APPLICATION OF ARTIFICIAL NEURAL NETWORK TO IDENTIFICATION OF LOW RESISTIVITY RESERVOIR IN THE NORTH BLOCK OF MINQIAO OILFIELD

More and more unconventional reservoirs have been discovered recently in oil and gas exploration. The reserves and production of low resistivity reservoirs are also increasing. As an unconventional one, the low resistivity oil-bearing strata are affected by many factors and the common logging interpretation method is not so efficient in evaluation of the oil reservoir. The Minqiao Oilfield is a complex faulting structure and a typical low resistance reservoir. The logging data is difficult for interpretation. The BP artificial neural network algorithm, learnt from known samples, is applied in this case using the self-compiled software. We have recognized the low resistivity reservoir E₁f₃^1-1 successfully in the north fault block of the Minqiao oilfield. The results have led to the increase in 8 new oil wells, 0.53 km² of new oil-bearing area and 23.62 million tons of new geological reserves of oil as the economic benefits in the reservoir of E₁f₃^1-1. It is also contributed to the deepening of understanding of the oil and gas distribution pattern and reservoir type in the fault block, with great theoretical significance.