Gunn M. G. Teige

The lack of relationship between overpressure and porosity in North Sea and Haltenbanken shales

Abstract Disequilibrium compaction (undercompaction) is probably the most accepted mechanism for explaining overpressured formations. This mechanism relies on the hypothesis that overpressured shales have higher porosity than normally pressured shales for a given depth. Converted to porosity units, log response data from two Jurassic shales at Haltenbanken and nine Cretaceous and Tertiary shales in the Norwegian sector of the North Sea have been investigated to see whether such a relationship exists in these areas. The depth of the studied shale formations varies between 0.5-5 km. The degree of overpressuring within one formation typically varies by 20 MPa. The log response data show that neither the porosity of massive North Sea shales (derived from sonic and density logs) nor the porosity of Haltenbanken intra-reservoir shales (derived from density and neutron logs) differ significantly between the overpressured and normally pressured stratigraphic units. Distinct differences between the pressure regimes are, however, apparent in the resistivity and sonic log data from deeply buried shales at Haltenbanken. These differences do not reflect porosity variations and are interpreted to reflect textural changes resulting from the overpressuring. These observations demonstrate that the Haltenbanken and North Sea shales do not conform to an undercompaction mechanism for overpressuring, as the overpressured wells fail to show elevated porosity. On the contrary, each formation seems to have been individually compacted according to burial depth, independent of pressure regimes.

Origin of overpressures in shales: Constraints from basin modeling

Techniques for detection, evaluation, and prediction of pore pressures in low-permeability rocks and equations for fluid-pressure computations in most integrated basin-modeling software are based on relationships between porosity and effective stress in shales. However, recent data show that overpressured shales in the North Sea do not exhibit higher porosities than the normally pressured shales of the same formation at similar depths.To further evaluate the existence of porosity vs. effective stress relationships in shales, fluid-flow simulations and porosity modeling in a typical high-pressure and high-temperature well in the North Sea were undertaken. The parameters in the permeability and porosity equations were adjusted until a satisfactory fit was achieved between the observed and modeled porosity and fluid pressure at present. However, the modeled porosity and pore pressure vs. depth history of the sediments deviated significantly from known porosity and pore pressure vs. depth relationships that have been observed in North Sea shales and elsewhere today.Because the results from basin modeling based on porosity-stress relationships were unacceptable, irrespective of parameter choices, and the well data from the North Sea show no signs of elevated porosities in the overpressured shales, it is inferred that effective stress-driven compaction alone has not generated the hard overpressures observed in deeply buried North Sea shales. These conclusions are suggested to be generally applicable to shales with low porosities and hard overpressures worldwide, both because of the physics involved and because similar results can be extracted from published modeling in the Niger Delta.

Capillary resistance and trapping of hydrocarbons: a laboratory experiment

Low permeability cap rocks retain oil by capillary forces when the pore throats of the seals are sufficiently small to prevent a flux of oil into the cap rock. In order to investigate the influence of aquifer overpressures on oil retention, water pressure was applied to a water-wet, highly permeable (1988 mD) core sample, which was oil-saturated to irreducible water saturation Swi and mounted with a low-permeability and water-wet membrane at the outlet. A water pressure difference of 0.5 MPa was applied across the core. This pressure was high enough to ensure fluid flow through the sample. The experiment was designed to see whether the water pressure would force oil through the membrane or if capillary forces at the sandstone–membrane interface would retain the oil, in which case water flow might take place in the (residual) water in the core and through the membrane. The experiment showed that oil was kept in place by capillary forces while water flowed through the core and the membrane. Accordingly, residual water can move through sandstones that are saturated to Swi. The experiments also demonstrated that the permeability associated with this residual water is high enough to prevent overpressures in the aquifer below the oil–water contact from pushing oil through a membrane seal. Thus, even for this highly permeable sandstone, the overpressure in the aquifer will not cause capillary seal failure.

