The final objective is to populate all the grid blocks of interest with fluid compositions that honour the observed aerial and vertical variations in the fluid compositions and properties in the appropriate reservoirs.

The lumped EOS should only be based on the qualified historical data and that collected from the most recent wells. Starting with the detailed models from the QC processes, the number of components is gradually reduced in a step wise fashion. At each stage, tuning is minimised and optimised such that the resultant EOS retains its physical integrity and accurately predicts the qualified PVT properties.

The lumped model is then validated using the reported and qualified historical data. Where appropriate, predictions of EOR processes and ashaltene and wax deposition are also validated.

Based upon our previous experiences, we are able to advise on the best methods to populate the grid block of the compositional reservoir simulation model honouring the distribution of fluid compositions.

The simplest case is where there are no aerial or vertical changes in fluid compositions. However, in our experience, aerial variations are often the case and whereas there may be changes with depth, these are often not uniform and are due to the charging history of the formations.