DETERMINATION OF WATER SATURATION FROM WELL LOGS HONOURING THE SHALY SAND PHYSICAL MODEL IN THINLY BEDDED RESERVOIRS

The accurate determination of reservoir parameters in thinly bedded reservoirs encounter problems of being interpreted as shales or water bearing sands, underestimation of reserves or the reservoir entirely missed partly due to log resolution. Although there are a number of saturation equations, most log dependent methods assume that only one clay type is present and similarity between properties of the clay in adjacent shale beds and in reservoir which may not be true due to diagenetic processes. Others like Waxman-Smit although have some theoretical basis, are difficult to obtain from logs and requires an iterative process for the determination of water saturation. On the otherhand, core measurements of water saturation apart from being expensive, run into error since it is difficult to preserve the water contained in the core under insitu conditions.

Due to the foregoing, a model based on parallel resistor network, that honours the shaly sand physical model and its parameters, makes provision for the effect of clay mineralogy, and can be computed from well log data is developed considering not just the volumetric but also geometric aspects of current flow path and heterogeneity of the formation to current flow.

Several NorthSea reservoirs portions with greater than 10% clay content have been used for testing the model and it has been found to work well. Comparison with other water saturation models was made and results show that Archie, Indonesian, Hossin’s model overestimates (i.e are pessimistic) indicating that the zone is water bearing with >60% water saturation. Crossplots of water saturation were made against formation true resistivity and resistivity index at different m and n values. The emphasis on it honouring the shaly sand physical model creates the potential for being generic. This will enhance more accurate discrimination of formations above and in the transition zone and for permeability estimations which are known to be sensitive to errors in water saturation and porosity.