Electrical submersible pump design parameters recalibration methods, apparatus, and computer readable medium

1. A method of analyzing and recalibrating parameters for an electrical submersible pump (ESP) application model, the method comprising the steps of:

generating, by a processor, a data set comprising a plurality of sets of ESP application model input parameters and associated theoretical output parameters for an ESP application model substantially matching a corresponding set of measured ESP output parameters defining a plurality of winning data sets or rows; and

performing an averaging analysis, by the processor, on the plurality of winning data sets or rows to obtain a set of ESP application model input parameters usable to recalibrate the ESP application model, wherein the step of generating a data set comprises performing the following steps for each of a plurality of training data sets:

performing a cluster analysis, by the processor, on one of the plurality of training data sets in order to segregate the training data set into a plurality of clusters, each cluster containing a plurality of data rows from the training data set, each data row in the same one of the plurality of clusters having at least similar characteristics;

mapping, by the processor, the set of measured ESP output parameters to one of the plurality of clusters responsive to a similarity analysis performed on the theoretical output parameter within each cluster to identify a winning cluster, the winning cluster being the one cluster of at least several data rows collectively having a set of output parameters most closely matching the corresponding set of measured ESP output parameters; and

identifying, by the processor, a winning data row within the winning cluster that yields outputs parameters which are the most closest to the corresponding set of measured ESP output parameters to define one of the plurality of winning data rows.

2. A method as defined in claim 1 , wherein the cluster analysis comprises a Self-organizing map analysis.

3. A method as defined in claim 2 , wherein each of the plurality of training data sets comprise a plurality of sets of at least substantially random ESP application model input parameters and corresponding theoretical output parameters.

4. A method as defined in claim 3 , further comprising the steps of:

retrieving, by the processor, a set of input parameters and theoretical output parameters for the ESP application model from a data repository for one of a plurality of ESP application models, the theoretical output parameters generated by ESP application modeling software;

retrieving, by the processor, the set of measured ESP output parameters related to the theoretical output parameters of the ESP application model, the measured ESP output parameters retrieved from a SCADA system;

determining, by the processor, differences between the measured ESP output parameters and the theoretical output parameters of the ESP application model generated by the ESP modeling software; and

generating, by the processor, the plurality of training data sets when the differences are found to be outside preselected error tolerances.

5. A method as defined in claim 4 , wherein the step of generating the plurality of training data sets comprises performing the following steps for each separate one of the plurality of training data sets:

generating, by the processor, a random set of ESP application model input parameters;

applying, by the processor, the randomly generated set of input parameters to the ESP modeling software; and

repeating, by the processor, the steps of generating and applying until a preselected number of sets of ESP application model input parameters and corresponding theoretical output parameters forming the respective training data set have been reached.

6. A method as defined in claim 5 , wherein the step of performing an averaging analysis on the plurality of winning data sets or rows results in a final winning data set or row, the method further comprising the step of performing a quality control analysis comprising the steps of:

feeding, by the processor, the ESP application model input parameters in the final winning row defining recalibrated ESP model input parameters to the ESP modeling software to obtain corresponding theoretical output parameters;

performing, by the processor, a comparison of the theoretical output parameters based upon the recalibrated ESP model input parameters to the corresponding set of measured ESP output parameters to verify that the theoretical output parameters are within error tolerance; and

providing, by the processor, an alert when the theoretical output parameters are outside the error tolerance.

7. A method of analyzing and recalibrating parameters for an electrical submersible pump (ESP) application model, the method comprising the steps of:

performing, by a processor, the following for each of a plurality of training data sets to generate a data set comprising a plurality of sets of ESP application model input parameters and associated theoretical output parameters for an ESP application model substantially matching a corresponding set of measured ESP output parameters defining a plurality of winning data sets or rows, each of the plurality of training data sets comprising a plurality of sets of ESP application model input parameters and corresponding theoretical output parameters:

performing, by the processor, a cluster analysis on one of the plurality of training data sets in order to segregate the training data set into a plurality of clusters, each cluster containing a plurality of data subsets or rows from the training data set, each data subset or row in the same one of the plurality of clusters having at least similar characteristics, mapping the set of measured ESP output parameters to one of the plurality of clusters responsive to a similarity analysis performed on the theoretical output parameters within each cluster to identify a winning cluster, the winning cluster being the one cluster of the plurality of clusters collectively having a set of theoretical output parameters most closely matching the corresponding set of measured ESP output parameters, and identifying a data set or row within the winning cluster that yields outputs parameters which are the most closest to the corresponding set of measured ESP output parameters to define one of the plurality of winning data sets or rows; and

performing, by the processor, an averaging analysis on the plurality of winning data sets or rows to obtain a set of ESP application model input parameters usable to recalibrate the ESP application model, wherein the cluster analysis comprises a Self-organizing map analysis.

8. A method as defined in claim 7 , wherein the a plurality of sets of ESP application model input parameters and corresponding theoretical output parameters in each of the plurality of training data sets comprise a plurality of sets of at least substantially random ESP application model input parameters and corresponding theoretical output parameters.

