Tahe oilfield is a famous marine Paleozoic giant oilfield dominated by paleokarstic reservoirs of the Ordovician carbonates.Caverns and fractures around them are the main pore space.Numerous small fracture-vugs act as important pore space and and flowpaths.Although the small fracture-vugs can be recognized on outcrops

cores and image logs

they are not readily recognizable in intervals without core and image logging data

seriously influencing the study of fractured-vuggy reservoirs and the establishment of accurate reservoir geological model.To tackle the problem

we calibrated the conventional logging data with cores and image logs and selected five vug-sensitive parameters and three fracture-sensitive parameters.These parameters were used to build a norma-lized and weighted function for identification of small fracture-vugs in Tahe oilfield.We used DEN-V

sh

R

LLD

-V

sh

cross plot method to recognize the types of fillings

including clastics

breccias and chemical fillings.The results of recognition are fairly good compared with that from core observation and image log interpretation.The vug recognition rate is above 81%

fracture recognition rate is larger than 82%

and cave fills recognition rate is over 84%.The method can meet the basic requirements for identification of small vugs and fractures.The application of the method to 117 wells in Tahe oilfield proves to be quite encouraging and the results also show the potential of the method for identification of caverns and fractures with a height over 0.5 meters and their fillings.It provides a new approach for fine study of reservoir space of paleokarstic reservoirs.