Relationship between burial history and seismic signatures in the Horda Platform area, Norwegian North Sea

The Horda Platform area, in the Norwegian North Sea, has a complex burial history, including rifting and fault block rotation followed by uplift end subsequent erosion. These processes have affected the elastic properties of the rocks, thus it is important to make proper use of this information when predicting AVO (amplitude versus offset) responses in the area. The net-uplift in the study area was estimated using several different techniques. First, sandstone modeling was performed with the integration of rock physics, cement volume estimation and burial history, honoring both mechanical and chemical compaction. Sec- ondly a seismic interval velocity cube was utilized to predict the amount of uplift, using a background trend calibrated for Eocene shales. Even though the latter approach resulted in more spacious and structural information, both these methods generally provided similar results. Generally, increasing uplift towards the east was observed, with magnitudes up to 1500 m at certain locations which coincides well with published Miocene drainage paths and ice sheet flow lines. A seismic section was, in addition, used to predict net-uplift relative to a reference point in the Viking Graben center, indicating relative movements in a similar pattern as the other methods suggests. Vitrinite reflectance data was also used as an uplift indicator, however no conclusive results were obtained using this method. Analysis of shale trends in the region revealed some typical "hydrocarbon responses", this in both uncompacted as well as organic rich shales, which could act as pitfalls in AVO analysis. AVO signatures were modeled using the same sandstone modeling method as described above, by utilizing the uplift estimates as an input parameter. The results show good correlation with responses from seismic data at several locations, even relatively far from well positions. Finally, rock physics classification of inversion data was performed and these results generally agreed well with both real- and modeled seismic responses