Contents - Previous - Next


This is the old United Nations University website. Visit the new site at http://unu.edu


Options for sequestering CO2

The most-discussed option for CO2 disposal has been piping CO2 to depths greater than 3 km in deep ocean basins. So doing would eliminate the rapid transient build-up of CO2 in the atmosphere and delay equilibration with the atmosphere by several hundred years; resulting interactions with calcite-rich sediments would probably reduce the long-term (> 2000-year) atmospheric concentration by a significant amount (~50 per cent) (Wilson 1992). But many questions remain about the dynamics of the processes involved, and many environmental issues have been raised - including concerns about the effects on ocean life of pH change from CO2 injection, and the impacts on benthic organisms and ecosystems as hydrate particles are deposited on the ocean floor. Much more research is needed on such issues before deep ocean disposal can be pursued with confidence that the environmental risks are acceptable (Turkenburg 1992).

Among other options, disposal in depleted natural gas and oil fields stands out as being especially secure, as long as the original reservoir pressures are not exceeded (van der Burgt et al. 1992; Summerfeld et al. 1993), and potentially low in cost (Blok et al. 1989; van der Burgt et al. 1992; Koide et al. 1992; Hendriks 1994). The global sequestering capacity associated with past production, proved reserves, plus estimated undiscovered resources is estimated to be 410 GtC for natural gas fields and 105 GtC for oil fields (Hendriks 1994). For comparison, global CO2 emissions from fossil fuel burning totalled 6.0 GtC in 1990.

The capacity of natural gas fields to sequester carbon at the original reservoir pressure is generally greater than the carbon content of the original natural gas and depends on the depth of the reservoir, the geothermal gradient, and the pressure gradient. Hendriks (1994) has shown that, for typical gradients, the ratio of carbon in CO2 to that in the original natural gas is 3.0, 1.8, and 1.4 for depths of 1, 2, and 3 km, respectively, and that worldwide, on average, about twice as much carbon can be stored (as CO2) in depleted reservoirs as was in the original natural gas.

If a hydrogen production facility could be sited near a depleted natural gas field the costs of long-distance pipeline transport of CO2 could be avoided. Moreover, the cost of injection into the reservoir could be offset to some extent by recovery of additional natural gas. When primary production of natural gas at a reservoir ceases, it is not because the contained natural gas has been exhausted but rather because the reservoir pressure falls below a certain level (typically of the order of 30 bar) at which it is no longer economic to continue pumping out natural gas. But with CO2 injection, the reservoir is depressurized, so that enhanced natural gas recovery is possible. Because of the order of 80 per cent of the natural gas in place is recovered in primary production, the amount of additional natural gas that can be produced is not large - but the enhanced production could pay for part of the incremental costs of CO2 storage.

Based on data characteristic of large natural gas reservoirs in the Netherlands, Blok et al. (1997) have assessed the prospects for enhanced natural gas recovery with CO2 injection, the implications for the net costs of CO2 sequestration, and the impacts on the costs of producing hydrogen or methanol from natural gas. For the reservoir conditions they studied, extra natural gas production would contribute between 3.5 per cent and 8.5 per cent of the natural gas feedstock requirements for a hydrogen production plant over the first 15 years of its assumed 25-year operating life, depending on whether there were large or small permeability differences, respectively, between adjacent layers of the natural gas reservoir. The material balances for their analysis of the case where there are small permeability differences are shown in figure 6.1. In this example, the reservoir is capable of sustaining a CO2 injection rate of 15,000 tonnes per day. Injection begins after 20 years of primary natural gas production, by which time the reservoir pressure has fallen to 30 bar, from the original pressure of 350 bar. During the next 15 years there is enhanced natural gas recovery that makes it possible to provide about 8.5 per cent of the natural gas requirements for the hydrogen production facility, as a result of reservoir depressurization via CO2 injection. (This corresponds to the case with "high credit for enhanced natural gas recovery" in table 6.4.) The reservoir pressure reaches about 130 bar after 15 years of CO2 injection. The recovered gas is about three quarters natural gas, the rest being CO2. It is assumed that the small level of overall contamination (< 5 per cent of the total gaseous input to the hydrogen production plant is CO2) is readily managed in the hydrogen production process. The hydrogen production plant is sized to match the depleted reservoir capacity for accepting CO2. For details, see Blok et al. (1997).

