The studied reservoir consists of turbiditic slope channels which form elongated sandy bodies with dimension of 4 to 12 Km in length and 1 to 4 Km in width. The thickness is varying from few metres to almost one hundred. These channels are hydraulically separated and characterized by high values of porosity and permeability.
These sand bodies were identified and porosity characterized by means of seismic data. The first three wells have proved the correlation between gas-bearing sands and "low-impedance" seismic responses. Nevertheless, since those wells did not cross all the sand bodies (only 6 out of 14) the main problem at that stage was the risk assessment of the hydrocarbon occurrence in the remaining bodies.
In order to assess this risk the first approach was to identify, for all the bodies, relevant seismic criteria (Amplitude strength, Signal noisiness, Fragmentation and Recon AVO) to use for a qualitative evaluation defining a ranking of the undrilled bodies. The following step was to apply a Bayesian approach to assess the relevant probability of hydrocarbon occurrence. This technique requires the estimation of two parameters: prior probability based on geological area knowledge and likelihood derived from seismic indicators.
On the base of the probability ranking a limited number of scenarios was defined. For each scenario a dedicated plan of development was assigned. A new appraisal well was drilled after this evaluation. This allowed verifying the validity of the ranking system used to estimate probability of success with regards to the undrilled segments. The results show good agreement between seismic data and well’s results with one (positive) exception.
Moreover a traditional "Risk Analysis" was performed in perforated bodies to sample all the critical parameters (static and dynamic) and to build several models. By these models a set of volumetrics and production profiles were obtained.