Abstract
Net-To-Gross (NTG) is the ratio of pay rock volume to the gross rock volume estimated over an interval used to compute hydrocarbon-initially-in-place (HCIIP). Identifying NTG as a series of discrete data points along the well path where the rock is ascribed as reservoir (NET = 1), or non-reservoir (non-NET = 0) based on a suitable cut-off estimated from a combination of petrophysical logs is a first step. In a deltaic setting, prevailing depositional processes result in inherent vertical and lateral variability with the effect that the determination of NTG is scale sensitive and interval dependent. This reality imposes a requirement of deliberate focus in 3D NTG modelling to preserve the depositional heterogeneity and ensure that any integrated reservoir modelling effort is geologically representative. This paper discusses different practices in the use and upscaling of NTG from well data and distribution in a static model, by comparing three approaches: (1) upscale and distribute using Sequential Gaussian Simulation (SGS) biased to facies, (2) upscale and distribute independent of the facies model but biased to established NTG regional trend, and (3) generate a discrete model based on a combination of the facies model and NTG data from the wells. The best method, which is still a simplification of reality, will be one where the model can rightfully predict NTG in undrilled areas and assess the impact on recovery estimation. Assessing these methodologies using a mature field in the coastal swamp depobelt in the Niger delta helps build a case for model simplification in an adjacent partially appraised field in the same depobelt. Thus, providing a simplified approach and clarifying insights on the conundrum of utilizing discrete NTG ‘logs’ to build a continuous property that is useful for HCIIP estimation but has minimal impact in the dynamic behavior of the reservoir.
