Risk assessment of a CO2 pipeline network for CCS – A UK case

Carbon Capture and Storage technology (CCS) requires transporting large amounts of CO2 over large distances, as capture plants are expected to be situated near power plants and other large industrial sources such as steel and cement works, while storage locations are expected to be in remote geological formations, typically offshore. CO2 can be transported using one or a combination of transport media: truck, train, ship or pipeline. Transport by pipeline is considered the preferred option for large quantities of CO2 over long distances, and is the subject of this paper. The CO2 pipeline network most appropriate for a country is clearly a function of the specific location of sources and storage points, their capacity and a number of other factors such as population centers and geographical features (rivers, mountains, railroads, motorways, etc). The phased roll-out and initial design of the onshore part of a CO2 pipeline network for the UK, suitable to deal with the distribution of forecasted CO2 amounts captured at major sources, was proposed by Lone et al, 2010, based on a techno-economics analysis. The analysis resulted in proposed sizing and location of various pipeline segments in a three-phase rollout, dealing with largest duties first, and details of CO2 flows and pressures for each segment in each roll-out phase. This paper describes the quantitative risk analysis of this pipeline network, and in particular an assessment of consequences due to the possible CO2 releases. First, the probability of various accidental events is determined. Then, the estimation of consequences is made with the PHAST software using its “long pipeline” release model, for two types of release: i) from a whole with diameter equal to 20% of section area and ii) from full bore rupture (catastrophic release). Accidental events in a CO2 pipeline can produce a spray release followed by a dense gas dispersion, and the high concentration of CO2 can cause fatalities. To determine possible health effects it is not only the CO2 concentration but also the duration of the exposure, as the gas cloud evolves are quantified. For the calculation of risk, the consequences are associated to the Probit function, which calculates the fraction of individual deaths in a population. The network examined passes near residential areas. For this situation, the consequences produced by a possible release are calculated and various corridors of risk are identified in terms of population at risk, utilising population distribution data. Finally, the trade-offs achievable between population risk decrease and additional pipeline costs arising from alternative pipeline pathways are demonstrated by means of a specific example.