System and method for automated detection of fracture driven interactions

Abstract

A method is described to detect, analyze, and characterize fracture driven interaction (FDI) events in unconventional resources using production data from a well and its nearest neighbors. The method provides a rigorous statistical analysis of the production data of a well and its nearest neighbors, and utilizes a combination of signals including pressure, rate, water-oil ratio (WOR), and fluid production to identify and characterize FDI events. The method is executed by a computer system.

What is claimed is: 
     
         1 . A computer-implemented method, comprising:
 a. receiving well completion and production data for a plurality of wells;   b. identifying parent-child well pairs in the plurality of wells;   c. generating event labels based on dynamic pressure exceptions, fluid ratio behavior analysis, and anomaly detection for the parent-child well pairs;   d. training a model to identify and characterize fracture driven interactions using the event labels and the dynamic pressure exceptions and fluid ratio behavior;   e. receiving subsurface reservoir property information including at least one of Shmin direction, compass angle, Pclay, Psand, vclay, vsand, distance from child stage and parent stage to natural fractures, and frac conductivity; and   f training the model using the subsurface reservoir property.   
     
     
         2 . The method of  claim 1  further comprising validating the model using blind event data validation and hyperparameter optimization, by segmenting the well completion and production data and the subsurface reservoir property information into a training portion and a testing portion. 
     
     
         3 . The method of  claim 1  further comprising applying the model to a second set of well completion and production data to identify and characterize fracture driven interactions. 
     
     
         4 . The method of  claim 1  wherein the identifying parent-child well pairs is based on proximity and timing relevance. 
     
     
         5 . The method of  claim 1  wherein the dynamic pressure exceptions and fluid ratio behavior analysis consider pressure and water-oil ratio (WOR) triggers as well as timing relationship to a nearest fracturing stage event. 
     
     
         6 . The method of  claim 1  wherein the model may be one of a neural network, decision tree, random forest, XGBoost, or other machine learning model, and has an architecture including multiple layers, with each layer consisting of a set of nodes or neurons. 
     
     
         7 . The method of  claim 6  wherein an input layer receives features extracted from the well completion and production data including at least one of pressure, rate, water-oil ratio (WOR), fluid production, and subsurface reservoir property information. 
     
     
         8 . The method of  claim 6  wherein an output layer generates predictions for potential fracture driven interactions. 
     
     
         9 . A computer system, comprising:
 one or more processors;   memory; and   one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions that when executed by the one or more processors cause the system to:
 a. receive well completion and production data for a plurality of wells; 
 b. identify parent-child well pairs in the plurality of wells; 
 c. generate event labels based on dynamic pressure exceptions, fluid ratio behavior analysis, and anomaly detection for the parent-child well pairs; 
 d. train a model to identify and characterize fracture driven interactions using the event labels and the dynamic pressure exceptions and fluid ratio behavior; 
 e. receive subsurface reservoir property information including at least one of Shmin direction, compass angle, Pclay, Psand, vclay, vsand, distance from child stage and parent stage to natural fractures, and frac conductivity; and 
 f train the model using the subsurface reservoir property information. 
   
     
     
         10 . The computer system of  claim 9  further comprising instructions that when executed by the one or more processors cause the system to validate the model using blind event data validation and hyperparameter optimization, by segmenting the well completion and production data and the subsurface reservoir property information into a training portion and a testing portion. 
     
     
         11 . The computer system of  claim 9  further comprising instructions that when executed by the one or more processors cause the system to apply the model to a second set of well completion and production data to identify and characterize fracture driven interactions. 
     
     
         12 . The computer system of  claim 9  wherein the identifying parent-child well pairs is based on proximity and timing relevance. 
     
     
         13 . The computer system of  claim 9  wherein the dynamic pressure exceptions and fluid ratio behavior analysis consider pressure and water-oil ratio (WOR) triggers as well as timing relationship to a nearest fracturing stage event. 
     
     
         14 . The computer system of  claim 9  wherein the model may be one of a neural network, decision tree, random forest, XGBoost, or other machine learning model, and has an architecture including multiple layers, with each layer consisting of a set of nodes or neurons. 
     
     
         15 . The computer system of  claim 14  wherein an input layer receives features extracted from the well completion and production data including at least one of pressure, rate, water-oil ratio (WOR), fluid production, and subsurface reservoir property information. 
     
     
         16 . The computer system of  claim 14  wherein an output layer generates predictions for potential fracture driven interactions.