1 - Introduction To MEM's

1.01 - Significance Of Mechanical Earth Models (18 min.) Sample Lesson

2 - Case Study: Duvernay Example

2.01 - Introduction To Duvernay (5 min.)
2.02 - Elastic Properties (22 min.)
2.03 - Unconfined Compressive Strength (15 min.)
2.04 - Confined Compressive Strength (23 min.)
2.05 - Tensile Strength, Friction Angle (6 min.) Quiz: 2.05 - Tensile Strength, Friction Angle
2.06 - Pore Pressure And In-Situ Stress Analysis (26 min.)
2.07 - Frac Modeling & Common Landing (14 min.) Quiz: 2.07 - Frac Modeling & Common Landing
2.08 - Fracture Extension, Net & Breakdown Pressures (7 min.) Quiz: 2.08 - Fracture Extension, Net & Breakdown Pressures
2.09 - Rock Competency Analysis - Part 1 (20 min.)
2.10 - Rock Competency Analysis 2 (6 min.)

3 - Montney Case Study Example

3.01 - Montney Elastic Properties (8 min.)
3.02 - Pore Pressure and In-Situ Stress Analysis (10 min.)
3.03 - Fracture Extension, Net and Breakdown Pressures (5 min.)
3.04 - Summary (1 min.) Quiz: 3.04 - Summary

Mechanical Earth Models / 1 - Introduction To MEM's

Lesson 1.01 - Significance Of Mechanical Earth Models

Access All SAGA Wisdom Training content Subscribe

Already a member? Sign in

Access All SAGA Wisdom Training content Subscribe

Already a member? Sign in

Transcript

01. Lesson 1.01: Significance of Mechanical Earth Models 02. The Mechanical Earth Model 03. What is a Mechanical Earth Model (MEM)04. Optimization Cycle for Unconventional Developments 05. 1D Mechanical Earth Model (Input Data)06. 3D Mechanical Earth Model

01. Lesson 1.01: Significance of Mechanical Earth Models

Hello everyone and welcome back with the new geomechanics series with SAGA. My name is Erfan Amini and I'm a geomechanics instructor. So in this chapter, I'll walk you through the process of construction of a mechanical Earth model that is essentially the numerical representation of geomechanical properties within the subsurface.

02. The Mechanical Earth Model

The mechanical Earth model, or the construction of the mechanical Earth model by far is one of the most important and crucial steps in technical evaluation and optimization of unconventional rock. It has a wide range of applications, for example in shale fracturing, for optimal selection of landing targets, for optimization of the frac job itself, or in optimization of the drilling process (optimizing the key drilling parameters), and well design and many, many other applications. So I recommend this course to everyone who is involved in the technical evaluation and optimization of unconventional rock. So as we'll see later, the Earth model is the initial step and input to many of our models. And in my experience the insufficiently accurate representation of it often will lead to poor analysis of our model and anything that follows. So I'll walk you through the process of construction of the Earth model. We'll discuss the limitations and assumptions and type of data that we need to build a robust model of the Earth model. So let's get started.

