Menu
Key Knowledge Document | NS051-SS-REP-000-00013
This technical report focuses on the Northern Endurance Partnership’s approach to building a a subsurface geological model for CO2 sequestration in the Endurance structure in the southern North Sea.
Building a geological model is critical to understand the potential storage volumes available for CO2, to assess risk and uncertainty in parameters defining the storage container and to identify data gaps. Essentially a model is built to represent, as accurately as possible, the shape and physical properties of a subsurface volume, in this case the Endurance Anticline. Geological models are generally built from information obtained from seismic data and well data. Since these data are often sparse, geological interpretation between the data points must be made to model the whole structure. The complexity of a geological model is related to the available data. If lots of good subsurface data are available, then a higher level of detail can be expected, along with a decrease in the uncertainty of the model.
Considering the location of the Endurance anticline, 145 km offshore and approximately 1000m below the sea floor there is a relatively good set of data available to build a geological model. These data are derived from several seismic studies, wells, and published material. More data could always be used but the combination of three wells drilled into the structure, seismic data and extensive published regional work is sufficient to characterise the structure for initial modelling work. The method used to construct the model followed industry standards and was very thorough, the analysis suggests the structure is ideal for injecting and containing large volumes of CO2.
This report is related to other key documents reviewed here. The Primary Store Geophysical Model and Report forms a critical part of the geological model as the geophysical data is used to define the upper and lower surfaces of the storage reservoir. Once a geological model is constructed it can then be integrated into a dynamic model that can be used to model fluid flow, pressure, and uncertainty within the structure (Primary Store Dynamic Model and Report). The dynamic model is critical for field development modeling and to help develop a CO2 monitoring plan. As more data becomes available through future drilling, seismic acquisition, and CO2 injection both static and dynamic models should be continually updated and refined.
In this document BP summarizes the work and methodology for construction the geological model of the Bunter sandstone reservoir in the Endurance anticline. This model is constructed using various vintages of 2D and 3D seismic, information from three wells drilled into the structure and information from offsetting wells and published work. The report outlines industry standard modelling workflow resulting in a very thorough and geologically reasonable model.
The depositional environment of the Bunter sandstone, core analysis, petrophysical analysis, facies analysis and upscaling are all critical in developing a geological model and are all thoroughly reviewed in the document. The result of the analysis demonstrates that the Bunter sandstone is a thick sandstone (average 275m) pervasive across the whole structure. The sandstone has a high net to gross (average 94%) indicating most of the interval is available to store CO2. The net sandstone has high porosity (average 22.5%) and high permeability (average 300 mD). The Bunter sandstone in this location would be considered a high-quality reservoir when compared globally to other potential reservoirs. There is variable cementation in the sandstone causing a decrease in porosity, but this has been captured in the model. The sandstone is relatively homogenous with some evidence of internal baffles. The extent of these baffles is not known and has been considered in the uncertainty of the model.
Critical to CO2 containment is a seal on top of the Bunter reservoir to prevent upward migration of CO2. The primary seal unit above the Bunter sandstone is approximately 10m of Rot Clay and on top of that approximately 100m of Rot Halite. Like the Bunter Sandstone the Rot Formation is laterally pervasive and appears remarkably consistent in thickness and lithology. Salts such as the Rot Halite are exceptionally good seals as the porosity and permeability of the salt is extremely low and its rheological characteristics make open fractures (conduits for CO2 leakage) within the salt very unlikely.
There appears to no identifiable faulting cutting through the reservoir or faulting across the top seal which is critical for CO2 containment. Where large faults are identified it is in the geological section above the seal. There may be some small-scale faulting at the very top of the reservoir which have been incorporated into the geological model. No large-scale faults affecting the entire reservoir or seal have been identified and therefore none have been incorporated into the static model.
Mapping the top and bottom surfaces of the Bunter sandstone and assessing any structural features such as faulting is reviewed in the Primary Store Geophysical Model and Report.
