| Popis: |
Improving the accuracy of petrophysical analyses is critical in optimizing formation evaluation. Petrophysical accuracy is affected by errors inherent in log and core measurements and reservoir heterogeneity. These errors can lead to imprecise differentiation of pay and non-pay resulting in either overlooked pay or additional testing costs. Error reduction will improve the differentiation of pay/non-pay but, due to remaining errors, the distinction will still be imprecise. Thus to optimize formation evaluation, it is imperative to quantify remaining errors. This paper presents a newly developed technology that quantifies petrophysical uncertainty to maximize the utility of petrophysics in formation evaluation by incorporating Monte Carlo numerical analysis to account for uncertainties in log, reservoir and petrophysical parameters. In comparison to traditional petrophysics, this technique does not assume unique values for the input parameters used in calculation of hydrocarbon pore volumes and deliverability. This technique incorporates random errors in logging tool responses and core measurements as well as reservoir heterogeneity in a Monte Carlo simulation. Output from the Monte Carlo simulation is then used in appropriate petrophysical models (e.g. Archie, Dual Water, etc.) to calculate statistical distributions for saturation and permeability to define productive and nonproductive zones. The novel aspect of this methodology is that it also identifies zones where the random measurement and model errors are too large to determine accurate saturation and permeability. In these zones of high uncertainty, saturation and permeability are "undetermined". Based on the amount of uncertainty, this technique can determine whether additional logging, coring or well test data are needed to reclassify "undetermined" zones as either productive or non-productive. This paper will outline the procedures used in developing the technique and will show its utility in several examples from Middle Eastern reservoirs. |
