Automatically selecting analogous members for new population members based on incomplete descriptions, including an uncertainty characterzing selection

1. A population comparison method comprising:

storing a list of population members and corresponding member characteristics parameters describing each listed member, wherein the population members are hydrocarbon reservoirs;

receiving a description for a new population member, said description missing one or more member characteristics parameters, wherein said at least one missing member characteristics parameters is a key parameter (KP);

automatically estimating a value for at least one missing member characteristic parameters responsive to stored said member characteristics parameters;

supplementing said description with each estimated value;

automatically comparing the supplemented description against stored descriptions for each listed member;

selecting a subset of listed members as analogous members for said new population member responsive to the comparison; and

extracting an uncertainty characterization of said selected subset automatically responsive to selection.

2. A population comparison method as in claim 1 , wherein a plurality of said KPs are identified as controlling KPs (CKPs), and one or more said at least one missing member characteristics parameters is a CKP.

3. A population comparison method as in claim 2 , wherein automatically estimating comprises:

generating a model from said list of population members references only to CKPs with known values in the new member data, missing said values for said KPs not identified as CKPs remaining unknown; and

applying said generated model to said new member data to estimate missing CKP values for said new member.

4. A population comparison method as in claim 3 , wherein extracting said uncertainty characterization comprises:

extracting a univariate probability distribution for each parameter from analogous members data; and

extracting a joint probability distribution of analogous members parameters.

5. A population comparison method as in claim 4 , before storing said list, said method further comprising pre-configuring said list, pre-configuring comprising:

pre-processing said raw member data, said raw member data including an entry for said each listed member, each said entry including said corresponding member characteristics parameters;

eliminating outlier members;

transforming, normalizing and standardizing listed member characteristics parameters; and

storing said list of population members.

6. A population comparison method as in claim 5 , wherein pre-processing selectively replaces hierarchical parameters, chronological parameters and ranking parameters, and after transforming, normalizing and standardizing, said method further comprises:

selecting a member as a target, one or more other members being previously identified as analogous members for said target; and

selecting member KPs as controlling KPs (CKPs).

7. A population comparison method as in claim 3 , wherein said model is a regression model.

8. A population comparison method as in claim 3 , wherein said model is a classification model.

9. A population comparison method as in claim 1 , wherein said population comprises hydrocarbon reservoirs, population members being known hydrocarbon reservoirs, and said new population member being a new hydrocarbon reservoir, said method further comprising automatically estimating further comprises estimating values for remaining unknown KPs responsive to said selected subset.

10. A reservoir valuation method comprising:

configuring known reservoir data including an entry for each known reservoir and corresponding reservoir characteristics parameters;

selecting controlling features from stored said reservoir key parameters (KPs);

storing the configured said known reservoir data in a refined list of known reservoirs;

receiving a description for a new reservoir, said description missing values for one or more KPs;

automatically estimating a value for at least one missing value responsive to said stored refined list;

supplementing the new reservoir description with each estimated value;

automatically comparing the supplemented new reservoir description against each listed reservoir;

selecting a subset of listed reservoirs members as analogous reservoirs responsive to the comparison; and

extracting an uncertainty characterization of said selected subset automatically responsive to selection.

11. A reservoir valuation method as in claim 10 , wherein configuring known reservoir data comprises:

selectively replaces hierarchical parameters, chronological parameters and ranking parameters;

selecting a known reservoir as a target, one or more other known reservoirs being previously identified as analogous reservoirs for said target; and

selecting KPs as controlling KPs (CKPs) eliminating outlier known reservoirs;

transforming, normalizing and standardizing listed known KPs; and

storing said refined list of known reservoirs.

12. A reservoir valuation method as in claim 10 , wherein one or more said at least one KP missing a value is a CKP, and automatically estimating comprises:

generating a model from said list of population members references only to CKPs with known values in the new member data, missing said values for said KPs not identified as CKPs remaining unknown; and

applying said generated model to said new member data to estimate missing CKP values for said new member, estimated values supplementing said description, missing

new reservoir KPs not identified as CKPs being automatically estimated responsive to said selected subset.

13. A reservoir valuation method as in claim 12 , wherein said model is a regression model and extracting said uncertainty characterization comprises:

extracting a univariate probability distribution for each parameter from analogous members data; and

extracting a joint probability distribution of analogous members parameters.

14. A reservoir valuation method as in claim 12 , wherein said model is a classification model and extracting said uncertainty characterization comprises:

extracting a univariate probability distribution for each parameter from analogous members data; and

extracting a joint probability distribution of analogous members parameters.

15. A computer program product for comparing members of a population, said computer program product comprising a computer usable medium having computer readable program code stored thereon, said computer readable program code causing one or more computer executing said code to:

store a list of population members and corresponding member characteristics parameters describing each listed member;

receive a description for a new population member, said description missing one or more member characteristics parameters;

automatically estimate a value for at least one missing member characteristic parameters responsive to stored said member characteristics parameters;

supplement said description with each estimated value;

automatically compare the supplemented description against stored descriptions for each listed member;

select a subset of listed members as analogous members for said new population member responsive to the comparison; and

extract an uncertainty characterization of said selected subset automatically responsive to selection.

