Well placing using bayesian network expert system

A method for automatic determining of a well placement and a well completion type, the method comprising:

• accessing, from a database and using one or more hardware processors, geological and geophysical data associated with a candidate area for drilling a first well and with an area of a second, nearby well;
• performing, using a Bayesian Decision Network (BDN) model and the one or more hardware processors, a first assessment of a data availability score value, saturation data, fracture data, and seismic data included in the geological and geophysical data associated with the candidate area for drilling the first well;
• outputting, based on the first assessment and using the one or more hardware processors, a first probability value that indicates a potential level of production of the first well if drilled in the candidate area;
• performing, using the BDN model and the one or more hardware processors, a second assessment of an offset data score value, stress data, thickness data, and porosity data included in the geological and geophysical data associated with the second, nearby well;
• outputting, based on the second assessment and using the one or more hardware processors, a second probability value that indicates a potential level of production of the first well if a lateral wellbore is placed in a particular layer and at a particular azimuth direction in a reservoir associated with the candidate area;
• performing, using the BDN model and the one or more hardware processors, a third assessment of tortuosity data, washout data, porosity data, and a net-to-gross value associated with the first well;
• and
• outputting, based on the third assessment and using the one or more hardware processors, a third probability value that indicates the well completion type for the first well.

A non-transitory machine-readable storage medium comprising instructions that, when executed by one or more processors of a machine, cause the machine to perform operations comprising:

• accessing, from a database, geological and geophysical data associated with a candidate area for drilling a first well and with an area of a second, nearby well;
• performing, using a Bayesian Decision Network (BDN) model, a first assessment of a data availability score value, saturation data, fracture data, and seismic data included in the geological and geophysical data associated with the candidate area for drilling the first well;
• outputting, based on the first assessment, a first probability value that indicates a potential level of production of the first well if drilled in the candidate area;
• performing, using the BDN model, a second assessment of an offset data score value, stress data, thickness data, and porosity data included in the geological and geophysical data associated with the second, nearby well;
• outputting, based on the second assessment, a second probability value that indicates a potential level of production of the first well if a lateral wellbore is placed in a particular layer and at a particular azimuth direction in a reservoir associated with the candidate area;
• performing, using the BDN model, a third assessment of tortuosity data, washout data, porosity data, and a net-to-gross value associated with the first well;
• and
• outputting, based on the third assessment, a third probability value that indicates the well completion type for the first well.