EAGE/SEG Research Workshop on Reservoir Rocks - Understanding reservoir rock and fluid property distributions - measurement, modelling and applications

Drilling for sub-salt Paleozoic carbonate reservoirs in the North Caspian sea, offshore Kazakhstan, presents numerous subsurface and surface challenges. The goal of such drilling is to safely penetrate carbonate reservoirs that range in depth from 4 to 5 km, which are likely to contain over-pressured H2S-rich hydrocarbons.

Fluid transport in sedimentary basins is a well-known phenomenon. It happens due to stresses caused by tectonic events and/or gravitational force, due to sediment loading or erosion. Typically, if the loading process is rapid, fluid expulsion through compaction can be severely `retarded', especially, in fine-grained sediments with `low' permeabiliry, such as silts or clays. This results in stress re-distribution withm the column - a greater proportion of the overlying weight of the sediments is borne by the fluids than in the case had the sediments compacted normally, causing a decrease in the stress acting on the rock framework. This phenomenon is known to occur in most sedimentary, clastics basins (geopressure).

Ultrasonic P and S wave travel-time measurements have made on Cretaceous sandstones from the Locahaline silica mine in West Scotland whilst being subjected to increasing pressure in a Hoek cell, The resulting dryframe P and S wave vebcities for a sequence of pressure values between 0 and 70MPa are then transformed into elastic modeli using density calculated from pore volume squeeze-out on a sister sample. Both the exces normal and the tangential compliance are calculated by using differential effective medium theory to compensate for host matrix softening due to evolution of the equant porosity.

Sandstone samples from the Carnarvon Basin in the Australian North West Shelf have been tested under triaxial loading conditions with independent control of pore pressure. Ultrasonic waveforms were recorded during these tests.

The appraisal process entailed in determining whether or not seismic monitoring was worthwhile in the special context of an HP HT field involved working with different kinds of data from different sources and at different scales.

Tree elastic parameters necessary for Gassman fluid replacement modeling, namely the velocity of the pressure wave (Vp), velocity of the shear wave (Vs) and bulk density (Rhob). These can be estimated from seismic data through inversion, using direct wireline log measurements or by conducting laboratory measurements on dry cores using ultra sonic frequencies. Whereas the three elastic parameters are measured directly on cores and in wells, seismic is unique in the sense that we can most accurately invert for acoustic impedance (Z-p) and with 4-C seismic also quite accurately estimate gamma, or Vp/Vs ratio, for a sufficiently thick reservoir interval.

Summary not available

Seismic recording offsets have historically been restricted for the following reasons: expense, operation difticulty, and the good use of longer offsets is technically more difficult. For current ocean bottom acquisition technology the rost and operational difficulty of acquiring Jonger offsets quickly becomes prohibitive but in the case of conventional marine streamer and onshore acquisition these issues are relatively trivial leaving only the extra technical difficulty as an obstacle. While the longer offsets do present a challenge to the seismic processor, the information gained in unraveling their kinematic and amplitude behaviour provides us with the opportunity to team much more about the subsurface than just the usual structural image.

The prediction of the spatial distribution of petrophysical properties within a heterogeneous reservoir is affected by significant uncertainties when based only on well information. However, integrating additional constraints such as 3D seismic data and sedimentary concepts can significantly improve the accuracy of reservoir models and help reduce uncertainties on predictions away from wells.