Artificial intelligence (AI) evaluation of current reservoir pressure distribution based on oil production data

The paper investigates a quick-look approach for the assessment of the distribution of current reservoir pressure based on production data.The method is based on an algorithm that includes calculation of the current distribution of values of stream functions, potentials and flow velocity in a selected area.The method allows monitoring the factual distribution of the current reservoir pressure of the producing horizon in the area under consideration, as well as evaluating the effectiveness of the impact on the reservoir in order to maintain reservoir pressure.Based on the proposed method, it is possible to create Artificial Intelligence (AI) technologies for analyzing operational data, machine learning to predict changes in reservoir pressure.The use of neural networks in the integration of geological, geophysical and operational data, operational risk management allows to create automatic expert systems to optimize the process of development and operation of oil and gas fields in conditions of insufficient information.The accomplishment of the investigated approach carried out applying data samples from Oil Rocks field (Horizon X, Block V) provided high accuracy for values obtained by calculations.The average relative error rate of the calculated values of reservoir pressure to the actual values of bottomhole pressure measurements in wells is no more than 1%, and the average calculated value of reservoir pressure in the productive strata in the study area is in conformity with its actual reduced value.