Abstract: The reservoir of PU-Formation in the Punan Oilfield is so thin and poor that it is very difficult to solve the question that the distinguished accuracy of the narrow&laminated sandstone is low by inversion of seismic wave impedence. This paper builds predicated model of reservoir lithology integrating geology statis-tics and modeling technologies, adopting the Gauss continuance
simulating methods in virtue of consequences of wave impedence inversion to carry through lithology modeling of natural electric phase body and deep side resistance body, aiming at characteristics of this area, mudstone and sandstone can be recognized obviouslynatural electric phase curves and it's easier to distinguish the oil-bearing property of reservoir by means deep side resistance curves, based on the original datum of inversion of seismic wave impedence. The geological statistics characteristic of the simulating consequences match with ihosr, of original datum, and the matching rate of predicted thickness of sandstone reaches 78% according to newly drilled 18 wells. It indicates that the predicted problem can be well solved by this technology.

Key words: thin&, poor reservoir, inversion of wave impe-dence, random simulation, reservoir prediction