Process to analyze mwd surveys from different bha runs in the same wellbore

Abstract

A process to analyze MWD surveys from different BHA runs in the same well bore in a combined fashion is presented herein. A processor is used to receive and analyze data provided through the generation of one or more BHA runs of a survey tool. The processor regulates the interaction of the various components within a control unit and facilitates calculation and analysis of the data produced to identify faults and their respective causes. For example, control unit may obtain Bayesian estimates and calculate an estimate of the mean survey error and a posteriori covariance corresponding to a full list of survey errors as multiple runs are completed. The process utilizes data from the entire wellbore to improve the observability of each error source. Consequently, fault detection, isolation, and recovery are much more robust.

What is claimed is: 
     
         1 . A method of drilling a borehole including a processor and communications system, comprising:
 obtaining vector measurements from a first survey tool of magnetic and gravitational fields at multiple survey stations in a borehole;   obtaining reference values describing magnetic and gravitational field components from at least any of a variety of reference field models source or an in-field data source;   obtaining specifications describing a priori accuracy of the first survey tool and reference field parameters;   defining a list of survey errors;   defining subsets of survey errors from a list that partitions a probability space of all combinations of listed survey errors;   calculating an estimate of a mean survey error and a posteriori covariance corresponding to the list of survey errors for each subset using the processor, the estimate being calculated using the survey tool measurements, expected reference field values, and only the a priori accuracy specifications corresponding to errors not included in the subsets;   obtaining a Bayesian estimate of the mean survey error and the a posteriori covariance of the list of survey errors from the set of statistical estimates obtained in the previous step, and a set of specified a priori probabilities for each partition of the probability space;   obtaining vector measurements from one or more additional survey tools of the magnetic and gravitational fields at multiple survey stations in a borehole;   obtaining a Bayesian estimate of the mean and a posteriori covariance the survey errors corresponding to each of these additional runs of survey tools into the borehole;   obtaining a combined statistical estimate of the survey errors corresponding to all survey tools run in a borehole by utilizing the Bayesian estimates from individual runs as correlated measurements; and   altering a drilling parameter using at least one of estimated deviations and a residual error covariance matrix.   
     
     
         2 . The method of  claim 1 , wherein the Bayesian estimates that are combined include survey runs from multiple boreholes originating from at least one of the same well pad and the same field. 
     
     
         3 . The method of  claim 1 , further comprising:
 partitioning the probability space encompassing all combinations of measurement and reference field faults for each survey run in the wellbore.   
     
     
         4 . The method of  claim 1 , further comprising:
 performing multiple runs of the first survey tool.   
     
     
         5 . The method of  claim 1 , further comprising:
 statistically combining the Bayesian run estimates for each run of the survey tool to form a combined estimate of the reference field component offsets and the sensor errors for each survey run.