3.10 Seismic exploration – part 2

Oil is a massively important product in today's society. We use it for fuel. It's in plastics. It's all around us. So this week we've come here, to Imperial College School of Mines, to talk to Dr. Rebecca Bell about how we find it.

So, Becky, uh …

[LAUGHS]

…why have you bought us a cake?

So, a cake actually provides a really good way to demonstrate what seismic data is and why we need it. So, going to an area of hydrocarbon exploration without seismic data is rather like having a cake like this, covered with icing, and trying to guess what flavour the cake is, how many layers the cake has, if it's got jam or cream in the middle. It would be very difficult. Take a guess?

I'm going to go with coffee and walnut.

OK, coffee and walnut. Seismic data, though, provides us with the means to have a kind of – to take a virtual cross-section through the earth, to take a look at the layering of the rock, get some idea or some sense what the lithologies might be, if there are any structures, and, on the rare occasion, actually be able to see the hydrocarbons directly. So acquiring seismic data is rather like cutting through this cake. So we can then expose what the layers in the Earth looks like.

So we can now see that we've got three different layers. We’ve got the brown one, a kind of beigey one, and a white one, separated by some white stuff, and then some brown layering at the bottom. At the moment, though, we don’t know if this one is chocolate or if this one is caramel or this one's vanilla. We can make a hypothesis, but we don't know for sure until we taste it. Seismic data is exactly the same. So, seismic data has limitations. It’s really useful to see the different layers in the earth, get some sense if there are folds or faults. We can start to make some hypotheses about what the rocks are. You might be able to hypothesise that maybe the bottom layer is chocolate, but you don't know before you eat it.

Right.

So seismic data really works best where we have a well drilled into the rock, where we can take core, and then we know exactly what the rock types are. But the seismic data allows us to decide exactly where would be the best place to drill.

OK!

Should we go ahead and …

Taste it now. Sure!

… look at some real data?

REBECCA BELL:

Let's do that, after our cake.

OK.

[LAUGHS]

So, my job involves mainly interpreting seismic data. So, seismic data is data that's collected using sound waves, where we produce sound waves which travel through the earth. They get reflected off rock types. We detect those reflections, and we start to build up a picture of what the earth looks like beneath our feet. Depending on the time it takes those reflections to be detected, we could start to build up a picture of all of the different layers, based on the time they take to come back to us. So that gives us an approximate image of what the earth looks like.

So what are we looking at here?

OK. So what we're looking at are different reflections from different rock layers. So, wherever you see a blue or a red line, that means we've got a boundary between a rock which has a different density and velocity from the one above or below it.

So don't think of this data as just a series of 2D lines. We actually have 3D data everywhere within the volume. So we can, in fact, scroll through this data set any way that we like to take a look at the structure.

And we're actually looking inside the earth, here.

We are, here, yes. So this is the seabed, this top reflection. And we're going down, here, to about six kilometres' depth – something like that.

So these are different layers of rock within the earth.

Exactly. So don't think of these as individual rock units. This isn't a rock unit, it's the boundary between two different rocks. So, can you see where the faults are?

So that's these bits, here, where the different layers are offset, right?

Exactly. Yeah. So you can see this really bright red is clearly that bright red. They look the same. But it's been dropped down by probably about 100 metres maybe, maybe more. On this side of the fault, the stratigraphy's been tilted backwards. So you can see everything dipping to the right of this image. But, very curiously, we've then got this horizontal feature which completely crosscuts that dipping stratigraphy, which is a little bit strange. So probably, in this environment the most likely explanation for that is that horizontal feature is a fluid contact.

So we're actually seeing the oil in the earth, right here.

Yeah, this is a fairly rare situation where we have what's called a "direct hydrocarbon indicator".

So the next step is that we need to try and characterise how big this reservoir could be. Is it big enough to actually be worthwhile getting all of our drilling equipment, paying millions of dollars to actually go and bother to drill? Or is it just too small and pathetic and not worth it? Just because we see a flat spot doesn't mean it's economically viable. What we would need to do, firstly, is to interpret the top of that reservoir. So, trace that reflector out throughout the entire 3D data set, so we can start to see its 3D geometry.

And this is what we've done on this map. So we saw that there was a nice flat spot here. And now we've interpreted that top reservoir horizon, we can see very nicely that we've got this area where the contours are all closed. And this is called a "four-way closure". So we're effectively making a kind of bowl shape with the top of the reservoir. So any gases that exist here will end up travelling upwards. They want to go as shallow as they can, because they're light and they rise. And they'll all end up accumulating in this bowl. So looking for these four-way closures is really key to working out where would be the optimal place to drill your well.

And have these been drilled?

They have been drilled, yep. So there has been a well drilled, in fact, exactly into this location.

Particularly with 3D data, now we're not restricted to just drilling on the location where we have data, we can actually map and make a geological model and decide exactly where is the best place to target. And since the advent of 3D data, there have been far less dry wells.