Well Integrity Evaluation Using Pseudo Attributes

Abstract

A method and system for identifying corrosion. The method may include disposing an electromagnetic (EM) logging tool into a pipe string configured to perform measurements at one or more depths and creating a log from the measurements at one or more depths taken by the EM logging tool in the pipe string. The method may further comprise computing one or more pseudo attributes of one or more pipes with the log and determining a remedial operation based at least on the one or more pseudo attributes. The system may include an electromagnetic (EM) logging tool into a pipe string configured to perform measurements at one or more depths and an information handling system.

What is claimed is: 
     
         1 . A method comprising:
 disposing an electromagnetic (EM) logging tool into a pipe string configured to perform measurements at one or more depths;   creating a log from the measurements at one or more depths taken by the EM logging tool in the pipe string;   computing one or more pseudo attributes of one or more pipes with the log; and   determining a remedial operation based at least on the one or more pseudo attributes.   
     
     
         2 . The method of  claim 1 , wherein the pseudo attributes of the one or more pipes are pipe thickness, metal loss, magnetic permeability, or electrical permeability of the one or more pipes. 
     
     
         3 . The method of  claim 1 , wherein computing the one or more pseudo attributes comprises building a regression model to estimate at least one pseudo attribute from the one or more pseudo attributes. 
     
     
         4 . The method of  claim 1 , wherein computing the one or more pseudo attributes comprises training a machine learning model to estimate at least one pseudo attribute from the one or more pseudo attributes. 
     
     
         5 . The method of  claim 1 , further comprising generating a first database with at least one or more pipe attributes. 
     
     
         6 . The method of  claim 5 , further comprising computing one or more model-based attributes with a model-based inversion. 
     
     
         7 . The method of  claim 6 , wherein one or more pseudo attributes are used to extract prior knowledge on a direction of metal loss progression. 
     
     
         8 . The method of  claim 7 , wherein the prior knowledge and the one or more pseudo attributes are used to determine one or more regularization parameters. 
     
     
         9 . The method of  claim 8 , wherein the model-based inversion comprises an optimization algorithm of one or more regularization parameters of the model-based inversion. 
     
     
         10 . The method of  claim 6 , wherein the model-based inversion comprises a cost function using gradient descent methods or a brute-force search of the database. 
     
     
         11 . The method of  claim 10 , wherein the cost function is defined in terms of the one or more pseudo attributes. 
     
     
         12 . The method of  claim 6 , further comprising performing a quality control on the one or more model-based attributes, wherein the quality of the one or more model-based attributes at a measurement depth is determined based on the one or more pseudo attributes at the measurement depth. 
     
     
         13 . The method of  claim 12 , wherein a Monte Carlo algorithm is used to estimate the one or more pseudo attributes or the one or more model-based attributes for different regularization parameters of measurement weights or random noise. 
     
     
         14 . The method of  claim 6 , further comprising comparing the model-based attributes to one or more pseudo attributes to form a comparison. 
     
     
         15 . The method of  claim 14 , wherein the comparison comprises a cross-covariance between individual pipe thicknesses or metal loss, misfit of a cost function, difference between model-based attributes and pseudo attributes. 
     
     
         16 . The method of  claim 15 , wherein acceptable ranges for the comparison are determined through statistical analysis. 
     
     
         17 . The method of  claim 16 , further comprising adjusting regularization parameters of the model-based inversion if the comparison is outside of an acceptable range. 
     
     
         18 . A system comprising:
 an electromagnetic (EM) logging tool into a pipe string configured to perform measurements at one or more depths; and   an information handling system configured to:
 create a log from the measurements at one or more depths taken by the EM logging tool in the pipe string; 
 compute one or more pseudo attributes of one or more pipes with the log; and 
 determine a remedial operation based at least on the one or more pseudo attributes. 
   
     
     
         19 . The system of  claim 18 , wherein the EM logging tool operates in time-domain or frequency domain. 
     
     
         20 . The system of  claim 18 , wherein the EM logging tool comprises at least one transmitter coil and at least one receiver coil.