- [Voiceover] First, look at some of the inputs and how we pay attention to them. There's some inputs that require a lot of work like the environment or the pollution. When you build or come up with the conception of environment or pollution it is harder. However, there are others mean more work that could be done by the petrophysicist or by the geophysicist. These are all the things they could do also. One key for you here is to always watch the data that goes into the process, and the data that comes out of the process. Make sure that distributions or the variance in distributions is as minimal as possible. That is the key thing to keep in mind. So I talked about the example of the environmental deposition. How is it involving? It's quite highly involving here, because it could involve any range of data, and it typically would use both your well, your seismic, and your regional information. And we rely on principles, some fundamental, some a bit more involved. And with seismic, you create attribute maps, which you use to characterize and interpret your environmental deposition. I have a case study here, which is-- I did this, this is a RGB Blending with an A Channel, also. And so, essentially you have expected comp done on the same volume at three different frequencies and this volume is, uh, is one that was built with minimums and lixiscope. So if we could hide the faults that's I've put on here going from northeast to southwest in this base. And every feature that you see not running from the northwest to southeast direction is a stratigraphic feature. And what we see here is that you can go in there, and you took your features. By the way, this is a new gene delta, so you can, so it allows you to, going to this extent allows you to be able to resolve very fine channels. So this channels a resolve. But how do I know that these channels are important at all? Well, by extracting one more attribute that is a reflection of the extensive measure of the culture of the reservoir, I'm able to show here that when you look at the channels here, lets take a look at this channel here, that ends somewhere around here, that way when you go back here you see it's populated hot yellow, there's sand there. So this is able to allow you to describe your environmental deposition better, and have the evidence to back that up. This is not to be read, it's just a checklist that you can take home with you. If you have all of this information, you're good to go to build your reservoir model. Let's talk about-- thanks to Evanson for this slide about scale, but you're always dealing with data on different scales. Some type of geological information is scale independent. However, some really depends on scale, and you have to be very careful about how useful that is. Seismic velocities have an impact on your reservoir. How much of an impact do they have inside your reservoir that's 50 feet or just 80 feet, compared to the impact they have when you're looking at lateral variation over tens of miles or you're looking at a distance of about 2,000 feet. So, it's important to understand all of these when you're looking at uncertainties, factoring the scale at which they operate or they exist. That's pretty important. And that's the example I was giving about the importance of the scale. For instance, you have a place localized to a little place on your Well Log. If it turns out that, that's because of some dueling process, it's not really something to carry as a certainty for the rest of your field. So you have to be mindful of these kinds on Scale on Uncertainties, and how we'll work with them. I have this slide here, and this information is from Matt Hall, it just shows us the level of integration that we get to be involved in, and there's two dimensions to this. If you're following the horizontal line, and you're going top and you slide that line up and down, what you get to see is the spatial resolution that's what you get to see. Or rather, that's what you get to see with the vertical resolution. And when you're sliding the vertical line left to right, you see the spacial resolution. What's important here, for instance, is to look at all that, your core and your well logs. Look at how many scales of magnet you have to move to get them to get to your grid marker to that dimension of your typical, average gird cell. This grid cell here is a hundred by hundred feet grid. And it's a structured grid rectangle, or rather it's a square. Now compare that, that scale to the seismic. Seismic is fairly regional and the magnitude is fairly wide. And you're moving aside all the way to your right to get to the tip of your model. You have to be mindful of all of these things. After you build the model, you have to QC that model. And what I've included here is the matrix that explains your devolve that we use to QC models. And we'll use this to QC a lot of models in different parts of the world, and it works very well. There are five pillars upon which this methodology is based on. And rather than going to look at every single button a person clicks, there are five pillars. Is this model framed to be effective? Meaning that, is it gonna actually resolve in a solution? Is this framed for statistical, is this statistically consistent? Is it geologically plausible? Does the model have utility? And then we look at the mechanics you click in the button. Behind this you've got, when you look under the hood, you've got a set of questions that help you arrive at the solution. Now, I know a lot of you are leaning forwards and trying to read this slide, and wishing you could see more. So there's more for you here. What you get to see here is, I've got an example here and you see the ranking here for the first three parameters in this QC here. This model is good for framing for effectiveness. It's also very good, because if you consider geologically plausible for that environment or the pollution. It gives an average plus mark for statistical consistency. Maybe many places, there's some parameters that were not fully accounted for. When you have all of this information, what this XPSG Model QC matrix does for you, it allows you to summarize the work you've been doing for months into one simple chart, one simple spider plot here. And what that does for you is if you look at geology, that's geological plausibility, five out of five here. So it comes in here, looks really good here. It's green, that means it's good for that. And if you look here, this chart has three bands of color. The red, the yellow, and the green. Those are thresholds. The black one presents the result for this model. So everywhere the black is close to the green, it tells you that the model has passed the mark. That model is good, it has passed the test, it's met all of the criteria that has been outlined. And everywhere the model is close to the red, it probably requires a second look. Sometimes a model may need you to find out what's wrong with it. So here's a new table there I show the method that we tend to do that XPSG a lot. We do a GiT Analysis, and it's useful. You can deploy this back at your office, break down the process and the gates, and look at every stage in the process. Look for the input and the output. And see what has changed, and that will be very useful.