Abstract
A history-matched reservoir model cannot be unique, because the same production history could be fit by different reservoir scenario. This paper presents a field story where the risk arising from such non-uniqueness is highlighted.
The first history matching on the field was performed to understand its drainage mechanism. Interaction of gravity and viscous forces on vertically heterogeneous system clearly explained the production performance. Based on that finding, an infill program was proposed and was executed. The first well hit the remaining gas, but its productivity was less than expected. Spatial distribution of sand grain size was reviewed and it was presumed that the formation productivity at that location was originally good, but had been impaired over the production period. As the reservoir pressure had declined significantly, excessive permeability reduction due to rock compaction was suspected as the reason, for which no core data was available. The second simulation study was conducted to access that effect. The history was matched with different scenario than the first study and the location for the second infill was proposed, where an existing well happened to be producing from other horizons. Although its core obtained on the target zone suggested very high permeability even after compaction, the second infill turned out to be dry. The productivity impairment proved to be true, but permeability compaction was not enough reasoning for the dry hole. Underestimation of trapped gas saturation under overburden conditions was suggested by its well log, which was consistent with other evidence. The third history match was made and finally concluded that significant amount of gas was left, but as a non-continuous phase and therefore, was practically unrecoverable.
One must be extremely careful in the choice of reservoir parameters to be evaluated in history matching. Preparing more than one reservoir scenario will be preferred.
