- I'd like to continue with discussion of the cost of heterogeneity in productivity in naturally fractured reservoirs. So far, we've talked about the heterogeneous results of the sampling that we do of a subsurface natural fracture system, from the standpoint of drilling into it with a borehole. That's a rather simplistic view, where you consider either hitting fractures, or not hitting fractures, but indeed, there's quite a range of variability in fracture characteristics, just due to the size and geologic processes that have acted on the fracture system since it formed. So, I'd like to discuss some of those right now, and look at the geologic factors that lead to reservoir performance, heterogeneity. Let's begin taking a look at an outcrop of a thick sandstone bed. The horizontal layers that you see here correspond to bedding planes. The vertical fractures are highlighted here, and I just highlighted a few of them to call your attention to the fractures. A discussion might have more than one, so when a reservoir engineer asks me the question, "What's the average fracture spacing? "I need it for the simulation model." This is always a puzzling question to answer for me or other geologists, but really, the question is a reasonable one, but it's one based on the tools we have at hand, it's not really based on the actual geology of the fracture system, in most cases. And usually, the tool that we have at hand that's motivating the reservoir engineer's question is a dual porosity model, which is a rather abstract and simplistic representation of a natural fracture system. And for the most part, it assumes a fair amount of homogeneity in the model, when in fact, as we've been discussing here, there's quite a bit of heterogeneity. If we look at the height distribution of the fractures, you see here that we have a lot of short fractures, and a very asymmetric distribution with very few tall fractures, and indeed only one very tall fracture way off on the right-hand end of this fracture height histogram distribution profile. What this can lead to is significant heterogeneity, you can well imagine if you drilled a borehole and hit that big fracture, you might get very good flow, very good drainage from the reservoir, or, on the other hand, you may have a lot of water that comes up into the borehole. As opposed to the result, if we hit a much smaller fracture from further off to the left side of the outcrop here, the production characteristics could be very different. So, here what we're looking at is a vertical aerial photograph of a sandstone bed, in fact, it's the same sandstone formation that we looked at in the previous slide. Looking down here, we're basically looking at just one layer in the sandstone. You can see the scale here shows that this is a fairly large exposure, there's multiple fracture sets present, let's just look at the fracture set that's exposed here that runs from the northeast to the southwest, and if we look at a histogram of the length distribution of this fracture set, what we see is a very strongly asymmetric distribution, very similar, in fact, to the height distribution that we saw in the previous slide. That is, a lot of rather short fractures, and a strong taper and a long tail indicating longer fractures as well, so a very heterogeneous distribution of fracture lengths, which is also going to play into our heterogeneous productivity of the reservoir, depending upon not just whether we hit a fracture or not when we're drilling into it, but indeed, what the size of the fracture that we're intersecting is. So, what we're finding, from what we see here, is the frequency distributions of length and height are both similar in the sense that they're strongly asymmetric in general. It turns out that the spacing of fractures is also generally nonlinear, quite often, it'll have a shape maybe more like a lognormal shape. Aperture distribution varies in a strongly asymmetric fashion, not unlike the way that we see frequency and length distributed. So we have strong asymmetry in each of the geometry factors that describe our fractures in a subsurface fracture system. Not all fractures are effective for fluid flow. When I use the term effective, that's what I mean, is effective conduits for delivering fluids throughout the formation in the subsurface. The way we judge whether a fracture in a borehole is effective is to compare various logs to see whether the fracture shows any indication of flow, such as a flow meter log, PLT log, a flowing temperature log, maybe lost circulation right at that interval in the well. Stoneley wave log, PF log under some circumstances, all of these can help us to understand whether individual fractures are effective. So let's look at the record from a well, I'm showing a plot near the center of the screen here, it shows a plot of 122 fractures that were measured on a borehole image log, the plot itself shows depth on the y-axis, and on the horizontal or x-axis, it shows DIP of the fracture. So the markers that are plotted here are the individual fracture planes that were interpreted. You see the blue ones are open fractures, the red are healed fractures. That is, healed with secondary mineralization. The tail on these points points in the direction of the DIP direction of the fracture, so we call this a tadpole plot, because the markers tend to look like tadpoles, or baby frogs. Big head, little tail. The overall strike orientation of the fractures on this plot are shown underneath the tadpole plot, and you see, it shows that the fracture strike overall is northwest-southeast in orientation. Now, this shows that there are certainly a large number of open fractures in this well, but when we go and qualify these fractures to try to identify which are effective fractures, it's a much smaller number, in fact, it's the seven fractures that are seen in the right-hand tadpole plot. So only seven effective fractures out of 122, that's something on the order of 5% of the total fracture population, are actually delivering fluids to the well bar. And you see that even the strike orientation of the fractures is different, it's northeast-southwest, compared to that total population. All of this indicates that there's a high degree of heterogeneity among the fractures that are present in the subsurface. Why should this be? Why should we have only a small number of effective fractures when we have so many open fractures? Well, let's remember what we saw when we looked at the length of fractures in the aerial photograph, and what we see, as you recall, is most of the fractures are very small, and a small number of them are large. Now, I don't know how big any of the fractures in this tadpole plot are, when we see fractures on a borehole image log, we can see what their height is within the borehole, but even that, it's usually nowhere near their true height. So what I'm suggesting is perhaps the effective fractures are the large fractures in a distribution like this. Or perhaps the effective fractures are the only fractures that are connected to other fractures that eventually form part of an extensive fracture network in the borehole. It's quite hard, it's impossible to know, in general, based on the data that we're able to collect now at the present time in the subsurface for fracture systems. So why should there be only a small number of fractures that are effective for fluid flow when we have such a large population of open fractures in this well? And this is typical of many wells, where it's a decided minority of fractures that are effective. And I think it's just that the effective fractures are either the big fractures, or the fractures that are well connected in a network of fractures that extends throughout a significant part of the reservoir, but it makes it hard when we have so many fractures, to really understand what's controlling the fluid flow within our reservoir. But for sure, what we want to do is know more about the effective fractures, and indeed, most experts who interpret and characterize naturally fractured reservoirs try to focus on these effective fractures, and to gain a better understanding of them in order to understand the controls on fluid flow within the reservoir.