Image of the week #3

Another one from the Suez Rift. This is my former colleague Chris Ott (hi Chris!), standing in front of a syn-rift boulder conglomerate containing various pre-rift and syn-rift clasts (the one to the right of Chris is probably a Precambrian basement basalt clast). These are actually some of the smaller clasts in the conglomerate: lower down in the same unit there are limestone blocks that are several tens of metres across, which makes mapping in the area "interesting". The unit is in the steeply dipping limb of a fault-propagation syncline in the hanging wall of a major block-bounding normal fault that was active during deposition. It is interpreted as a debris flow triggered by slope collapse.

Geology students are happy

This article in the Guardian was good to see. Apparently, a survey of some 220,000 students in the UK suggested that 95% of the geologists were happy with their degrees, the highest percentage for any subject.

It was interesting to see the reasons given for this. According to Paul Nathanail of the University of Nottingham "geology...gives students, at an early stage in their studies, the chance to be part of cutting-edge science. If a student sees a new rock, they can begin to challenge the established way of thinking". While I'd like to think this is true, I suspect it's only really a factor for the brightest and most motivated students. I think that the chance to get out in the fresh air and study something with clear applications is perhaps a bigger factor. Whatever the reason, it's a good thing, and should help geology departments recruit more students.

Image of the week #2

This is from my PhD thesis. Although at first sight it probably looks like a cross-section through a fold/thrust belt, the scale is somewhat different. It's actually a scan of a large thin section. The folds and thrusts are a small-scale slump in finely interbedded sandstone and mudstone of the Albert Formation of New Brunswick, Canada. The base of the slump cuts down through the underlying stratigraphy from left to right. Having measured up a large number of folds in similar slump beds, I found that the palaeoslope was roughly perpenicular to the local faults. So these slumps may have formed during fault-induced slope instability.

Oil prices: it's not about the geology

This article by Nils Pratley was in this weekend's Observer. It mainly concentrates on the influence of OPEC, the cartel that controls around 40% of the worlds oil supply. It re-iterates that oil prices generally don't have too much to do with how much oil there is available to produce: it's more about how much actually gets produced, for reasons that tend to be economic and geopolitical as much as geological.

The oil price is back down to $107.30 today, from a high of $147 in early July. That's a 20% fall in just two months. This has little do with the state of oil production, and everything to do with speculation. Now the OPEC countries have got used to oil at $100+ a barrel, they're ready to take action, by cutting production, to keep oil prices high. And, of course, the west needs relatively high prices to drive investment in technically and/or politically difficult fields, now that most of the easy oil has been found.

It's a strange situation. If there was a cartel that was fixing prices in, say, the British airline industry, something would be done about it. But oil is not like other commodities.

One thing in the article that bothered me a little was this:

The Saudis guard closely their data on the oil reserves and production capability. Why? "Peak oil" theorists argue it's because the big reserves aren't as big as advertised.

This makes 'peak oil' sound like a conspiracy theory. In fact, the peak oil hypothesis says nothing about the size of Saudi Arabian reserves. What it says is that global oil production will follow a similar pattern to sub-sets of global production (such as production from single basin, or a single country): that is, it will increase until roughly 50% of the oil has been produced, and then it will fall. Some people, who you might call early peak oil theorists, think that OPEC reserves are overstated, but that doesn't alter the fact that oil production will peak: it just shifts the timing of the peak.

Image of the week #1

This is an idea that I've blatantly stolen from the Clastic Detritus blog. Each Friday I'll put up an image: probably mainly field photographs, but also computer-generated images and such like from my digital outcrop mapping research.


This week's image is a nice example of a fault zone from the Suez rift, Egypt (There will probably be a lot of images from Egypt in this series). This is the Nukhul fault, juxtposing Cretaceous chalk of the pre-rift Sudr Formation against the syn-rift Miocene Abu Zenima and Nukhul formations. A sliver of Eocene pre-rift Darat Formation is caught between the paired slip surfaces of the main fault zone, and is internally deformed. In the hanging wall of the main fault zone (to the right), a series of minor faults occur in a damage zone about 50 m wide. The minor faults tip out downward, with one fault showing a duplex pattern near its tip as it merges into bedding.

Geoscience in the key of Radiohead

Recently Radiohead produced a video for their song "House of Cards", using data acquisition technology similar to that which I'm using for my research. They used a Velodyne LIDAR system to capture images of the band. This gives me a flimsy excuse to write a bit about the use of LIDAR technology for geological mapping in the research group I work with at the University of Manchester.

LIDAR data collection in the Suez rift.

LIDAR (LIght Detection And Ranging) uses a laser beam to scan objects, in our case rock outcrops, and returns series of points that are accurately located in space (within a couple of centimetres) and also have a value for the intensity of the returned laser pulse. This produces what is known as a 'point cloud' dataset. We use a calibrated digital camera to give colour to the points. A differential GPS with sub-metre accuracy allows us to accurately georeference the data. The system that Radiohead used collects up to 1.8 million points a second. The system we use manages 12,500 points a second, which is still fast enough for us to scan an outcrop on a scale of the order of a few hundred metres by a hundred metres in 10 to 15 minutes, with a point spacing of 5 to 10 cm. One of our datasets captures an entire half-graben in the Suez rift, and comprises something on the order of 5 billion points.

Point cloud dataset, coloured using digital images, with geological interpretation added.

Digital Elevation Model derived from one of our Gulf of Suez datsets, and coloured with a 60 cm resolution satellite image. Red dots show the positions of the LIDAR scan stations. The whole dataset comprises more than 5 billion points and covers an area of approximately 9 km2, with a point spacing of 5 to 10 cm.