16. A computer program product for comparing members of a population as in claim 15 , wherein a plurality of said member characteristic parameters are key parameters (KPs), a plurality of said KPs are identified as controlling KPs (CKPs), and one or more said at least one missing member characteristics parameters is a CKP.

17. A computer program product for comparing members of a population as in claim 15 , wherein causing said one or more computer executing said code to automatically estimate said value causes said one or more computer to:

generate a model from said list of population members references only to CKPs with known values in the new member data, missing said values for said KPs not identified as CKPs remaining unknown; and

apply said generated model to said new member data to estimate missing CKP values for said new member.

18. A computer program product for comparing members of a population as in claim 17 , wherein causing said one or more computer executing said code to extract said uncertainty characterization causes said one or more computer to:

extract a univariate probability distribution for each parameter from analogous members data; and

extract a joint probability distribution of analogous members parameters.

19. A computer program product for comparing members of a population as in claim 18 , wherein said population comprises hydrocarbon reservoirs, population members being known hydrocarbon reservoirs, and said new population member being a new hydrocarbon reservoir before storing said list, said computer readable program code further causing said one or more computer executing said code to:

pre-process raw known reservoirs data, said raw known reservoirs data including an entry for said each listed reservoir, each said entry including said corresponding reservoir characteristics parameters;

eliminate outlier reservoirs;

transform, normalize and standardize listed reservoir characteristics parameters;

and storing said list of known reservoirs.

20. A computer program product for comparing members of a population as in claim 18 , wherein pre-processing selectively replaces hierarchical parameters, chronological parameters and ranking parameters, and after transforming, normalizing and standardizing, said one or more computer further being caused to:

receive selection of a known reservoir as a target, one or more other known reservoirs being previously identified as analogous reservoirs for said target;

receive selection of KPs as a CKPs;

receive weights weighting said KPs in a similarity function to compare said target to other known reservoirs, and interactively adjust the weights and similarity function to select said previously identified analogous reservoirs;

return to receive selection of another known reservoir as a target until all known reservoirs have been selected; and

store said weights and similarity function, said similarity function being applied to said description, said weights being applied to said new member CKPs to estimate missing new reservoir CKPs.

21. A computer program product for valuating reservoirs, said computer program product comprising a non-transitory computer usable medium having computer readable program code stored thereon, said computer readable program code causing a computer executing said code to:

configure known reservoir data including an entry for each known reservoir and corresponding reservoir key parameters (KPs);

select controlling features from stored said KPs to identify controlling KPs (CKPs);

store configured said known reservoir data in a refined list of known reservoirs;

receive a description for a new reservoir, said description missing values for one or more said KPs;

automatically estimate a value for at least one missing value responsive to said stored refined list;

supplement the new reservoir description with each estimated value;

automatically compare the supplemented new reservoir description against each listed reservoir;

select a subset of listed reservoirs members as analogous reservoirs responsive to the comparison; and

extract an uncertainty characterization of said selected subset automatically responsive to selection.

22. A computer program product for comparing members of a population as in claim 21 , wherein:

causing said one or more computer executing said code to configure known reservoir data causes said one or more computer to:

pre-process raw known reservoir data, eliminate outlier known reservoirs, transform, normalize and standardize KPs for listed known reservoirs, and

store said list of known reservoirs; and

causing said one or more computer executing said code to select controlling features causes said one or more computer to:

receive selection of a known reservoir as a target, one or more other known reservoirs being previously identified as analogous reservoirs for said target, receive selection of KPs as a CKPs, receive weights weighting said KPs in a similarity function to compare said target to other known reservoirs, the weights and similarity function being adjusted to select said previously identified analogous reservoirs, return to receive selection of another known reservoir as a target until all known reservoirs have been selected; and

store said weights and similarity function, said similarity function being applied to said description, said weights being applied to said new reservoir characteristics parameters to estimate missing new reservoir CKP values.

23. A computer program product for comparing members of a population as in claim 22 , wherein pre-processing selectively replaces hierarchical parameters, chronological parameters and ranking parameters, one or more said at least one missing new reservoir characteristics parameters is a key parameter, and causing said one or more computer executing said code to automatically estimate causes said one or more computer to generate a model from said list of population members references only to CKPs with known values in the new member data; and

apply said generated model to said new member data to estimate missing CKP values for said new member, missing new reservoir KP values not identified as CKPs being automatically estimated responsive to said selected subset.

24. A computer program product for comparing members of a population as in claim 23 , wherein causing said one or more computer executing said code to extract an uncertainty characterization causes said one or more computer to extract a univariate probability distribution for each parameter from analogous members data; and extract a joint probability distribution of analogous members parameters.