Method for forecasting the production of a petroleum reservoir utilizing genetic programming

1. A computerized method of forecasting production profiles for petroleum reservoirs comprising the steps of:

(a) utilizing genetic programming to generate a historical proxy and a forecasting proxy;

(b) qualifying each of a plurality of reservoir models with the historical proxy as either a set of acceptable models or a set of unacceptable models, each model of the plurality of reservoir models having different parameter values; and

(c) applying the forecasting proxy to the set of acceptable reservoir models to produce a range of production forecasts for the set of acceptable reservoir models.

2. The method of claim 1 which includes producing the plurality of reservoir models utilizing a uniform sampling methodology that generates a sampling distribution of a parameter space for a predetermined number of runs.

3. The method of claim 1 which further includes identifying common characteristics of the reservoir models in the set of acceptable reservoir models.

4. The method of claim 1 which further includes identifying common characteristics of the reservoir models which were not included in the set of acceptable reservoir models.

5. A computerized method of forecasting production profiles for petroleum reservoirs comprising the steps of:

(a) producing an initial sample of reservoir models;

(b) generating a historical production profile and a future production profile for each of the initial sample of reservoir models;

(c) qualifying each of the initial sample of reservoir models as either acceptable or unacceptable with respect to the historical production profiles to produce a historical set of qualifications;

(d) inputting the historical set of qualifications into genetic programming to generate a historical proxy;

(e) applying the historical proxy to a plurality of reservoir models, the plurality of reservoir models being larger than the initial sample of reservoir models, and qualifying each model in the plurality of reservoir models as either acceptable or unacceptable to identify a set of acceptable reservoir models;

(f) quantifying each of the initial sample of reservoir models with respect to the future production profiles to produce forecasting characterizations;

(g) inputting the forecasting characterizations into genetic programming to generate a forecasting proxy; and

(h) applying the forecasting proxy to the set of acceptable reservoir models to produce a range of production forecasts for the set of acceptable reservoir models.

6. The method of claim 5 , wherein the initial sample of reservoir models is produced by a uniform sampling methodology that generates a sampling distribution of a parameter space for a predetermined number of runs.

7. The method of claim 1 , wherein the initial sample of reservoir models is a subset of the plurality of reservoir models.

8. The method of claim 1 which further includes classifying the reservoir models in the set of acceptable reservoir models.

9. The method of claim 1 which further includes identifying common characteristics of the reservoir models in the set of acceptable reservoir models.

10. The method of claim 1 which further includes identifying common characteristics of the reservoir models which were not included in the set of acceptable reservoir models.

11. The method of claim 1 which further includes performing an outlier study utilizing the genetic programming on the initial sample of reservoir models.

12. A system for forecasting production profiles for petroleum reservoirs, including:

a historical proxy generated by a genetic program, the historical proxy qualifying a plurality of reservoir models as either acceptable or unacceptable; and

a forecasting proxy generated by the genetic program, the forecasting proxy being applied to the acceptable reservoir models to produce a range of production forecasts.