03. What is a Mechanical Earth Model (MEM)

So I'll start with a couple of introductory... , kind of setting the stage, and lay out the context for what we'd like to achieve by the end of this course. So I'll start with a couple of basic definitions here. What is a mechanical Earth model? So a mechanical Earth model (sometimes called M-E-M or 'MEM'), is essentially nothing but the numerical representation of geomechanical properties within the subsurface.
There are so many elements and geomechanical parameters in construction of the mechanical Earth model. So when I think about the mechanical Earth model, I try to bucket the properties that are similar in nature, kind of like (for example), the first category is mechanical properties such as Young'smodulus, Poisson's ratio. And I tend to bucket them with properties like strength, like unconfined compressive strength, confined compressive strength, tensile strength, cohesion and shear strengths and some of the properties like internalfriction angle. So that's one group. The second one is on forces, like pore pressure, the in-situ stresses (min, max, overburden) and any other byproduct of this (basic curves like volumetric stress or deviatoric stress) are all bucketed in the same category. The third one relates to the frac'ing itself. So things like frac breakdown pressure, fracture extension pressure, or frac net pressure.
As I mentioned before, the mechanical Earth model has a wide range of applications. For example in optimizing the drilling process. The key drilling parameters like weight on bit, like rate of penetration, torques and drags and things like that. Or even for well design, selection of the proper casing size, or adjusting the mud weight in problematic zones. It has also applications in hydraulic fracturing optimization, proper selection of optimal landing targets for optimization of the frac job itself. It's essentially the input of the frac model. Also in reservoir simulation since it involves the frac'ing itself. If it does, it requires some representation of the geomechanics because the system is dynamic and requires some geomechanics embedded in it.
Now we can also classify an Earth model based on the dimensionality of the model representation. For example, we can have a 1D mechanical Earth model. The 1D mechanical Earth model considers only variation in 1 dimension which is the vertical direction (that's where most of the variabilities are). And then we have another one which is the 3D, is the most comprehensive representation of geomechanical properties in the subsurface, which allows us to represent both lateral and vertical heterogeneity in the model. Now what we're going to focus on today is the 1 dimension, because the 1 dimension is the foundation of construction of a 3 dimensional Earth model, which requires an integration of multiple 1 dimensional MEM, seismic inversions, fault stress analysis, and all of that is a very complex and exhaustive process. It needs multiple cycles of optimization, numerous iterations as well. So just to kind of come back on this, so we going to focus only in the 1 dimensional mechanical Earth model, and we're going to focus only around the reservoir (a little bit above, a little bit under). We're not going to talk about the uphole alot. I have almost one slide which kind of walks you through how to build that for uphole as a very similar process. The reason for that is because I want to focus on the reservoir, because most of the data that we need to build a robust model, they are all coming from the reservoir. If it's core drive geomechanical properties or if it's like diagnostic fracture injection testing (all coming from the reservoir), that allows us to build a robust model. The uphole is a little bit like that. It heavily relies on the drilling experience, but we'll go through that quickly as well in the upcoming slides. So that's #2. And the #3 is that there are many, many commercial software out there that can help you to construct a mechanical Earth model. But I also want to recognize that it's not everyone's passion and expertise to build a detailed mechanical Earth model using the commercial software. So what I did, I wanted to empower everyone to build their own model. So I set up a semi-automatic spreadsheet that helps to get started with something.

04. Optimization Cycle for Unconventional Developments

This slide tries to summarize the way I think the mechanical Earth model fits into that unconventional development optimization. So ideally we have a 3-dimensional geological model, which itself is built off of a lot of data like seismic, the full suite of logs, some reverse geosteering. So an example of that is essentially here. So ideally we do have that. And then what we're going to do is we're going to go in and build the 1-dimensional mechanical Earth models for as many wells as we need, but using geomechanical logs, core drive geomechanical properties,DFIT, drilling data, anything that we have which allows us to build a robust model. Once we have those 1D MEMs that we can integrate into our 3-dimensional geological model, then we can bring in the fault stress analysis, the positional changes, depth and anything that we need to build that complex 3-dimensional representation of geomechanical properties in 3D.
After that, when we feel happy about that, now we can go to the next step. We can do some fracture modeling. And obviously those frac models that we build will have to be validated against some sort of field diagnostics. Some examples might be seismic, fiber optics, tracers, etc., and that helps us to validate it. So if you don't feel happy about it and we feel the frac model is not as calibrated, then we have to go back and adjust our Earth model, because often what happens is that those Earth model representations may not be unique, so they have to be revisited and adjusted. And once we do that, we come back and we run the model again and we compare them against the field diagnostics to see if there's any consistency.
And once we feel happy about that, we can go to the next step which is the production modeling and forecasting. And that also has to be consistent or validated with production data such as rate and pressure. If you don't feel happy about it, we don't see consistency then we go one step back. We do frac modeling, field diagnostics. If you don't feel happy about that, we go back and we do multiple cycles of trials and iterations until we optimize the process. So we also recognize the fact that this process is very long, exhaustive. So what we try to do in this course is to walk you through this process of how we can build a good and robust model, to kind of reduce the number of iterations required to get from the initial point to the last, where we have an optimized cycle for our dimensional development.

05. 1D Mechanical Earth Model (Input Data)

So there's a lot of input data to go into the construction of a mechanical Earth model. This slide tries to summarize what sort of data we require. So we have all those geomechanical buckets, now we haveelastic properties like Young's and Poisson's and shear and bulk moduli. So we need some geo-mic logs such as sonic logs, compressional and shear arrival times, the bulk density, some lab tests. Similarly for strength properties like tensile, UCS, CCS, cohesion, and friction angle. We need DTC (again), DTS, bulk density, some lab tests. For the pore pressure, again DTC, diagnosticfracture injection tests, or things like leak-off tests or XLOT. For the in-situ stresses like overburden, min and max, again DTC, bulk density, DFITs, image logs. So we're going to walk you through each of those steps in detail, but you can kind of see that we need a lot of DTC, DTS, and bulk density. They all are geomechanic logs and these are the building blocks of a mechanical Earth model.

