Uncertainty Analysis of Two-Dimensional and Three-Dimensional Modeling in Exploration Context Based on Multivariate Computational Experiment

The same geological object can be modeled both in two-dimensional space and in three-dimensional one. Currently, both three-dimensional and two-dimensional modeling is carried out using specialized programs (often the same ones) and includes the same stages of import and raw data preparation, performing cross-well correlation, so the construction of a 2D model is just as laborious as a 3D one. In this situation, the issues of reliability of the models, how the model is similar to the studied object come to the fore. Two-dimensional geological modeling has a number of known limitations, not allowing taking into account the vertical heterogeneity of the simulated object, the vertical separation of reservoir layers and their communication in space, as well as data for wells located outside the saturation circuit. Three-dimensional modeling by introducing an additional dimension increases the complexity of the model and the number of uncertainty factors. Since the reliability of the created models significantly affects the rationality of investment decisions, the ratio of the contribution of uncertainty factors of two-dimensional and three-dimensional modeling requires careful study. This work offers one of the ways to solve this problem, based on the formulation of a series of multivariate computational experiments.