Carbonate Reservoirs are well known for their heterogeneity in terms of porosity and permeability. In this field case from UAE onshore a sharp degradation of petrophysical quality was noted at the gas-water contact, in relation to diagenetic cementation, and led to an independent modelling for the aquifer, as well as an independent modelling of the reservoir with a data filter for the gas pool only. A merge of the 2 grids was then performed.
In this field a major bias on well data distribution from crest to aquifer affects the geostatistical histogram evaluation: wells are concentrated in the crest down to mid-flanks while there are few wells from mid-flanks to the aquifer. Consequently distribution histograms of petrophysical data from the whole model should not respect well data histograms (whether, core-, log- or cell-derived).
The methodology of separating gas and aquifer modelling, and of separating well derived histograms from model-derived ones led to the following results:
– Better capture of the reservoir degradation with depth in the gas pool, – Better capture of the sharp degradation break in the aquifer.
This paper is focusing on the methodology of how to build two separate models in the gas pool and the ‘aquifer for each petrophysical property and how to combine them in one property model to honour reservoir heterogeneity.
Integration of all data at all scales and constant QC between database sources (logs, cores, seismic, dynamic history) were the means to produce a geomodel capturing the key heterogeneities of the reservoir, those with major impact on fluid front migration during production.