Applying State-of-the-Art Modeling Workflows to Brown Field Assets: Dibi Field, Niger Delta
The conventional reservoir management approach to a brown field has involved history-matching a single static geological model. Many modeling practitioners have used multipliers and/or a one-variable-at-a-time approach, basing uncertainty range size on the ability to maintain the quality of the history match. In the past, Chevron Nigeria Limited (CNL) had based reserves booking on the outcome of deterministic reservoir simulation with mixed results.
To avoid oscillations in booked reserves and ensure consistent P6 to P1 movement, a workflow similar to a green field uncertainty assessment has been employed. This case study highlights the workflow’s application to Dibi, one of CNL’s key fields based on P1 reserves. Methodology includes multiple faulted reservoir grids covering an appropriately wide range of uncertainty for input parameters, utilizing facies-based modeling via Multiple Point Statistics (MPS), combined static and dynamic experimental design, probabilistic history matching, and use of the CLIWorkflow Manager. The result is a range of auditable, easy-to-rebuild, history-matched models, providing a range of reserves. Discrete P10-P50-P90 models can be established to optimize the asset development plan for drilling opportunities, facilities updates, etc.
Dibi was deposited in a tidally-influenced environment, different from most other CNL reservoirs, which are predominantly shoreface. Dibi’s tidal channel bars can have permeability of an order of magnitude higher--as well as water saturations significantly lower--than the shoreface sands into which they downcut. Some of the previous history match attempts required incongruous model parameters yielding questionable forecasts. MPS allows SCAL-derived transforms to be applied to geologically intuitive geometries and hence better control of connectivity in the model. Additionally, the novel “Continuous MPS” technology allows distribution of petrophysical properties within facies with regard to varying geobody orientation.
Probabilistic assessment is necessary to capture the range of possibilities for proper reserves estimates and facilities planning. The procedure automatically numerically screens out models that cannot match using a given set of degrees of freedom. The result is a continuously narrowing range on all input uncertainties as well as the implied degree of correlation between them. Once a reservoir has history, uncertainties should no longer be considered completely independent.
AAPG Search and Discovery Article #90090©2009 AAPG Annual Convention and Exhibition, Denver, Colorado, June 7-10, 2009