Methods and systems for predicting joint networks in subsurface layers

Abstract

Methods and systems for synthetic joint network prediction are disclosed. The method may include obtaining a plurality of outcrop pavement images of a plurality of joint networks and determining, using a first machine learning (ML) network, a plurality of detected joint networks using the plurality of outcrop pavement images. The method further includes determining, for each of the plurality of detected joint networks, a set of surface geostatistical properties and creating a database of surface joint properties including the plurality of sets of surface geostatistical properties. The method still further includes generating a synthetic joint network predictor by training a second ML network, using the database of surface joint properties, to produce a synthetic joint network. In addition, the method includes using the synthetic joint network predictor to predict a predicted joint network from a geostatistical description of observed subsurface joints.

What is claimed is: 
     
         1 . A method, comprising:
 obtaining a plurality of outcrop pavement images of a plurality of joint networks;   determining, using a first machine learning (ML) network, a plurality of detected joint networks using the plurality of outcrop pavement images;   determining, for each of the plurality of detected joint networks, a set of surface geostatistical properties;   creating a database of surface joint properties comprising the plurality of sets of surface geostatistical properties;   generating a synthetic joint network predictor by training a second ML network, using the database of surface joint properties, to produce a synthetic joint network; and   using the synthetic joint network predictor to predict a predicted joint network from a geostatistical description of observed subsurface joints.   
     
     
         2 . The method of  claim 1 , further comprising:
 obtaining subsurface data from a subsurface region having a joint network and comprising a wellbore;   determining a set of subsurface geostatistical properties of the joint network from the subsurface data;   determining a target subsurface layer from the set of subsurface geostatistical properties;   wherein using the synthetic joint network, the predicted joint network is predicted in the target subsurface layer according to a subset of target subsurface geostatistical properties selected from the set of subsurface geostatistical properties.   
     
     
         3 . The method of  claim 1 , wherein the plurality of outcrop pavement images comprises outcrop pavement images from a plurality of different geological facies. 
     
     
         4 . The method of  claim 1 , wherein the set of surface geostatistical properties comprises joint orientation. 
     
     
         5 . The method of  claim 1 , wherein creating the database of surface joint properties comprises correcting for structural discontinuities that are not joints. 
     
     
         6 . The method of  claim 1 :
 wherein the second ML network is a generative adversarial network comprising a generator and a discriminator;   wherein the synthetic joint network predictor is derived from the generator.   
     
     
         7 . A method, comprising:
 obtaining subsurface data from a subsurface region having a joint network and comprising a wellbore;   determining a set of subsurface geostatistical properties of the joint network from the subsurface data;   determining a target subsurface layer from the set of subsurface geostatistical properties; and   using a synthetic joint network predictor, predicting a predicted joint network in the target subsurface layer, based, at least in part, on a subset of target subsurface geostatistical properties selected from the set of subsurface geostatistical properties.   
     
     
         8 . The method of  claim 7 , wherein the synthetic joint network predictor is generated by a second machine learning (ML) network trained with a database of surface joint properties. 
     
     
         9 . The method of  claim 8 , further comprising:
 obtaining a plurality of outcrop pavement images of a plurality of joint networks;   determining, using a first ML network, a plurality of detected joint networks using the plurality of outcrop pavement images;   determining, for each of the plurality of detected joint networks, a set of surface geostatistical properties; and   creating the database of surface joint properties comprising the plurality of sets of surface geostatistical properties.   
     
     
         10 . The method of  claim 7 , further comprising:
 updating, using a well planning system, a portion of a planned well in the target subsurface layer based, at least in part, on the predicted joint network; and   drilling, using a drilling system, a well guided by the planned well.   
     
     
         11 . The method of  claim 7 , wherein determining the target subsurface layer further comprises determining, at least, two-dimensional boundaries of the target subsurface layer. 
     
     
         12 . The method of  claim 7 , wherein predicting a predicted joint network in the target subsurface layer further comprises correlating the set of target subsurface geostatistical properties and a set of surface geostatistical properties to specify an input for the joint network predictor. 
     
     
         13 . The method of  claim 7 , further comprising determining reference analogues between the predicted joint network in the target subsurface layer and one or more related surface pavements. 
     
     
         14 . A system, comprising:
 a plurality of devices configured to be disposed along a wellbore and configured to obtain subsurface data;   a computer configured to:
 receive the subsurface data, 
 determine a set of subsurface geostatistical properties of a subsurface joint network from the subsurface data, and 
 determine a target subsurface layer from the set of subsurface geostatistical properties; and 
   a synthetic joint network predictor configured to predict a predicted joint network in the target subsurface layer based, at least in part, on a subset of target subsurface geostatistical properties selected from the set of subsurface geostatistical properties.   
     
     
         15 . The system of  claim 14  wherein the synthetic joint network predictor is generated by a second machine learning (ML) network trained with a database of surface joint properties. 
     
     
         16 . The system of  claim 15 , further comprising:
 a first ML network configured to:
 receive a plurality of outcrop pavement images comprising a plurality of joint networks, and 
 determine a plurality of detected joint networks using the plurality of outcrop pavement images, 
   wherein the computer is further configured to:
 determine, for each of the plurality of detected joint networks, a set of surface geostatistical properties, and 
 create the database of surface joint properties comprising the plurality of sets of surface geostatistical properties. 
   
     
     
         17 . The system of  claim 14 , further comprising a well planning system configured to update a portion of a planned well in the target subsurface layer based, at least in part, on the predicted joint network. 
     
     
         18 . The system of  claim 17 , further comprising a drilling system configured to drill a well guided by the planned well. 
     
     
         19 . The system of  claim 14 , wherein the joint network predictor is configured to determine reference analogues between the predicted joint network in the target subsurface layer and one or more related surface pavements. 
     
     
         20 . The system of  claim 14 , wherein the subsurface data comprises wellbore images.