Saline aquifers and depleted hydrocarbon fields situated beneath the North Sea are currently being proposed as storage repositories for anthropogenic CO2 captured from point source emitters in the UK and mainland Europe. Two experimental sites are already operating successfully offshore Norway: Sleipner since 1996 and Snøhvit since 2007, collectively storing several million tonnes of CO2/year in the sub-surface. Despite the apparent success of these current projects, one of the major public and scientific concerns is the ability of storage sites to retain CO2 on the millennial timescales required for CO2 plume stabilisation and dissolution. Some areas of the North Sea are also known to contain palaeo-gas seepage pathways within overburden sediments that overlie deeper hydrocarbon reservoirs (e.g. Witch Ground Graben). These areas either need to be avoided for CO2 storage or rigorously assessed in terms of leakage risk. Since the Sleipner storage site lies within such a province, this thesis delivers a detailed evaluation of the Nordland Group overburden and a critical assessment of its long-term sealing capability for CO2. From interpretation and detailed mapping of a baseline 3D seismic dataset (acquired before CO2 injection operations commenced in 1996), we have identified numerous palaeo-migration pathways and high-amplitude seismic anomalies within the Nordland Group overburden sediments deposited above the Sleipner CO2 storage site. We attributed these features to thermogenic or biogenic gas migration, accumulation and bio-degradation over geological time. We also mapped a complex network of sand-filled, glacial channels and tunnel valleys distributed within a few hundred metres below seabed and highlighted their significance as potential fluid migration networks and/or secondary storage containment for leaking CO2. Of further significance, we confirmed that these overburden features also create spatial density variations that impact on the accuracy of seismic time-depth conversions, resulting in the probability of topographic distortions being propagated into seismic interpretations and models. To the best of our knowledge no such detailed mapping of the Nordland Group overburden at Sleipner has been undertaken previously. To determine whether the top layer of the CO2 plume at Sleipner might encounter these relict pathways as it ascends and migrates laterally beneath the caprock, we evaluated the critical column heights required for a CO2 accumulation to enter such a pathway under a range of storage conditions for a CH4/CO2/brine system; assuming that these pathways currently contain methane gas. Risking scenarios were based on a range of phase saturation, pressure, temperature, density, viscosity, interfacial tension and wettability conditions likely to be encountered at depths commensurate with the caprock at Sleipner. We concluded that given certain conditions at the caprock, CO2 could leak more easily into palaeo-migration pathways than CH4 (i.e. at lower entry pressures and therefore smaller column heights), assuming that brine densities and, most importantly, pore radii have not changed significantly over geological time (i.e. no cementation or dissolution has taken place). To further understand the dynamic significance of these palaeo-migration pathways, channels and tunnel valleys, including their ability to form inter-connected leakage/migration networks, we constructed a series of high-resolution 3D models of the Sleipner storage site and overburden, then used stochastic basin modelling and simulation techniques to investigate the processes involved during the introduction of CO2 into the storage site over a prolonged time period. Models were populated with geological, stratigraphic and structural information derived from our seismic interpretation. Flow simulations were calibrated to published data and matched to the present-day plume distribution. The absence of observational reservoir pressure and temperature data from Sleipner introduces significant uncertainty to model outcomes with respect to CO2 density and column height estimates and to surmount this difficulty we constrained the caprock temperature to CO2 density estimates obtained from the most recent gravity data observations at Sleipner. We concluded that the overburden heterogeneity is significant and palaeo-migration pathways, highpermeability channels and tunnel valleys at Sleipner may become potential migration pathways for CO2 as the plume continues to spread laterally over the coming decade, but the possible storage response is difficult to quantify given the absence of sufficient overburden rock property information and accurate pressure and temperature data for the storage site. The overall conclusion from this work is that insufficient information was collected within the Sleipner area prior to storage site development and too many significant studies which should have been performed as a pre-requisite (e.g. obtaining a caprock sample for laboratory testing of potential seal capacity), were actually performed some years after CO2 injection operations had already commenced. The pressure and temperature conditions at the caprock depth for the Sleipner storage site are also marginal in terms of maintaining CO2 above critical point conditions in dense phase and thus maximising storage efficiency. Most significantly, no rigorous overburden mapping and risking was performed for Sleipner (such as the work described in this thesis), thus the fact that no leakage has been detected at Sleipner is more due to good fortune than following best practices. Hopefully, our work has highlighted these key deficiencies so that future CO2 storage site feasibility and development studies will be performed more diligently.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:578424 |
| Date | January 2012 |
| Creators | Nicoll, Grant Douglas |
| Contributors | Haszeldine, Stuart |
| Publisher | University of Edinburgh |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/1842/7813 |
Page generated in 0.0133 seconds