Carbon capture and storage (CCS) is increasingly seen as one of the mechanisms that can make a useful contribution to the reduction of carbon dioxide emissions over the next 50 years.

In order to have a significant impact, CCS will have to be installed on a large scale (Gt carbon year-1). Some of the storage will be in terrestrial sites such as saline aquifers, from which there is a risk of carbon dioxide leakage through faults in the caprock, borehole seals and other sources. There is also some risk of leakage from the pipeline infrastructure used to transport CO₂.

This is important for several reasons: first, accurate data on the quantity of carbon that has been sequestered will be needed for trading and accounting purposes; second, any leakage will negate the original purpose of the storage; third, the leaking CO₂ might damage surface ecosystems. Hence, more information is needed on the responses of surface ecosystems to CO₂ leaks.

The results will be used both to inform planning applications and environmental impact assessments for proposed terrestrial CCS schemes, and to detect the occurrence of leaks from existing schemes. The University of Nottingham has an ongoing programme of research on these topics.

The University has established a soil gassing facility which will simulate the elevated soil CO₂ concentrations caused by a CCS leak. The facility, seen below at the beginning of the 2006 season, is known as ASGARD (Artificial Soil Gassing and Response Detection) and is a development of the successful previous PRESENSE investigation of the impact of pipeline natural gas leaks on overlying vegetation.

Carbon dioxide from cryogenic cylinders is supplied to (initially) 12 experimental plots, each 2.5 x 2.5 m. The gas flow is individually regulated by mass flow controllers, before being injected into each plot from a diffusive outlet 0.6 m below the ground surface. Concentrations of CO₂ and O₂ in the soil atmosphere are measured daily. Another 12 ungassed plots act as experimental controls.

The experimental programme for 2006/7 involved the establishment of the site itself (hardware and services, funded by a HEFCE/SRIF3 award), together with one PDRA (funded by a Research Councils TSEC/UKCCSC grant). Since June 2006 the facility has been fully operational and has been used to carry out initial experiments on pasture grass, spring barley and linseed. The measurement programme includes routine gas analysis, growth samples and regular monitoring of spectral reflectance. The yellow objects in the picture above are the gas injection pipes; the vertical black tubes are access tubes for soil gas measurement. We are also working closely with the British Geological Survey (BGS) at Keyworth, who are a major player in the UK programme on CCS and are characterising the ASGARD site.

Objectives

The ASGARD facility gives us a unique capability to research the consequences of CCS leaks for terrestrial ecosystems. In 2006, our study involved the following topics:

• Measurement of spectral reflectance signatures which can be used for the satellite remote detection of leaks;
• Measurement of plant stress responses such as leaf area and chlorophyll content;
• Measurement of carbon isotope ratios as a sensitive discriminator of fossil-fuel from natural CO₂;
• Evaluation of a range of gas measurement techniques.

Further potential topics include:

• Measurement of changes in soil chemistry and its interactions with soil water;
• Measurement of responses of soil organisms;
• Mass-balance work to measure the relationship between injected and emitted CO₂. This would be achieved by measuring surface fluxes with a flux box. An additional facility would involve supplementing the ground injection plots with surface soil tanks to contain lateral gas migration;
• Comparison of leak effects on different soil types and structures, using tanks or excavated soil blocks;
• Comparisons between different methods of measuring the soil gas concentration;
• Measurement of the balance between shoot stimulation and root repression for overlying plants. The high soil gas concentrations are expected to damage plants by asphyxiation, but the enhanced flux of CO₂ from the surface will then elevate the atmospheric concentration which will stimulate plant growth;
• Comparison of responses to CO₂ in different species;
• Monitoring ecological changes in mixed canopies.

Results

The distribution of CO₂ across the plots was measured by barholing. The measurements on 27 July 2006 for the eight grass plots are shown above. Each gassed plot shows a similar peak and an approximately symmetrical gas distribution across the plot. This pattern gives confidence that the Asgard system can produce reliable replicate treatments for analysis. Concentrations in the control plots are low, but there is some evidence of contamination from the neighbouring gas plots. Because we were concerned about this issue, we also established "remote control" plots for grass that were a minimum 10 metres distance from the gassed zone.

Hyperspectral remote sensing techniques were tested for their ability to distinguish subtle signs of plant stress. The stress index shown above is the ratio of two subsidiary peaks of the first derivative spectrum and is sensitive to changes in plant pigments: a higher ratio indicates a healthier canopy. All plots showed a decline through the early summer, which was notably dry, with recovery from the end of July. The gassed plots were more stressed than the controls from early July onwards. Significant differences between the local controls and the remote controls appeared towards the end of the season.

Gas samples from the plots were monitored for analysis of stable carbon isotopes. Biogenic carbon dioxide has a δ¹³C value (relative to standard) of about -23 ‰. The injected gas is isotopically lighter (-31 and -34 in the examples above) and can be readily distinguished from the biogenic carbon dioxide. The curves in the figure show the values predicted by gas mixing; the measurement points above the curve indicate isotope fractionation.