in cores and static values), I really interested to know about rock physical methods have been used to calculate exact values of rock mechanical properties.
    You can find my email address below. Regards

    1. Hi Sahar, there are rock physical methods such as Gassmann’s model that can take care of replacing the fluid content (e.g., gas to water). The effect of pore pressure/stresses on dynamic/static properties, however, is more complex and it is usually studied by performing tests such as compressibility. These effects are usually studied for the purposes of 4D Seismic Monitoring or to account for the effect of stress disturbtion on dowhnhole tool measurements (e.g., sonic logs) and unfortunately the models in this area are less developed so far. As I mentioned, it is important for rock physics to come up with models that can account for all these effects at the same time. Below is a link to a paper of mine that provides an example of the tests used to study the effects of pore pressure variation on static and dynamic properties. https://www.geoconvention.com/archives/2016/294_GC2016_Lab_Invesigation_of_Static_and_Dynamic_Mechanical_Response.pdf

Leave a Reply

Your email address will not be published. Required fields are marked *