Decision making in field development has been crucial in petroleum exploration and production activities where numerous uncertainties exist in the complete process. In the development process of new fields, any additional data to minimize uncertainty is important, as it decreases risks and provides better insight and robustness in decision making. Reservoir analogue techniques provide valuable information in this process. In the reservoir analogue method, oilfield historical data is used along with descriptive statistics and case-based reasoning as a guide or tool in carrying out field-development scenarios, which enables the field-development teams to make decisions while knowing the limitations of past real-life performances in fields. This will greatly help field-development teams make decisions based on what to expect economically, as they will have a reference on recovery factors and other indicative parameters to make more informed decisions on optimization parameters and field development. With advancements in technology and data-driven techniques, data has become more readily accessible; however, turning data into knowledge has become a challenge. Using reservoir analogues is not a new concept; however, the rationale in finding the analogues has not been studied or published in detail so far. The reservoir analogue technique involves a delicate use of categorical and parametrical attributes, classification of attributes, along with strong engineering judgment on individual cases. Thus, illustration of the proper use of algorithms plays an important role in obtaining successful results. In our paper, we proposed two approaches employed in the case-based reasoning system to find the reservoir analogues, where each reservoir is represented by not only numerical attributes but also categorical attributes. In each approach, several different distance metrics and similarity measures for numerical and categorical attributes, respectively, are studied.

You can access this article if you purchase or spend a download.