Optimization of reservoir performance is strongly related to feasible evaluation of reservoir uncertainty. In modern reservoir engineering, a substantial amount of effort is allocated to maintenance and upgrading of what is believed to be the current best realization model of the reservoir, where reservoir uncertainty is demonstrated through parameter variation. This technique is frequently applied even though the nature and extent of uncertainty associated to reservoir characterization, strongly suggests a multiple model approach, i.e. visualization of reservoir uncertainty through realization of alternative models. Awareness of this problem is particular important in the case of strongly faulted gas- and condensate reservoirs.
The model described in this paper is designed for production from gas or gas condensate fields, where modelling of highly faulted and compartmentalized reservoirs is severely impeded by the nature and extent of known and undetected faults and, in particular, their effectiveness as flow barriers. The model redefines individual fault block traps in terms of block volume and cross-flow probability to adjacent blocks, where geological uncertainties are carried through in the model by error propagation techniques. Three different realizations of reservoir models are created; one representing the most likely (average) reservoir while two others represent the extreme alternatives, reflecting the reservoir uncertainties. The fault block evaluation is treated as input to material balance calculation, where reservoir production is assumed to take place from individual block volumes.
The reservoir model is easy to modify by varying block volumes and/or cross-flow probabilities, according to improved knowledge about the reservoir or simply as a sensitivity test of the degree of communication across fault barriers. The model also estimates the total well drainage coverage and the sequence of blocks from where wells should start production. The model may also certify the research area of interest, where improved information about the reservoir has the greatest impact on improved reservoir performance.