The application of Mathematical Programming methods to oil and gas field development planning

This paper describes some applications of Mathematical Programming methods to the problem of planning the long term exploitation of oil and gas resources.Several previously reported oil and gas planning models are reviewed and a newly developed model is described. The earlier models are characterised by the use of implicit representations of reservoir behaviour, using suitable variables and constraints within the problem matrix. This approach gives the greatest freedom to find the most effective planning solution and probably requires the least data, but is often limited in application by the degree of realism that can be achieved in representing reservoir behaviour within available computational resources.The new model was developed to overcome these limitations. It is formulated as a Mixed Integer Programming model using Special Ordered Sets to identify alternative reservoir production profiles. This permits the description of reservoir production to be as good (or as bad!) as the available data, but limits the variation of investment and production between fields to the range of explicitly defined production profile choices. Possible extensions to the model to allow development profiles to be varied within the model, so reducing the number of explicitly defined profiles needed, are discussed and illustrated.