Search

The Sleipner CO 2 storage operation has been injecting CO 2 since 1996, and the growth of the plume has been intensively monitored using time-lapse seismic techniques. Detailed history-matching of the topmost CO 2 layer has proven challenging. This paper summarizes results from a series of flow simulations examining two key parameters affecting CO 2 mobility: permeability heterogeneity and fluid temperatures within the plume. The best match to the observed distribution of CO 2 was achieved by including high permeability channels in the reservoir flow model, as observed on seismic data. Thermal models suggests that CO 2 enters the top sand layer 7 °C warmer than the ambient reservoir. The resulting reduction in the density and viscosity of CO 2 does not significantly improve the fit between seismic and simulation.

Time-lapse 3D seismic reflection data, covering the CO2 storage operation at the Snohvit gas field in the Barents Sea, show clear amplitude and time-delay differences following injection. The nature and extent of these changes suggest that increased pore fluid pressure contributes to the observed seismic response, in addition to a saturation effect. Spectral decomposition using the smoothed pseudo-Wigner–Ville distribution has been used to derive discrete-frequency reflection amplitudes from around the base of the CO2 storage reservoir. These are utilized to determine the lateral variation in peak tuning frequency across the seismic anomaly as this provides a direct proxy for the thickness of the causative feature. Under the assumption that the lateral and vertical extents of the respective saturation and pressure changes following CO2 injection will be significantly different, discrete spectral amplitudes are used to distinguish between the two effects. A clear spatial separation is observed in the distribution of low- and high-frequency tuning. This is used to discriminate between direct fluid substitution of CO2, as a thin layer, and pressure changes that are distributed across a greater thickness of the storage reservoir. The results reveal a striking correlation with findings derived from pressure and saturation discrimination algorithms based on amplitude versus offset analysis.

Demonstrating secure containment is a key plank of CO2 storage monitoring. Here we use the time-lapse 3D seismic surveys at the Sleipner CO2 storage site to assess their ability to provide robust and uniform three-dimensional spatial surveillance of the Storage Complex and provide a quantitative leakage detection tool. We develop a spatial-spectral methodology to determine the actual detection limits of the datasets which takes into account both the reflectivity of a thin CO2 layer and also its lateral extent. Using a tuning relationship to convert reflectivity to layer thickness, preliminary analysis indicates that, at the top of the Utsira reservoir, CO2 accumulations with pore volumes greater than about 3000 m3 should be robustly detectable for layer thicknesses greater than one metre, which will generally be the case. Making the conservative assumption of full CO2 saturation, this pore volume corresponds to a CO2 mass detection threshold of around 2100 tonnes. Within the overburden, at shallower depths, CO2 becomes progressively more reflective, less dense, and correspondingly more detectable, as it passes from the dense phase into a gaseous state. Our preliminary analysis indicates that the detection threshold falls to around 950 tonnes of CO2 at 590 m depth, and to around 315 tonnes at 490 m depth, where repeatability noise levels are particularly low. Detection capability can be equated to the maximum allowable leakage rate consistent with a storage site meeting its greenhouse gas emissions mitigation objective. A number of studies have suggested that leakage rates around 0.01% per year or less would ensure effective mitigation performance. So for a hypothetical large-scale storage project, the detection capability of the Sleipner seismics would far exceed that required to demonstrate the effective mitigation leakage limit. More generally it is likely that well-designed 3D seismic monitoring systems will have robust 3D detection capability significantly superior to what is required to prove greenhouse gas mitigation efficacy.

Time lapse seismic surveys covering the CO2 injection plume at Sleipner are used to test novel techniques which estimate the thickness of a spreading CO2 layer. Utilising the spectral content of the data, the methods extend the limited vertical resolution encountered with time domain data.Spectral decomposition using the smoothed pseudo Wigner-Ville distribution extracts monochromatic reflection amplitudes from the topmost CO2 layer and is used to assess the lateral variation in peak tuning frequency. This provides a direct proxy for temporal thickness which show consistency with true layer thicknesses derived from structural analysis.A spectral inversion method is also applied to a subset of the upper layer, with limited success. It is noted that high signal-to-noise ratios and the absence of overlying and underlying reflections are required to utilize the technique fully.

To transfer responsibility for a CO2 storage site to the competent authority, the EU Storage Directive requires it to be demonstrated that CO2 will be completely and permanently contained and that the storage site is evolving towards a situation of long-term stability. As outlined in current EU guidance, there may be room for interpretation by both the national competent authority and the storage operator in the implementation of this part of the Directive. Here we postulate that, for storage in dipping open aquifers, the requirements of the Directive can be met by modelling various combinations of a boundary constraint, a migration velocity constraint and a constraint on the minimum fraction of injected CO2 that is likely to be trapped at a defined time (1000 or 10,000 years) after injection ceases. A key outcome is that there is a strong trade-off between storage efficiency and storage security when storage is maximised under different combinations of constraints. To increase confidence in security under all scenarios, modelling longer post-injection periods is desirable; however there may be computational limits to this.

