Auto-detection interpretation model for horizontal oil wells using pressure transient responses

摘要: Directional drilling is an excellent option to extend the limited reservoir reach and contact offered by vertical wells. Pressure transient responses (PTR) of horizontal wells provide key information about reservoirs drilled. In this study multilayer perceptron (MLP) neural networks are used correctly identify models from pressure derivative curves derived To end, 2560 for six distinct generated design a machine-learning classifier. A single hidden layer MLP network with 5 neurons, trained scaled conjugate gradient algorithm, selected as best This smart classifier provides total classification accuracy 98.3%, mean square error 0.00725, coefficient determination 0.97332 over whole dataset. Performance proposed verified real field data, synthetically noisy PTR, some signals outside range initially assessed training plus testing data subsets. The developed can reservoir-flow model probability close 0.9. novelty work that it employs large dataset (not vertical) well tests applied includes train verify reliably achieve high-level prediction accuracy. CIted : Moosavi, S.R., Vaferi, B., Wood, D.A. Auto-detection interpretation oil using responses. Advances in Geo-Energy Research, 2020, 4(3): 305-316, doi: 10.46690/ager.2020.03.08