Recent studies of field growth worldwide indicate that, over the latest two decades, reserve addition from the existing fields has added about similar quantity of oil reserves has the contribution from new discoveries. A large portion of that reserve growth comes from the improved recovery. Thus future reserve growth should be included in the estimated ultimate resource (EUR) in a resource assessment. On the other hand, volumetric calculation of resource potential involves different levels of variable dependencies from geological risk evaluation, volumetric calculation to resource integration at a basin or higher scales. The variable correlations are so common among the geological variables that ignoring the interdependencies may lead to a biased resource estimation and result in an under-estimated uncertainty range. However, obtaining sufficient data to define the correlation is often a challenge, particularly in a frontier region. Other difficulties of implementing the variable dependencies include the formulation of the variable correlation when the volumetric variables do not follow a multi-normal or lognormal distribution as well as using analogy. Recent methodology development in financial risk modeling indicates that the use of copulas allows more flexibility in incorporating variable dependency using analogy and can easily handle non-standard normal or lognormal distribution, providing a natural way to study and measure dependency between random variables. This paper illustrates the use of copula for implementing variable dependency through example from the oil resource assessment of the established plays in BMB. The field reserve growth from improved recovery in the assessment is handled as a variable correlation problem based on world wide statistics of positive correlation between in place volume and ultimate recovery factor. Comparisons of oil resource estimates from scenarios of positive correlation and independence in the BMB suggest that when positive correlations among the variables are implemented in the assessment, the mean value of basin oil EUR is almost about 1.5 times that of the EUR estimated assuming complete independence as well as giving an EUR with a much greater uncertainty range.