Probabilistic Forecasting for Mature Fields With Significant Production History:A Nemba Field Case Study

摘要

The Nemba Field is the third largest oilfield in terms of dailyrnproduction in the Block 0 Concession in the Republic ofrnAngola. Oil production from the field began in January 1996rnthrough an Early Production System (EPS). The field reachedrna peak production rate of approximately 110 MBOPD inrnAugust 2002 and is currently undergoing crestal gas injectionrnat an instantaneous Voidage Replacement Ratio (VRR) ofrnapproximately 0.85. The field is being evaluated for therninstallation of a new compression train to allow for gasrninjection at or above a VRR of 1.0.rnTo support this effort, along with other reservoirrnmanagement activities, a reservoir characterization andrnsimulation study was performed in 2003/2004. The coarse-grid,rnreservoir simulation model generated from this study wasrna six component equation-of-state (EOS) model containingrnapproximately 246,906 total cells (136,675 active) withrnapproximately 33,794 non-neighbor connections to accountrnfor complex faulting and truncation across stratigraphicrnintervals. The workflow for this study consisted of datarnintegration, fine-grid model development, scale-up, coarse-gridrnmodel development, history matching, andrndeterministic/probabilistic forecasting. What distinguishes thernprobabilistic forecasting of this brown-field development fromrnthat of comparable green-field developments is the availablernproduction history and the need to retain the integrity of thernhistory match while fully assessing the impact of uncertaintyrnlevels in key geologic and project variables (uncertainrnparameters).rnThe methodology employed in this study is similar to thernframework proposed by Castellini et. Al.1 , but was adapted tornuse more off-the-shelf technologies. This methodology usesrnboth single variable input modifications (sensitivity analysis) and simultaneous, multi-variable input modificationsrn(experimental design). The sensitivity analysis were used torndetermine the appropriate uncertainty ranges for individualrnuncertain factors while the experimental design was used torndetermine the interactions of these uncertainty factors.rnDuring the experimental design portion of the analysis,rnmodifications were made to the history matched input data andrnthe simulator was run through both the historical andrnprediction periods of the project. In order to retain the integrityrnof the history match, a Quality of History Match (QoHM)rnvariable was defined to quantify and rank any degradation tornthe history match caused by perturbation of uncertain data.rnProxy equations for the simulation response were thenrngenerated for the QoHM variable and the project resultrnvariables: Estimated Ultimate Recovery (EUR) and discountedrncumulative oil production. Finally, a two-step Monte Carlornsimulation approach was used to develop the cumulativerndistribution functions (CDFs) for all alternatives consideredrnfor the gas compressor project.rnIn this two-step Monte Carlo simulation, the uncertainrnfactors were sampled from their individual CDFs for use in arnMonte Carlo trial. The sampled data were first input into thernproxy equation for the QoHM variable to determine the impactrnon the History Match. If the selected input data were found tornmeet predetermined acceptance criteria for the QoHMrnvariable, then they were passed to the second step: input to thernproxy equations for the project result variables. Input datarncombinations that did not meet the acceptance criteria wererndiscarded. This process was repeated for the desired number ofrnMonte Carlo trials to complete the simulation.rnTopics to be discussed in detail in this paper include:rndevelopment of the reservoir simulation model, application ofrnthe proposed methodology for probabilistic forecasting,rndevelopment of the QoHM variable, selection of the QoHMrnacceptance criteria, the two-step Monte Carlo simulationrnapproach, final project results, and lessons learned / bestrnpractices in probabilistic forecasting of brown-fieldrndevelopment projects.