Best Modelling Approaches on Carbonate Reservoirs, a Case Study From the Majella Mountain, Italy

The oil-rich North-West flank of Majella Mountain was investigated by drilling about 200 wells, for exploitation purposes during 1942/43. We digitalized the dataset of Lettomanoppello-Fonte di Papa-Fonticelle area including average bitumen percentage property logs derived along the vertical profile of the wells (43). We were able to build a base case that represents the almost real distribution of bitumen in the subsurface, thanks to the data density in respect of reservoir variability. We tested different modeling tools using the Petrel software, to assess the best approach to model such reservoir. We conducted several blind well tests, in order to find the best approach to replicate the well toggled off. The best approach at the whole reservoir scale resulted to be the SGS algorithms with a variogram since it seems to be the best in mirroring the variability of the property of bitumen percentage. This solution can be used in the first step of exploration phase for similar reservoir, to estimate the percentage of hydrocarbons as we have done for the Cigno Oilfield.

SGS algorithm with Trend Model, limited to the vertical mean trend built from the upscaled logs, resulted to be the best approach in the blind well test.