| Summary: | Passive microseismic monitoring provides a non-invasive method of monitoring deformation and changes in the stress distribution within a rock mass. Recently the petroleum industry has applied and focussed such studies on the monitoring of hydraulic fracturing stimulation: the fracturing of a reservoir by the injection of a high pressure fluid into reservoir causing the rock to fracture, increasing the permeability. However, the development of passive seismic monitoring of a fully operational oil field has been slow largely due to economic constraints limiting the geophone array to be deployed within a single vertical borehole. The challenge is thus to refine the seismic methods applied to datasets from the vertical arrays and to explore the extent to which further investment in more sophisticated seismic arrays would advance the understanding of the reservoir architecture and evolution. The Ekofisk reservoir in the North Sea provides an excellent opportunity to address these issues and is the focus of this study. Discovered in 1969 the Ekofisk field is located within the Central Graben of the Norwegian North Sea and is operated by Conoco-Phillips. It was one of the world's first economically viable chalk reser- voirs and comprises two main oil bearing intervals: the Ekofisk and Tor Formation chalks, separated by a relatively impermeable siliceous chalk. Since the onset of production in 1971, the reservoir has expe- rienced appreciable subsidence. A water injection program was initiated but has done little to mitigate the problem. In April 1997, one of the first hydrocarbon passive micro seismic monitoring studies was undertaken over an 18 day period at Ekofisk in an attempt to understand the mechanism of deformation. Data were acquired using a six station geophone array deployed in a vertical borehole. Fundamental to the study of seismicity at any scale is the accurate determination of the event source. Events for borehole microseismics are usually located using a ID velocity model, P- and S-wave arrival times and the polarisation azimuth of the P-wave particle motion. However, in the case of all sensors being deployed within a vertical or near-vertical bore hole such analysis leads to an inherent 1800 ambiguity in source location. In this study, this ambiguity is removed by using the back-projecion of the dip of the particle motion from multiple stations until they converge on one side of the well or the other as a priori information to constrain the initial source location. The procedure is developed and tested using a dataset acquired from another field during hydraulic fracture stimulation, where event locations are known. Applying this procedure to the Ekofisk dataset 627 events are successfully located with coherent features clustering around production/waterflooding wells and fractures. Most are located less than 250 m away from the monitoring well and at a depth of ",3 km in the Ekofisk chalk formation. Little seismicity is observed from the underlying Tor Formation chalk, which is separated from the Ekofisk Formation by an impermeable layer of siliceous chalk. There is no evidence of seismicity in the overburden. Repeating earthquakes (multiplets) are identified and relatively re-located in order to enhance the resolution of active features and gain further insight into the mechanism of deformation. Qualitative analysis of the waveforms of the multiplets shows a number of potential mechanisms such as production/waterflooding induced activity, fault re-activation and stress triggering. Having established precise hypocentres for the Ekofisk microseismicity the final goal of this study is to gain information along the source-receiver raypath using a shear wave splitting study of anisotropy. An automated shear wave splitting approach is applied to the Ekofisk dataset yielding 1125 reliable measurements. A number of near-vertical fracture sets with fracture strike orientations of NE-SW and W-SE agrees with previous core based studies of Ekofisk. In summary, this thesis shows that a single borehole deployment of geophones can be used to gather detailed information about the spatial and temporal variation in seismicity in a hydrocarbon setting. The seismic data can then be used via study of seismic anisotropy to place fundamental new constraints on the state of stress, mineral alignment, layering of sedimentary structures or the presence of aligned fracture sets all of which have profound implications for reservoir management.
|
|---|