The relatively thick sandstone with high porosity and permeability makes the Bunter sandstone an ideal target for CO2 disposal. The anticline is a large structural feature, and the aerial extent and height of the structure indicates it can store a considerable quantity of CO2. The halite seal above the target reservoir is thick and laterally pervasive and no large-scale faulting has been identified in the reservoir formation or seal which is critical for containment.
An unusual feature of the Endurance anticline is that the Bunter sandstone within the anticline may be in hydraulic continuity with an outcrop of the formation on the seabed approximately 20 km to the east. If this is the case the result is a hydraulically open system which may be beneficial to pressure management. The saline water displaced by injecting CO2 into the anticline may be able to dissipate, thereby decreasing the amount of pressure build-up. The CO2 being lighter than water would remain trapped in the anticlinal structural. Pressure management is discussed further in the Primary Store Dynamic Model and Report.
Discussion of potential volumes of sequestration, and uncertainty analysis are not discussed in this document but are reviewed and summarized in Primary Store Dynamic Model and Report.
The static geological model, as presented, appears to be a robust framework to integrate into a dynamic model. However, it should be noted the actual geological model itself was not available for review just the methodology and descriptions in the Key Knowledge Document.
If more detailed study of the overburden is to be done, core analysis could be beneficial to determine more accurate porosity-permeability relationships, as well as more accurate density and neutron-based porosity measurements, as opposed to porosity derived from available data (i.e. resistivity) and estimated petrophysical (a, m, n) values. Regarding absence of mud, and provenance of carbonate in the system, additional work could be performed on core and additional wells to delineate areas of cementation and determine carbonate source. Given the high Net to Gross (NTG) of the reservoir section (97-99%), Vshale percentages are small and PHIT=PHIE, but if heterolithic facies are to be included in the future, PHIE may be considered. Correctly the P10, Mean & P90 reservoir parameters have been chosen to represent the current interpretation of the Bunter Sandstone over the project area and beyond. These numbers could prove pessimistic as the few current logs and core within the structure have actual values that exceed the P10 case, NTG ~97-99%, PHIT ~20-24%, Permeability ~400-600 mD. This uncertainty would be decreased if more wells were drilled into the structure for disposal and/or monitoring.
NMR data shows very optimistic permeabilities, exceeding 1000 mD in the best quality sands. More work with these data could be done, with porosity-permeability crossplots for each electrofacies applied to future models. Low Net to Gross intervals could be included as well, resulting in an increase of modelled reservoir and injectable volume.
If dynamic modelling shows fluid movement to the northwest, it would be beneficial to add additional area to understand pressure and CO2 concentration beyond the spill point.
The two grids used, 2 million cell ‘Coarse’ and 100 million cell ‘Fine’ model, do an excellent job of testing the effect of vertical resolution on the distribution and concentration of CO2 over time, in this case 400+ years. Runs on both models showed a difference of 3.5% CO2 concentration at the ~200-year mark, this difference was no longer observed as the system equilibrated at the ~400-year mark. Moving forward, all modelling can be done using the coarse model, which will save static model building time as well as significant dynamic simulation time.
Variogram parameters used for petrophysical modelling are reasonable, and future versions will be able to further define the direction and extent of each facies based on 4D seismic and pressure data acquired during the life of the project.
Uncertainties in the model have been well defined and next steps could include further work on lateral continuity and architecture of baffles. The model is constructed such that more detailed modelling of low NTG intervals along with pressure data acquired in the future can easily be added to help delineate these baffles. Faulting has not been included in the current model, but has been considered and could be added if history matching requires it.
Document name: Primary Store Geological Model & Report
Reference number: NS051-SS-REP-000-00014
Document length: 70 pages
Topic area: Geological model for CO2 sequestration
Project: Net Zero Teesside / Northern Endurance Partnership
Original report date: August 2021
Original author: BP Exploration Operating Company
Reviewer names: Graham Simpson, PhD, P.Geo, Graham Dolce, P.Geo.
Reviewer organization: GLJ Ltd
Date of review: February 2023
The CCUS Hub is designed to support policy-makers, potential hub developers and emitters interested in setting up a CCUS hub by sharing learnings from the most advanced hubs and identifying new ones.