Differences between flow of injected CO2 and hydrocarbon migration

Abstract Knowledge of fluid flow processes in the subsurface is important for CO2 storage operations as well as for hydrocarbon exploration. Repeated seismic surveys for more than 10 years of CO2 injection into the Utsira Formation, in the Sleipner area, offer a unique dataset. This dataset holds information on fluid migration processes that can be analysed for the benefit of hydrocarbon exploration and CO2 storage considerations alike. Thorough analyses of these datasets reveal several features that give useful information of subsurface fluid flow processes. The CO2 in the Utsira Formation has flowed laterally beneath thin, intra-formational shales. At the same time, CO2 has flowed vertically through shaly horizons that would normally be considered as barriers to fluid flow. This flow has apparently taken place through vertically stacked flow conduits through the shales. These conduits may to some extent have existed prior to the start of CO2 injection, but may also have been augmented by the CO2 injection process. The calculated pushdown of seismic reflectors below the CO2 plume is less than that observed, which may point to the presence of hitherto unrecognized flow paths for the CO2. Hydrocarbon migration pathways are in general not recognizable in seismic data. This implies that such avenues are significantly thinner than those of the CO2 migration in the Utsira Formation. This result points to the presence of mixed-wet migration pathways, in which capillary flow resistance may not control the (sub-horizontal) flow path thickness. A circular depression at the top of the Utsira Formation that existed prior to the injection may be interpreted as a result of a deeper seated sand remobilization feature. Such features will also promote vertical hydrocarbon migration where they are present. A more widespread occurrence of such features may explain why hydrocarbons are generally found beneath thick shales, but are less likely to be found below thin intra-formational shales below the structural spillpoint of the top seal. These observations suggest that seal thickness is an important parameter, even if the capillary entry pressure of the sealing rock is sufficiently high to preserve significant hydrocarbon columns.

Seismic characteristics of fluid leakage from an underfilled and overpressured Jurassic fault trap in the Norwegian North Sea

Pre-drill estimates of hydrocarbon column heights are often uncertain, particularly where filling to structural spill point is questioned. Identification of locations where vertical leakage is concentrated can allow more reliable hydrocarbon column height predictions. The locations of vertical leakage were sought on seismic data over an underfilled and overpressured trap (35/10-2) in the Norwegian North Sea. It was hoped that one single location or one single narrow leakage zone coinciding with the gas–water contact would be found. If such a location or zone existed, it was expected to be in or above a fault plane, as leakage in the 35/10-area is thought to mainly result from shear failure along faults. The investigation found a zone with pronounced dimming above a triple fault intersection bounding the 35/10-2 structure. This zone stretches further downdip and is positioned above a fault plane. The shallowest part of this zone intersects the top of the reservoir at the depth of the proven gas–water contact. These observations suggest that a discrete leakage zone, of which the shallowest part controls the column height of the structure, has been identified and that the column height is limited by leakage resulting from shear failure. The occurrence of similar seismic features over undrilled structures can lead to safer assessments of hydrocarbon column heights, especially if the observations are consistent with the general knowledge of stress state and leakage processes in the area.

The Influence of Stress Regimes on Hydrocarbon Leakage

Hydrocarbon leakage through faults and fractures commonly limits in-place hydrocarbon reserves. Faulting and fracturing are controlled by effective stress changes, and such changes may therefore alter hydrocarbon column heights. The predictive power of stress history analyses in seal evaluation depends on how accurately the stress history and relationships between effective stress changes and hydrocarbon leakage can be determined. Stress history and hydrocarbon occurrence were examined in four different overpressured provinces of offshore Norway in the search for such relationships. These provinces have experienced different geological histories and variable amounts of hydrocarbon leakage. Because all these areas received fairly recent hydrocarbon charge, the work focused on the identification of recent geological events that may subsequently have influenced recent stress history, including the present-day stresses. Areas of recent structuring were found to be characterized by more extensive hydrocarbon leakage than areas with less such structuring. This increased frequency of hydrocarbon leakage was interpreted to be the result of shear failure at the trap crests, induced by the combined effects of elevated pore pressures, stress anisotropy, and recent stress changes. These results suggest that identification of recent stress changes based on the geological history of the study area could aid the prediction of hydrocarbon occurrence. It is inferred that stress history analyses can also reduce the uncertainty involved in seal analyses elsewhere.

