Net sand analysis of thin beds in The Dampier Sub-Basin and Exmouth Plateau of the Northern Carnarvon Basin.
Honours Degree 2009
University of Adelaide
Thin beds, defined as having a thickness between 3 and 60cm (Passey et al., 2006), pose a major problem for the oil and gas industry with relation to formation evaluation. Thinly bedded reservoirs fall below the resolution of many conventional wireline techniques, hence provide only an averaged report of a reservoirs mechanical properties. As a result, errors in estimations of net sand occur.
The overall aim of this thesis was to refine net sand analysis techniques with the focus on providing robust cut-off values in a range of depositional environments and fluid types. The areas of study were the heterolithic reservoirs of the Mungaroo and Brigadier Formations in the Northern Carnarvon Basin. An integrated geological approach has been recommended for this thesis, as a combination of conventional wireline logs has the ability to yield a higher vertical resolution, than is possible using one single method. In this context, integrated geological analyses included: manual logging of core, Core Red-Green-Blue, Static Formation Micro Imager, VShale and fluid column integration.
This study demonstrated that manual logging of core allowed for the most detailed and confident estimates of sandstone volumes, as intervals of core were qualified according to a visible percentage of sandstone. From these results, manual core logging was chosen as the benchmark of comparison for all previously stated analyses.
An investigation into the other techniques of net sand analysis used in this study revealed that CoreRGB produced the highest relative accuracy for thin bed determination. This was shown to be due to the production of high-resolution logs when converting photographs to red, green and blue (RGB) image logs. It was demonstrated that, relative to manual logging of core, estimations of net pay using the Core Red-Green-Blue analysis produced over-estimates averaging 11.50%.
Estimates of sandstone volumes for Static FMI analysis were undertaken both in an un-refined (no fluid column integration) and refined technique (fluid column integration), resulting in over-estimates of net sand by 105.21% and over-estimates in net pay by 21.55% respectively. These results clearly demonstrate that the refined Static FMI analysis was a superior method of measurement as it provided significantly more accurate estimates of thin-bedded net sand, due to the integrated knowledge of fluid types.
Finally, the Vshale technique gave the most variable approximation of all methods tested, with over and under-estimates of net pay averaging 23.48 and -12.49% respectively. It was shown that the observed variability in estimates was a result of external borehole conditions affecting the gamma ray tool.
These results indicate that an integrated geological approach has the ability to refine the assessment of formation evaluation within thin-bedded formations of the Northern Carnarvon Basin, as well as to decrease the range of net-to-gross uncertainties in many field volumetric evaluations