06. 3D Mechanical Earth Model

As I mentioned before, in this course we're kind of going to focus only on the 1D, a 1-dimensional mechanical Earth model that simplifies the Earth surface to a single dimension, which assumes the variability only happening in the vertical direction. As I mentioned before, the 3D is more intensive, more complex, and more comprehensive that accounts for lateral heterogeneities across the 3D domain. What I want to do here, I want to list some of the top drivers of those lateral heterogeneities that we need to be mindful of. And sometimes the 3D lateral heterogeneity may take over the 1D, so we need to make sure that we understand when that case becomes important. So one top driver of lateral heterogeneities are (by far) faults in structures. Perturbation of mechanical properties, especially the in-situ stresses caused by faults and structures. And the way I like to explain it is, if you think about a flat surface which is confined from all angles by some sort of compressional loads, the in-situ stresses will be pretty much uniform laterally, and that's when most of the variations are driven by 1-dimension (the vertical direction).
So let me use this paper to explain what that means. So let's say you have a flat surface here, and all the angles are compressed by far-field (which may be tectonic). And let's say (for any reason), if this flat surface is bent like that, then the in-situ stress distribution will be no longer uniform because of the bending of this surface. Now, one of the drivers of this bending is basically faults in structures. So imagine if we have a broken surface here, and then we have the mountain pushing from one side, and pushing this block on my left side over the other block on the right side, and creates some curving and bending essentially. That causes a redistribution of stresses along this flat surface, locally along the bending. And then obviously we have the fault here, so this is kind of a thrust fault system with the hanging wall being here on the left and the foot wall being here on the right. So the areas that are bended, most likely the rocks will accommodate more in-situ stresses, and the areas that are around the bending will be depressurized and unloaded, creating some sort of stress trough. And as you go away from that point locally, everything becomes equal to the in-situ stresses. So this bending is such that when we sum up all the net effective stresses, that sum has to be equal to 0.
And now why this is so important. This is very important for unconventional development. Let's say (for the case of fracking), if this pen is my horizontal well and we decide to go and frac or complete this well on the hanging wall of the bent surface (let's say like that), so hypothetically you're on the hanging wall here and you're sufficiently away from the bend. And we go and complete this from the toe all the way to the heel. And what happens? Everything is going well. The frac'ing is easy, sand placement is very easy from toe to heel. And we finish the frac job and we flow back, we produce the well. It's probably the productivity is through the roof. It's a perfect environment. Now as we get closer to the fault, things will start to change. So we start again with the toe. We frac from the toe to the heel. Everything's perfect up until the fracs at the heel, will start the field impact of this stress distribution, basically this stress trough. What happens then is that the fractures are going to get drawn to the fault, so we're going to see some asymmetric fractures, potentially some height suppression. The number of induced seismicity events will start to go up, we're going to lose our fractures to the fault and obviously we're going to see some deterioration in our productivity, as a result of this interaction with the fault.
Now we can also move over a little bit. And now we have some overlap with the bending section (let's say). Now again, at the heel we start the frac from the toe. And what happens? The same thing. We're going to get drawn to the fault. We're going to lose the fractures to the fault up until the fracs start to feel the bending or the critically stressed environment. And what happens then? The fracs maybe start shooting up or down. The frac placement becomes very difficult and maybe we start to open up the beddings and parting and all that. And the earth may start to move because of this force on the thrust fault and then maybe some possibility of casing deformation. Now we can move over a little bit. Now the toe has some overlap with the bending. And the same thing can happen again - shooting up or down, it's critically stressed, shooting up or down, lifting the bedding, moving the Earth, some casing deformation potentially. And then now, as we step back and close to the heel, we're going to get surrounded by that stress trough. Now our fracture is going to get drawn to the fault. Same thing, high suppression, losing to fracs at the fault, and again more degradation in our productivity. As we step further and further from the fault on the foot wall, then those sort of effects will go away. So this is an example of how important the faults are, especially if you have some sort of development of fracking near faults. You need to be mindful of those type of changes. So if you want to build a geomechanical model, then maybe 1-dimensional is no longer sufficient. So we have to bring other datasets, do some fault stress analysis and include that into our interpretation.
So that's one big driver of lateral heterogeneity. The second one is more like induced, like parent-child that causes pressure depletion, stress depletion as well. And the depletion itself, again can be caused by the fault effects as we see on the side of the bending stress trough. You may experience some sort of depressurization, natural fracturing that leads to depletion as well.
And the third one is the change in the depositional environment. The depositional environment can change, the depth can change, and the pore pressure stresses can change as well. So in order to understand if 3D actually is important to lateral heterogeneity, we need incorporation of advanced datasets such as seismic, borehole information, or geological surveys. So it's an exhaustive process but is an important process.