So far this perhaps sounds like an unnecessarily elaborate system for doing geological mapping. After all, you don't actually need much more than a map, compass and a bunch of colouring pencils to create a perfectly acceptable geological map. So what actual advantages does it give us? It allows us to do accurate geological mapping of surfaces that are exposed in vertical cliff faces, which is difficult to do in other ways because the faces are not easily accessible, and because it is difficult to map vertical faces accurately on horizontal maps or aerial photos. The data is digital, and that allows us to export our geological interpretations to modelling software of the type typically used in the oil industry. This allows us to use outcrop data in a similar way to how sub-surface data (mainly seismic) is used in the oil industry, but with much higher resolution (centimetres compared to tens of metres). In the Suez rift, we're using LIDAR data to improve our understanding of how fault systems evolve through time, and how that structural evolution influences the stratigraphy and sedimentology of the strata that are deposited in the rift. To do this, we need to understand how the thickness and rock types of geological strata change across the area, with respect to the structural geology. Those variations are subtle, and we need high resolution data to map them accurately. Finally, it allows us to easily communicate our data and results to the oil companies who sponsor us, and gives us nice, clear images we can use to illustrate scientific papers and conference presentations.

Three-dimensional representation of a geological surface, derived from LIDAR-based geological mapping

This being a fairly new technology for geological mapping, it has not always been straightforward to use. For example, software for geological interpretation of LIDAR data didn't exist. So we had to create our own: my colleague Dave Hodgetts has been working on this, and the resulting software is now being spun out. And we're still trying to figure out how we can use all the data to its best advantage. But we're getting there, and I hope the first publications from this work will be coming out soon. Watch this space...

Abiogenic oil

This is the first post of what I hope will become a renaissance for this blog. I've tended to neglect it, partly because I've been working hard, and partly because when I do get time to do some blogging, it's much more fun to go moron-baiting over at hawk/handsaw. The intention is to publish something here once a week in future.

This post is about abiogenic oil, a collection of hypotheses by which oil could be created in the mantle, rather than in sedimentary basins at relatively shallow depth. A recent(ish) Hedberg Conference of the American Association of Petroleum Geologists (AAPG) addressed this issue, and the results were written up in the AAPG Bulletin earlier this year.

First of all, why is this important? For two reasons. If significant quantities of oil are produced through inorganic chemical processes in the mantle, it is possible that there is far, far more oil available for production than is conventionally thought. The much-dreaded peak oil could be put off for a considerable period of time. Secondly, current exploration is based on the biogenic origin of hydrocarbons; that hydrocarbons are created during the burial of organic material. This hypothesis makes predictions about where hydrocarbons should be found. If a significant amount of oil really is abiogenic in origin, then we need to urgently revisit how we explore for oil and gas resources.

How could abiogenic processes produce hydrocarbons? There are two possibilities: firstly, degassing of the mantle, followed by polymerisation the low-molecular-weight compounds released. Secondly, the serpentinisation of ultramafic rocks combined with a Fischer-Tropsch reaction, in which carbon monoxide and hydrogen are combined, using a catalyst, to form hydrocarbons. What both these hypotheses have in common is that hydrocarbons are produced deep in the crust or in the mantle, and must migrate considerable vertical distances to reach the relatively shallow reservoirs in sedimentary basins where hydrocarbons are normally found.

Unfortunately, the evidence for large accumulations of abiogenic hydrocarbons seems to be somewhat thin. No-one disputes that non-commercial deposits of abiogenic hydrocarbons do exist, but commercial quantities seem to be elusive. The problem is really one of Occam's Razor; where there are suggestions that hydrocarbon accumulations might be abiogenic, they can also be explained within the biogenic paradigm. For example, hydrocarbons have been found in fractured basement rocks such as granite. At first glance, this seems to be contrary to the biogenic model. On closer analysis, though, these occurences can be explained coventionally. The hydrocarbon source rock is subjected to heat and pressure, generating hydrocarbons, which then migrate via fractures. If the horizon they are generated in is overpressured, they can be forced downward. Alternatively, uplifted areas of granitic basement might have source rocks adjacent to them across faults, for example.

Another problem for the abiogenic oil hypothesis is the presence of age-restricted biomarkers. These are organic compounds that are only produced from certain types of organism. For example, a compound called oleanane is derived only from angiosperms (flowering plants). It is therefore restricted to hydrocarbons derived from Late Cretaceous or younger source rocks. Not only that, but the radiation of angiosperms through time should mean that the proportion of oleanane increases as the source rock becomes younger. This is exactly what is observed in real hydrocarbon provinces. This is consistent with the biogenic origin of oil, and not with the abiogenic origin of oil.

A third problem is the existence of oil shales. These can be explained by the biogenic paradigm: they are the strata that contain the organic material that gets buried, matures into hydrocarbons, and migrates into reservoir units of high porosity and permeability. But the shales themselves are of very low porosity and permeability. So how could fluids migrate from deep in the crust or mantle and accumulate in them?

Does all this mean that there is no such thing as commercial abiogenic oil? Not necessarily. If you wanted to find abiogenic oil in exploitable quantities, you would have to look in areas where crustal-scale fault or fracture systems could allow fluids generated deep in the Earth to migrate towards the surface, perhaps in areas of basement where there are no sedimentary basins. No-one has been looking in those places, because the biogenic model works perfectly well. It would be hugely risky to go looking for hydrocarbons using the abiogenic theory as your guide, but it could ultimately be worth it. Especially if oil prices rise much further...