Depth Dependent Rock Physics Trends for Mesozoic Reservoirs in the Norwegian Barents Sea

Sabine Klarner¹, Caroline J. Lowrey², and Ron Borsato^1
1Reservoir, PGS, Moscow, Russian Federation.
²Reservoir, PGS, Lysaker, Norway.

Rock physics studies are used to improve our understanding of whether the properties of rocks encountered in a well allow for seismic delineation of target lithologies and pore filling fluids and to give an uncertainty estimate of amplitudes and corresponding elastic responses. To date, most of the published trends are derived from grain-supported, pure quartz sandstones. However, rock physics appropriate to mineralogically and texturally complex reservoirs, such as the main Triassic reservoirs of the Barents Sea are poorly understood. Lithologically complex reservoirs exhibit a wide range of elastic properties that depart, to varying degrees, from those of grain-supported, quartz dominated, clastic reservoirs. The current project was undertaken to develop a systematic approach to understanding the seismic response of these reservoirs and to draw conclusions about the possible general features of the elastic behaviour of lithic-rich clastics. The unique approach, of including the effect of compaction by introducing statistically valid depth trends of elastic properties, significantly improves the predictive power of the models at basin scale. The total porosity of the investigated reservoirs is strongly influenced by lithology distribution and sorting. Clean sandstones exhibit the highest porosity and lowest velocity, with a gradual transition to low porosity, higher velocity shaley sandstones. Despite this trend, the p-wave velocity shows a broad scatter and is not a reliable lithology indicator. To find a physical property discriminating the various lithology types, s-wave velocity was analyzed. Clean sandstones tend to have higher Vs than shaley sandstones or shales at the same Vp level, which makes the Vp/Vs ratio a physical lithology discriminator for the area. For a significant proportion of samples we observe a lower level of the s-wave velocities, than predicted by the empirical Greenberg-Castagna trends for quartz sandstones. These phenomena must be accounted for, by some means, when calculating synthetic s-wave logs in wells where s-wave is not recorded, and especially for calculation of missing logs in the Russian sector of the Barents Sea. We have been able to empirically extract local depth dependent trends for elastic properties of the main lithotypes. Based on analyses of the available well data, principle stochastic scenarios for different reservoir types and their seismic responses (changes in lithology, porosity, fluid properties) have been modeled to cover a wider range of probable cases in the subsurface. The established rock physics trends will provide a sound base for combining well and seismic data, to assess the feasibility of pre-stack inversion and litho/fluid prediction on seismic data. These models will also aid in the design of physically and economically optimized acquisition geometries for future seismic acquisition programs and form the base for any future 4D projects in the area.

 

AAPG Search and Discovery Article #90130©2011 3P Arctic, The Polar Petroleum Potential Conference & Exhibition, Halifax, Nova Scotia, Canada, 30 August-2 September, 2011.

����������������������������������������������������