Prediction of Poisson's ratio for carbonate rocks using ANN and fuzzy logic type-2 approaches

Young's modulus and Poisson's ratio describe the elastic behavior of rock. It is extremely important to determine these parameters in order to minimize the risk associated with the oil and gas well engineering. The estimation helps in several areas of drilling and production such as well placement optimization, design of completions, mud weight calculations, and hydraulic fracture geometry. Each one of these factors play a part in maximizing the recovery of hydrocarbons and in taking crucial decisions for an appropriate field development strategy. Poisson's ratio is second most important parameter in understanding the elastic behavior of the rock and plays a critical role in almost all processes such as drilling, reservoir simulation, and production. It is an essential component in geomechanical earth model (GEM). The Poisson's ratio is estimated based on empirical models and artificial intelligence models. These models are construction from data that has different types of uncertainties. This paper presents an Artificial Neural Network (ANN) as well as Fuzzy Logic Type-2 (FLT2) approach for prediction of static Poisson's ratio. FLT2 is able to incorporate the uncertainties in measurements and still give a robust solution to a given problem. Well log data is used as input and laboratory determined static Poisson's ratio is used as output in the artificial intelligence (AI) tool. The data were collected from a range of experiments conducted on carbonate rocks covering a wide range of input and output values. The model takes care of uncertainties in the input and output data and is therefore a better approach in establishing a relationship between them and in predicting static Poisson's ratio for new input data.