Fig. 6.1 Wellhead production of hydrogen from natural gas with injection of the separated CO2 into a depleted natural gas field and enhanced natural gas recovery as a result of reservoir depressurization (Source: based on data developed in Blok et al. 1997)

There would be considerable capacity in depleted natural gas fields for sequestering CO2 recovered from decarbonizing fuels other than natural gas. For example, when hydrogen is produced from natural gas, the CO2 recovery rate -10.7 kg C/GJ of produced hydrogen (see table 6.1) or 9.0 kg C/GJ of the natural gas from which it is derived (assuming all energy inputs are provided by natural gas) - is equivalent to just two-thirds of the carbon in the original natural gas (13.5 kg C/GJ). Since, on average, the CO2 sequestering capacity is equivalent to about twice the carbon contained in the original natural gas, the production of hydrogen from natural gas would thus leave about two thirds of the sequestering capacity available for CO2 derived from other sources. In addition, the sequestering capacity associated with past natural gas production and future production that will not be associated with the manufacture of hydrogen and the sequestering of the separated CO2 would be available.

The limitation on strategies for CO2 sequestration in depleted natural gas and oil fields is their limited geographical availability. This should not be a significant constraint on hydrogen manufactured from fuel conversion facilities sited near the natural gas fields, because the produced hydrogen could be distributed long distances via pipeline at acceptable costs (see table 6.7 below and Appendix C), serving markets just as natural gas pipelines do today. However, it would be desirable to have sequestering options that are more widely available as well.

The most widely distributed reservoirs for potential sequestering of CO2 are saline aquifers located deep below the earth's surface underlying most of the area of sedimentary basins throughout the world. The areal extent of these basins is equivalent to nearly half of the land area of the inhabited continents (see table 6.3). Aquifers are porous underground beds, consisting mainly of sand, that are permeable to the flow of fluids. The pore spaces are usually filled with water and, occasionally, with petroleum or natural gas as well. In order to be able to store the CO2 at high supercritical fluid densities, only aquifers deeper than 750 metres are considered as potential storage reservoirs for CO2. Aquifers containing fresh water are normally found at much shallower depths. If CO2 were stored at a depth of 750 metres or deeper, it would generally take 2,000 years or more to reach a freshwater reservoir, and even then it will probably reach the freshwater reservoir in low concentrations; the main effect of the CO2 that would enter the freshwater reservoir would be to increase the hardness of the water, because carbonates will dissolve (Hendriks 1994).

Table 6.3 Areas of sedimentary basins in relation to total land areas

   

Sedimentary basina

Region

Total land area (106 km2)

Area (106 km2)

% of total land area

Asia 27.60 11.34 41
Former USSR 22.40 9.71 43
Europe 4.93 2.76 56
South America 17.80 9.07 51
North America 24.20 9.21 38
Africa 30.30 15.23 50
Australia and Oceania 8.51 7.47 88
World 135.74 64.79 48

a. Source: Koide et al. (1992).

Hendriks (1994) has estimated the CO2 storage capacity of such aquifers under the assumption that the injected CO2 displaces water. He has made alternative estimates that depend on the extent to which structural traps are needed for secure storage (see table 6.4). Without a structural trap, the injected CO2 might eventually migrate from the injection site to other subterranean sites where storage is not desirable or even to where it can escape to the atmosphere. If structural traps are not needed, the estimated worldwide sequestering capacity of aquifers is about 15,000 GtC; if structural traps are necessary, the sequestering capacity is 60 GtC (Hendriks 1994). A recent multinational, multi-institutional study (Holloway 1996) carried out for the European Commission concluded that structural traps might not be necessary to achieve a reasonably high degree of confidence in the security of aquifer storage, if the CO2 is injected far enough from the aquifer boundary that it is predicted not to reach the boundary.

Table 6.4 Order-of-magnitude estimates of the CO2 storage capacity of aquifers (GtC)

 

Sequestering capacity

Region

Structural trap required

No structural trap required

Western Europe 3 700
Eastern Europe 8 2,000
North America 11 2,700
Latin America 11 2,700
Africa I 1 2,700
Middle East 3 700
Oceania and Asia 14 3,400
World 60 15,000

Source: Hendriks (1994).

It estimated that the underground CO2 storage potential for the European Union plus Norway is more than 200 GtC - equivalent to 250 years of total CO2 emissions for all of OECD Europe; most of this storage capacity is in aquifers under the North Sea. The study pointed out, however, that, to ensure adequate storage security in large aquifers, large amounts of data would be needed regarding reservoir integrity.


Contents - Previous - Next