Abstract

The gas fields developed by TOTAL in the Mahakam Delta consist of fluvio-deltaic, multi layered reservoirs of individual sands with variable extensions. No current commercial model is available to properly manage such reservoirs under the different constraints and development scenarios. It is obvious that a 3 dimensional gridded model is too heavy to use to predict the reservoir performances and to optimize development plans for different deliverability scenarios.

To overcome this problem, TOTAL has developed inhouse, a specific material balance based tank model, PREDIC. This model uses in addition to the Material Balance Equation:

• Deliverability Equation Q = C (P_r²-P²_BHFP)ⁿ or ΔP² = AQ_g + BQ_g²

• Pressure Loss Equation in Pipes of Cullender and Smith

The main particularities and features of PREDIC are described in this paper:

• Management of crossflow between reservoirs in the same well

• Distribution of the produced wellhead gas in a given surface network (which may include HP, MP and LP sub networks) vs limiting effect of rates

• Option of aquifer activity incorporated into model

• Condensate production prediction (vs decline law for condensate yield)

• Management of both deliverability targets and required swing and potentials

• Possibility of history matching -ie. adjustment of pressures (wellhead, reservoir) for a given production history.

• User friendly in its inputs and outputs

Such a model is necessary to make realistic predictions. It is now the basic tool used for reserves estimates and the definition of an optimum development plan to meet a given production profile. In addition, it enables the company to react to the changes in the gas demand as PREDIC is also a field/well management tool. These points are illustrated in this paper by the application of the model to the giant Mahakam gas field, TUNU.