9. Non-transitory computer readable medium having processor readable electrical submersible pump model design analyzing program code embodied on the computer readable medium, the processor readable program code for programming one or more processors to perform operations for analyzing and recalibrating parameters for an electrical submersible pump (ESP) application model, the processor readable program code comprising a set of instructions that when executed by the one or more processors, cause the one or more processors to perform the operations of:

generating a data set comprising a plurality of sets of ESP application model input parameters and associated theoretical output parameters for an ESP application model substantially matching a corresponding set of measured ESP output parameters defining a plurality of winning data sets or rows; and

performing an averaging analysis on the plurality of winning data sets or rows to obtain a set of ESP application model input parameters usable to recalibrate the ESP application model, wherein the operation of generating a data set comprises performing the following operations for each of a plurality of training data sets:

performing a cluster analysis on one of the plurality of training data sets in order to segregate the training data set into a plurality of clusters, each cluster containing a plurality of data rows from the training data set, each data row in the same one of the plurality of clusters having at least similar characteristics;

mapping the set of measured ESP output parameters to one of the plurality of clusters responsive to a similarity analysis performed on the theoretical output parameter within each cluster to identify a winning cluster, the winning cluster being the one cluster of at least several data rows collectively having a set of output parameters most closely matching the corresponding set of measured ESP output parameters; and

identifying a winning data row within the winning cluster that yields outputs parameters which are the most closest to the corresponding set of measured ESP output parameters to define one of the plurality of winning data rows.

10. A non-transitory computer readable medium as defined in claim 9 , wherein the cluster analysis comprises a Self-organizing map analysis.

11. A non-transitory computer readable medium as defined in claim 10 , wherein each of the plurality of training data sets comprise a plurality of sets of at least substantially random ESP application model input parameters and corresponding theoretical output parameters.

12. A non-transitory computer readable medium as defined in claim 11 , wherein the operations further comprise:

receiving a set of input parameters and theoretical output parameters for the ESP application model from a data repository for one of a plurality of ESP application models, the theoretical output parameters generated by ESP application modeling software;

receiving the set of measured ESP output parameters related to the theoretical output parameters of the ESP application model, the measured ESP output parameters retrieved from a SCADA system;

determining differences between the measured ESP output parameters and the theoretical output parameters of the ESP application model generated by the ESP modeling software; and

generating the plurality of training data sets when the differences are found to be outside preselected error tolerances.

13. A non-transitory computer readable medium as defined in claim 12 , wherein the operation of generating the plurality of training data sets comprises performing the following operations for each separate one of the plurality of training data sets:

generating a random set of ESP application model input parameters;

applying the randomly generated set of input parameters to the ESP modeling software; and

repeating the steps of generating and applying until a preselected number of sets of ESP application model input parameters and corresponding theoretical output parameters forming the respective training data set have been reached.

14. An apparatus for analyzing and recalibrating parameters for an electrical submersible pump (ESP) application model, the apparatus comprising:

a ESP model design parameter analyzing computer including a processor and memory coupled to the processor; and

an ESP model design parameter analyzing program configured to analyze and recalibrate parameters for an ESP model, the program stored in the memory of the ESP model design parameter analyzing computer and including instructions that when executed by the ESP model design parameter analyzing computer, cause the computer to perform the operations of:

generating a data set comprising a plurality of sets of ESP application model input parameters and associated theoretical output parameters for an ESP application model substantially matching a corresponding set of measured ESP output parameters defining a plurality of winning data sets or rows, and performing an averaging analysis on the plurality of winning data sets or rows to obtain a set of ESP application model input parameters usable to recalibrate the ESP application model, wherein the operation of generating a data set comprises performing the following operations for each of a plurality of training data sets:

performing a cluster analysis on one of the plurality of training data sets in order to segregate the training data set into a plurality of clusters, each cluster containing a plurality of data rows from the training data set, each data row in the same one of the plurality of clusters having at least similar characteristics;

mapping the set of measured ESP output parameters to one of the plurality of clusters responsive to a similarity analysis performed on the theoretical output parameter within each cluster to identify a winning cluster, the winning cluster being the one cluster of at least several data rows collectively having a set of output parameters most closely matching the corresponding set of measured ESP output parameters; and

identifying a winning data row within the winning cluster that yields outputs parameters which are the most closest to the corresponding set of measured ESP output parameters to define one of the plurality of winning data rows.

15. An apparatus as defined in claim 14 , wherein the cluster analysis comprises a Self-organizing map analysis.

16. An apparatus as defined in claim 15 , wherein each of the plurality of training data sets comprise a plurality of sets of at least substantially random ESP application model input parameters and corresponding theoretical output parameters.

17. An apparatus as defined in claim 16 , wherein the operations further comprise:

receiving a set of input parameters and theoretical output parameters for the ESP application model from a data repository for one of a plurality of ESP application models, the theoretical output parameters generated by ESP application modeling software;

receiving the set of measured ESP output parameters related to the theoretical output parameters of the ESP application model, the measured ESP output parameters retrieved from a SCADA system;

determining differences between the measured ESP output parameters and the theoretical output parameters of the ESP application model generated by the ESP modeling software; and

generating the plurality of training data sets when the differences are found to be outside preselected error tolerances.

18. An apparatus as defined in claim 17 , wherein the operation of generating the plurality of training data sets comprises performing the following operations for each separate one of the plurality of training data sets:

generating a random set of ESP application model input parameters;

applying the randomly generated set of input parameters to the ESP modeling software; and

repeating the steps of generating and applying until a preselected number of sets of ESP application model input parameters and corresponding theoretical output parameters forming the respective training data set have been reached.