One of the three key regulatory requirements in Europe for transfer of storage site liability is to demonstrate conformity between predictive models of reservoir performance and monitoring observations. This is a challenging requirement because a perfect and unique match between observed and modeled behavior is near impossible to achieve. This study takes the time-lapse seismic monitoring data from the Sleipner storage operation to demonstrate that as more seismic data becomes available with time, predictive models can be matched more accurately to observations and become more reliable predictors of future performance. Six simple performance measures were defined: plume footprint area, maximum lateral migration distance of CO2 from the injection point, area of CO2 accumulation trapped at top reservoir, volume of CO2 accumulation trapped at top reservoir, area of all CO2 layers summed, and spreading co-efficient. Model scenarios were developed to predict plume migration up to 2008. Scenarios were developed for 1996 (baseline), 2001, and 2006 conditions, with models constrained by the information available at those times, and compared with monitoring datasets obtained up to 2008. The 1996 predictive range did generally encompass the future observed plume behavior, but with such a wide range of uncertainty as to render it of only marginal practical use. The 2001 predictions (which used the 1999 and 2001 seismic monitoring datasets) had a much lower uncertainty range, with the 2006 uncertainties somewhat lower again. There are still deficiencies in the actual quality of match but a robust convergence, with time, of predicted and observed models is clearly demonstrated. We propose modeling-monitoring convergence as a generic approach to demonstrating conformance.

The Nordland Group is an important stratigraphical unit within the upper Cenozoic of the northern North Sea. At its base lies the Utsira Sand, a dominantly sandy regional saline aquifer that is currently being utilized for carbon dioxide sequestration from the Sleipner gas and condensate field. A ‘mudstone drape’ immediately overlies the Utsira Sand, forming the caprock for this aquifer. The upper part of the Utsira Sand was recently dated as Early Pliocene, but the precise age of the overlying Nordland Group mudstones has remained uncertain. Dinoflagellate cyst, pollen and spore, foraminiferal and stable isotopic analyses have been performed on these mudstones from a conventional core within the interval 913.10–906.00 m (drilled depth) in Norwegian sector well 15/9-A-11. The samples lie closely above the Utsira Sand. Results give a Gelasian (late Late Pliocene) age for this interval, with a planktonic foraminiferal assemblage at 913.10 m indicating warm climatic conditions and an age between 2.4 and 1.8 Ma. An abundance of the cool-tolerant dinoflagellate cysts Filisphaera filifera and Habibacysta tectata at 906.00 m, along with evidence from pollen and foraminifera, points to deposition during a cool phase of the Gelasian. Dating the mudstone drape provides useful insights into depositional processes. It seems likely that the Utsira Sand, a basinal lowstand deposit, became progressively starved of clastic input as sea level rose and shorelines retreated. The mudstone drape is interpreted as a highstand deposit, perhaps including a maximum flooding surface. Overlying prograding wedges of the Nordland Group form a regressive succession, characterized by increased sedimentary input and rates of deposition of at least 25 cm per 1000 years, which is more than five times that of the Utsira Sand. This is the first published study of a dinoflagellate cyst assemblage from the Upper Pliocene of the northern North Sea. The new dinoflagellate cyst species Echinidinium nordlandensis Head sp. nov. and Echinidinium sleipnerensis Head & Riding sp. nov. are formally described.

Geological storage of CO2 that has been captured at large, point source emitters represents a key potential method for reduction of anthropogenic greenhouse gas emissions. However, this technology will only be viable if it can be guaranteed that injected CO2 will remain trapped in the subsurface for thousands of years or more. A significant issue for storage security is the geomechanical response of the reservoir. Concerns have been raised that geomechanical deformation induced by CO2 injection will create or reactivate fracture networks in the sealing caprocks, providing a pathway for CO2 leakage. In this paper, we examine three large-scale sites where CO2 is injected at rates of ~1 megatonne/y or more: Sleipner, Weyburn, and In Salah. We compare and contrast the observed geomechanical behavior of each site, with particular focus on the risks to storage security posed by geomechanical deformation. At Sleipner, the large, high-permeability storage aquifer has experienced little pore pressure increase over 15 y of injection, implying little possibility of geomechanical deformation. At Weyburn, 45 y of oil production has depleted pore pressures before increases associated with CO2 injection. The long history of the field has led to complicated, sometimes nonintuitive geomechanical deformation. At In Salah, injection into the water leg of a gas reservoir has increased pore pressures, leading to uplift and substantial microseismic activity. The differences in the geomechanical responses of these sites emphasize the need for systematic geomechanical appraisal before injection in any potential storage site.