Advanced Eor Process Upscaling Using Multi-Scale Digital Verification Technology and FDP Application

Moreno Ortiz, Jaime Eduardo (Schlumberger) | Klemin, Denis (Schlumberger) | Savelyev, Oleg (Gazprom Neft Middle East B.V.) | Gossuin, Jean (Schlumberger) | Melnikov, Sergey (Gazpromneft STC) | Serebryanskaya, Assel (Gazprom Neft Middle East B.V.) | Liu, Yunlong (Schlumberger) | Gurpinar, Omer (Schlumberger) | Salazar, Melvin (Schlumberger) | Gheneim Herrera, Thaer (Schlumberger)

OnePetro 

Use of numerical models to characterize and evaluate reservoir potential is an industry wide practice, with increasingly more development decisions being substantiated by finite difference models. Advances on hardware and software, along with the ability to effectively incorporate accurate process physics, makes simulation a robust tool for field development decisions, particularly on complex operations such as enhanced oil recovery and/or reservoirs with challenging heterogeneity and pore structures. Use of these models does not come without its challenges where data requirements (and use of special characterization both at lab and field level) increase as does the reservoir characterization granularity and thus model sizes. Unsurprisingly the increase of model precision and data requirements amplifies non-uniqueness of the numerical solutions obtained during any field evaluation including field development planning (FDP). Incomplete/inconsistent datasets pose a further challenge to the accuracy (and arguably risk) of the forecasts by introducing further uncertainty on the process characterization. Use of complementary technology such as digital rock, that would enable mitigate impact of such uncertainties in a timely manner -either at field or laboratory level, is thus highly desirable particularly when dealing with enhanced oil recovery. Compounding the non-linearity effect of the EOR agent characterization is the effect of the augmented numerical artifacts (dispersion, dilution, etc) of which complex chemical implementations are prone to, making the upscaling process from laboratory dimensions to field more complex.