Isotope studies furnish evidence of the source of CO2 in certain natural-gas reserves, and of the long-term retention of such gas in unexpected environments such as ancient continental crust.
How long can natural gases be retained in geological formations deep underground? From a study of carbon and helium in such gases, described by Ballentine et al. on page 327 of this issue 1, it seems that they can be stored for much longer than has generally been assumed. This finding has implications for identifying and exploiting underground fluids, notably new reserves of hydrocarbons such as methane.
Hydrocarbons are generated in specific geological environments, the source rock formations. They then migrate underground until they encounter a natural barrier, the cap rocks, which prevent further movement upwards. There they may remain until the barrier's geometry is changed by tectonic activity and they are released. In a tectonically stable area, the limiting factor for hydrocarbon storage is believed to be their rate of diffusion through the cap rocks. In geological terms this diffusion is thought to be swift, so the gas should not be retained for long geological periods. But Ballentine et al. argue that in the area they studied, hydrocarbons, along with CO 2 and rare gases of different origins, have been accumulating and stored for up to 300 million years. The region in question consists of limestone formations known as the west Texas Permian basin. The researchers conclude that continental areas such as this, which are old and stable, could be fruitful targets for hydrocarbon exploration.
Accumulations of hydrocarbons, especially natural-gas reservoirs, often contain CO2. This can lower the economic value in two ways. If there is a lot of CO2 , there will be less hydrocarbon, and there are additional costs involved in separating the two. So the aim of oil companies is to locate high-quality deposits of natural gas in the Earth's crust, for instance by identifying the origin of gaseous components and how they were incorporated into the hydrocarbon reservoirs. There are two possible origins for the CO2: it may have been derived from the underlying mantle through degassing of magmas; or it may stem from thermal degradation of carbon-rich rocks, such as limestones, in the crust.
To find out which is the case, analysis of the isotopic composition of the carbon present is not very helpful. Carbon has only two isotopes, and the isotopic signature of mantle carbon lies between those of carbonate (limestone) and of organic carbon (methane, for example). Likewise, the chemical composition of natural gases is of limited value. It allows the conditions in which the gases were generated to be inferred, but not the origin of their component elements. Fortunately, natural-gas deposits contain rare gases, such as helium, which have proven to be good fingerprints of origin. They are inert, and so are not involved in the chemical speciation that occurs in hydrocarbon formation. And their source — whether in the atmosphere, ground water, continental crust or mantle — can be identified reasonably well from their isotopic composition.
So, together with the relative proportions of He to CO2, helium isotopes provide a unique way to identify the origin of CO2 (refs 2, 3). The isotopic composition of helium, 3He/4He, in the mantle is three or four orders of magnitude higher than that of pure crustal gases. This is because primordial 3He trapped during the Earth's formation is still present in the mantle, whereas the crust is degassed in mantle volatiles and is rich in radiogenic 4He produced by the radioactive decay of uranium and thorium isotopes. As a result, the CO2/ 3He ratio of the mantle is (1–2)×109 whereas that of crustal gases in stable continental areas is much higher — around 1011–1013.
Natural gases from the west Texas Permian basin contain from a few per cent to as much as 97% CO2. Ballentine et al.1 have made precise measurements of the 3He/4He and CO2/3He ratios in gases from several wells in the area, in an attempt to determine the origin of the CO2 and explore the paths and processes by which it became mixed with the hydrocarbons. They found 3He/4He ratios higher than those characteristic of radiogenic production in the crust, and CO2/ 3He ratios close to the mantle ratio, much lower than typical crustal ratios. So they conclude that this CO2 is magmatic in origin.
But how are gases originating in the Earth's mantle transferred into the crust? The most plausible process is magmatism — roughly speaking, deep-seated volcanic activity — and Ballentine et al. have attempted to identify the potential areas of magmatism that have contributed CO 2 to the west Texas Permian basin. Magmatism occurred during the development of the Basin and Range Province, which lies 100 km or more west of the Texas basin. But Ballentine et al. argue that this is an unlikely source because the CO2/3He mantle signal increases towards the south of the Texas Permian basin, suggesting a different pathway for gas migration. Instead, they propose that mantle-derived CO2 entered the basin during extension and uplift of the area around 300 million years ago, which is believed to be the last large tectonic episode to affect the region. In their view, hydrocarbon generation began 280 million years ago — that is, after CO2 had entered the basin.
Several reports indicate that natural gases, especially hydrocarbon-rich gases, are often associated with volatiles derived from the mantle. But such an association generally occurs in basins that have experienced tectonic activity comparatively recently (during the Cenozoic, the era running from 65 million years ago until the present). Examples of such basins are those that have been affected by Alpine tectonics, such as the Rhine graben or the Pannonian basin in Hungary3,4, or those formed in subduction zones where tectonic plates slide down into the mantle, such as the Green Tuffs in Japan5,6. If Ballentine and colleagues' view that gases can be stored in much older, currently stable continental areas is correct, it means that there may well be appreciable amounts of hydrocarbons in such regions waiting to be discovered and exploited.
More generally, their conclusion supports the geochemical view that fluids (ground water, as well as natural gas) can be stored in geological basins for very long periods of time, rather than the results of hydrodynamical modelling which suggest that underground fluids move comparatively rapidly in geological terms. The key issues in these debates are the variation in permeability of geological barriers and the diffusivity of fluid species through such barriers — both parameters are only poorly known and are difficult to assess over geological time periods. Further case studies that address these matters would be most welcome.