Quality Control of 3D GeoCellular Models: Examples from UAE Carbonate Reservoirs

3D geocellular static models are the key input for fluid flow simulations with the main aim to predict the future reservoir performance for a particular recovery scheme. Since the predictability of the dynamic model depends on the quality of the geocellular model, it is imperative that the input data, the modelling workflow, methodologies and approaches are verified and validated prior to the sanction of the geocellular model. The objective of this paper is therefore to discuss the process of performing quality assurance and quality control (QA/QC) of 3D geocellular models exhibiting real field examples from the Middle East carbonate reservoirs. 3D static models are built using data from multiple sources, at different scales and with different degrees of uncertainty. The validation and reconciliation of all the data is of paramount importance. The procedure to build any geological model is very similar provided all the data is available. Some variations in the procedure are expected depending on the complexity of the phenomena to model, but must of the time workflows divert based on data quality and data availability. In this paper we discuss the use of key validation checks for each step of the modelling process taking into account the data quality and field maturity, namely for the 1)- structural framework modelling, 2)- facies modelling, 3)- porosity modelling, 4)- permeability modelling, 5)- rock type modelling, 6)- water saturation modelling, 7)- upscaling and 8)- uncertainty analysis. The use and validation of the applicability of secondary variables in the petrophysical modelling, such as acoustic impedance from seismic inversion, is also addressed. From the analysis of multiple geocellular models, inconsistencies were detected at different stages of the modelling process, starting from the well surveying with implications to horizontal well positioning within the framework, to the modelling of facies and petrophysical properties, with inconsistencies on variogram model parameters. Also, the validation of the velocity modelling and time-depth conversion used for the structural framework was validated by comparing FWLs depths against spill points. Furthermore, the quality of the facies model could be verified against regional facies belt maps (similar variogram azimuths are expected) while the validation of the permeability scale-up at well level could be achieved by reconciling with well test kh data. These are just a few examples of the material discussed in this paper. The novelty of the quality assurance process pertained to 3D geological models is the identification of appropriate metrics with key performance indicators for each step in the modelling workflow. At the end of the QA/QC process the models are ranked in quality and technical gaps identified for subsequent model improvement. Guidelines and best practices are also presented in this paper.