4.2.3 The Longannet project (UK)

Both for the already cancelled Longannet and Kingsnorth projects non-confidential FEED studies were released by the British government (Kingsnorth Carbon Capture & Storage Project, 2010; UK Carbon Capture and Storage Demonstration Competition, 2011). All information in the sections on Longannet and Kingsnorth are based on the information from these FEED studies.

The Shell Longannet - Goldeneye project considered storage in a depleted gas field, with as target the Lower Cretaceous Valhall formation (Captain Sandstone member) at about 2 km depth (and to a lesser extent a secondary reservoir above, at a depth of about 1.5 km), in the central part of the UK sector of the North Sea. Each reservoir has a seal complex of mudstones and the seal complex above the primary reservoir also includes marls.

The monitoring plan comprises to a large extent the same elements as the ROAD project, namely:

  • Environmental baseline monitoring using multi-beam echo sounding and seabed sampling and continuous tracer injection;
  • Well integrity using pressure and temperature gauges, distributed temperature sensing (DTS), tubing integrity logging and seabed CO2 detection below the platform;
  • CO2 injection conformance using pressure, saturation and flow monitoring;
  • Lateral and vertical irregularity and plume conformance using time lapse seismic.

The main difference observed consists of the status of the time lapse seismic data. In the ROAD project, time lapse seismic data is considered as contingency monitoring, triggered by an irregularity observed on other monitoring data (essentially the pressure data), whereas at Goldeneye time-lapse seismic monitoring is part of the standard monitoring program. This can be justified by the fact, that in the shallower Goldeneye reservoir delineation of the plume in the reservoir is expected observable on the time-lapse seismic data, in contrast to the deeper P18-4 reservoir of the ROAD project.

Furthermore the seabottom imaging and sampling program seems more extensive at Goldeneye than for ROAD, which can be explained by the presence of existing pockmarks at the seabottom at Goldeneye and by the presence of seven abandoned wells in the area. These are absent in and above the P18-4 reservoir of ROAD.

The last major difference consists of the number of injection wells, originating from a conversion of the five existing gas production wells into injection wells. One of these wells will serve as a monitoring well at an early stage of the project to see the saturation front passing by. At ROAD the P18-4 reservoir is spatially much more confined by faults, with no additional wells penetrating the reservoir. The only well partially used for monitoring will be the P18-1 well, which penetrates the neighbouring P15-9 field. Regular checks for CO2 contents might be taken, as well as pressure measurements to investigate potential (though highly unlikely) communication between the two fields.

Otherwise the philosophy at Goldeneye and P18-4 (ROAD) is highly comparable in the sense, that the plan aims essentially at detecting irregularities and then triggers a more extensive contingency monitoring program.

Selection of the monitoring technologies at Goldeneye is based on the following factors: risk relevance, measurability, operational constraints, competitiveness and proven technology.

The Bowtie Method (DNV, 2010a) as well as Shell's own risk matrix are used for a systematic risk assessment of events with the potential to affect storage integrity and performance.

The program complies with the high level requirements of the storage directive and its implementation in the UK and addresses the EU ETS MRG guidelines as well, since leakages for the full chain are addressed. The shallow seabed monitoring of the Longannet project addresses comprehensively the OSPAR guidelines.