Novel NFDM Workflow: An Approach to NMR Data Processing for Accurate Fluid Typing and Saturation Determination

Abstract

Abstract Accurate fluid typing and quantitative saturation determination are essential to evaluate reservoir production potential, which is a well-known deliverable by NMR measurements. However, challenges remain in obtaining these critical reservoir properties due to the ill-posed nature of Inverse Laplace Transform (ILT)-based NMR data processing and partitioning of overlapping continuous T1/T2 or 2D T1T2 distributions. This work presents a straightforward and efficient data processing workflow, the NMR fluid discretization method (NFDM), based on a new discrete inversion method, Anahess. The first step in existing NMR data processing and interpretation is to apply the ILT methods with known issues of spurious peaks and broadening at the short relaxation time region. The overlapping peaks in 1D and 2D continuous distributions reflect the continuous oil and rock pore size distributions affected by the regularization factor. The second step of partitioning these overlapping peaks for fluid types is usually performed using manual cut-off, Gaussian decomposition, or machine learning methods, with considerable uncertainty and inconsistency. The new inversion method, Anahess, generates a limited number of unique discrete components that directly correspond to fluid types from the raw relaxation data. The new NFDM workflow, a commonly-used ILT and an improved ILT method are applied to T1T2 measurements on synthetic samples, bulk fluid mixture (oil and water), and shale samples from multiple wells. For the synthetic samples of discrete components, the new method returned the components of accurate amplitudes and T1T2 relaxation times, while the ILT methods only gave approximate results through manual partitioning. The NFDM results from mixed bulk fluids also agreed very well with those from separate measurements. The true potential of the new workflow was illustrated in the analysis of shale samples, where different fluid types such as free oil, free water, absorbed oil, and clay-bound water, were easily determined and quantified by direct assigning the discrete components based on their T1 and T2 values according to a general fluid type NMR scheme. The production of different wells is expected to correlate positively with the free oil saturation. In summary, fluid typing and saturation determination were achieved using the new NFDM workflow with the Anahess inversion method, which eliminates the uncertainty, ambiguity, and inconsistency of existing ILT-based two-step approaches. Further applications to well logging and other NMR measurements using discrete inversion methods, including Anahess, are worth investigating.