Seal Failure Related to Basin-scale Processes

The leakage of trapped petroleum is a major concern in hydrocarbon exploration and has led to a large number of exploration failures. Changes in stress state, fluid pressure, and cap rock permeability may all result in a loss of trapped hydrocarbons. Such changes may result from several different subsurface processes. This chapter describes an examination of important processes that control compaction, fluid flow in reservoirs, fault reactivation, and their influence on leakage from hydrocarbon reservoirs. It was concluded that seal-failure analysis is seldom based on more than a few of the operating processes and therefore does not reach its full potential. Especially, the effects of overpressures on sediment compaction and hydrocarbon leakage seem to have been oversimplified and commonly overstated. The conditions for hydrofracturing and the corresponding loss of hydrocarbons from structural crests are also commonly considered too superficially in sealing analyses. It was concluded that inadequate leakage assessments can result from neglect of some subsurface processes that influence stress, pore pressure, and hydrocarbon permeability in the seal. Seal integrity predictions can be improved if thorough analyses of the relevant subsurface processes routinely precede the sealing analyses.

Evaluation of caprock integrity in the western (high-pressured) haltenbanken area — a case history based on analyses of seismic signatures in overburden rocks

Abstract Seal failure is a significant risk factor in the western, high-pressured part of the Haltenbanken area. Accordingly, an investigation of seismic expressions of hydrocarbon leakage was initiated to aid further exploration in this area. The main objective of this caprock integrity study was to evaluate the caprock of the high-pressured Kristin structure in western Haltenbanken, and more generally to identify seismic expressions of hydrocarbon leakage. This was initially done on a regional 2D dataset, and later on a semi-regional 3D dataset. The results show a relationship between vertical caprock leakage from Jurassic reservoirs and strong seismic dim-zones on the 2D data. The correlation between dim-zones on the 2D data and dim-zones on the 3D data was very good. However, the strong leakage-related dim-zones on the 2D data were weaker on the 3D data, and consequently an equivalent relationship between caprock leakage and dim-zones on the 3D data could not be established. The implications of the fact that the relationship between reservoir leakage and strong dim-zones — which was clear on the 2D data — was unclear on the 3D data, is that no reliable tool for identifying hydrocarbon leakage from seismic 3D data exist in the Haltenbanken area. However, analyses from the 2D data have demonstrated that such data can be used as a positive contribution in exploration leakage assessments in this area.

How Accurate Can We Compute Hydrocarbon Column Heights in Overpressured Prospects

Computation of maximum column height is common during evaluation of overpressured traps. The main idea behind such computations is to assume tensile fracturing, and that the hydrocarbon buoyancy cannot exceed the difference between the fracturing pressure and the reservoir pore pressure. The uncertainties of such computations are rarely addressed and have to date not been quantified. We perform such quantification by using pre-production reservoir pore pressure variations to estimate the accuracy that can realistically be attained in evaluating pore pressures in undrilled prospects. We have also analysed uncertainties in fracture pressure determination from leak off tests. This was performed by comparing leak off pressures and fracture closing pressure for a large number of extended leak off tests from the Norwegian Continental Shelf. The uncertainties of the fracturing pressure is frequently 5 MPa or more, which is comparable to the buoyancy of a 1 km gas column. We therefore conclude that computation of maximum column heights should not be computed from individual wells, and that computations based on regional fracture pressure trends need further verification.

History-matching of CO2 Flow at Sleipner – New Insight based on Analyses of Temperature and Seismic Data

History matching of layer 9 of the Utsira Formation in the Sleipner area has so far been problematic, but could be achieved it higher CO2 temperature than has previously been estimated are invoked. Such high temperatures can to some extent be justified from hithero unknown DST measurements. Furhtermore, heat exchange between warm injected CO2 and the colder reservoir can be significant. Such heat exchange may possibly explain the discrepancy between modelled and observed CO2 fow in the Utsira Formation.