Model-driven surveillance and diagnostics

Abstract

Performing diagnostic of hydrocarbon production in a field includes generating a thermal-hydraulic production system model of a wellsite and a surface facility in the field, and simulating, using the thermal-hydraulic production system model, and based on multiple root causes, a hydrocarbon production problem to generate a feature vectors corresponding to the root causes. Each of feature vectors includes parameter values corresponding to physical parameters associated with the hydrocarbon production. Performing diagnostic further includes configuring, using the feature vectors, a classifier of the hydrocarbon production problem, detecting the hydrocarbon production problem in the field, analyzing, using the classifier, and in response to detecting the hydrocarbon production problem, surveillance data from the wellsite and the surface facility to identify a root cause, and presenting the root cause to a user. The classifier is configured to classify the hydrocarbon production problem according to the root causes.

What is claimed is: 
     
         1 . A method to perform diagnostic of hydrocarbon production in a field, comprising:
 generating a thermal-hydraulic production system model of a wellsite and a surface facility in the field;   simulating, by a computer processor, using the thermal-hydraulic production system model, and based on a plurality of root causes, a hydrocarbon production problem to generate a plurality of feature vectors corresponding to the plurality of root causes,
 wherein each of the plurality of feature vectors comprises a plurality of parameter values corresponding to a plurality of physical parameters associated with the hydrocarbon production; 
   configuring, using the plurality of feature vectors, a classifier of the hydrocarbon production problem,
 wherein the classifier is configured to classify the hydrocarbon production problem according to the plurality of root causes; 
   detecting the hydrocarbon production problem in the field;   analyzing, by the computer processor, using the classifier, and in response to detecting the hydrocarbon production problem, surveillance data from the wellsite and the surface facility to identify a root cause of the plurality of root causes; and   presenting the root cause to a user.   
     
     
         2 . The method of  claim 1 ,
 wherein the plurality of root causes comprises at least one selected from a group consisting of a change in a reservoir inflow performance, a change in a tubing characteristic, and a change in a surface characteristic.   
     
     
         3 . The method of  claim 1 ,
 wherein the plurality of root causes comprises at least one selected from a group consisting of a zero flow through a downhole pump, low flow rate through the downhole pump, and operating the downhole pump that is not submerged in liquid, and   wherein the plurality of physical parameters comprises at least one selected from a group consisting of electrical current to the downhole pump, electrical voltage at the downhole pump, frequency of the electrical current, well head tubing fluid temperature, well head tubing fluid pressure, downhole pump intake pressure, downhole pump discharge pressure, downhole pump intake fluid temperature, downhole pump motor windings temperature, and well head annulus fluid pressure.   
     
     
         4 . The method of  claim 1 , further comprising:
 obtaining a plurality of probability density functions, wherein each of the plurality of probability density functions represents a probability distribution of measurement noise associated with one of the plurality of parameter values,   wherein the classifier is further configured using the plurality of probability density functions.   
     
     
         5 . The method of  claim 1 , wherein analyzing the surveillance data comprises:
 generating, using the classifier and based on the surveillance data, a classification probability associated with each of the plurality of root causes,   wherein identifying the root cause is based on the classification probability associated with the root cause meeting a pre-determined criterion.   
     
     
         6 . The method of  claim 5 , further comprising:
 obtaining previous surveillance data at a previous time;   analyzing, using the classifier, the previous surveillance data to generate a previous classification probability associated with each of the plurality of root causes, wherein the classifier is a Bayesian classifier; and   obtaining the surveillance data at a current time subsequent to the previous time,   wherein generating the classification probability associated with each of the plurality of root causes comprises updating, based on the surveillance data, the previous classification probability associated with each of the plurality of root causes.   
     
     
         7 . The method of  claim 1 ,
 wherein the plurality of root causes comprise at least one selected from a group consisting of gas failing to flow into a bottom value in a gas lift well, a flowrate to a gas lift well being incorrect, a gas lift valve being stuck in an open position, and an injection through multiple gas lift values.   
     
     
         8 . A system to perform diagnostic of hydrocarbon production in a field, comprising:
 a wellsite and a surface facility in the field for performing the hydrocarbon production;   a surveillance and diagnostics computer system, comprising:
 a model generator executing on a computer processor configured to:
 generate a thermal-hydraulic production system model of the wellsite and the surface facility in the field, and 
 
 an analysis engine executing on a computer processor and configured to:
 simulate, using the thermal-hydraulic production system model and based on a plurality of root causes, a hydrocarbon production problem to generate a plurality of feature vectors corresponding to the plurality of root causes,
 wherein each of the plurality of feature vectors comprises a plurality of parameter values corresponding to a plurality of physical parameters associated with the hydrocarbon production, and 
 
 configure, using the plurality of feature vectors, a classifier of the hydrocarbon production problem, 
 wherein the classifier executes on a computer processor and is further configured to:
 classify the hydrocarbon production problem according to the plurality of root causes, 
 detect the hydrocarbon production problem in the field, 
 analyze, in response to detecting the hydrocarbon production problem, surveillance data from the wellsite and the surface facility to identify a root cause of the plurality of root causes, and 
 present the root cause to a user; and 
 
 
 a repository configured to store the surveillance data and the thermal-hydraulic production system model. 
   
     
     
         9 . The system of  claim 8 ,
 wherein the plurality of root causes comprises at least one selected from a group consisting of a change in a reservoir inflow performance, a change in a tubing characteristic, and a change in a surface characteristic.   
     
     
         10 . The system of  claim 8 ,
 wherein the plurality of root causes comprises at least one selected from a group consisting of a zero flow through a downhole pump, low flow rate through the downhole pump, and operating the downhole pump that is not submerged in liquid, and   wherein the plurality of physical parameters comprises at least one selected from a group consisting of electrical current to the downhole pump, electrical voltage at the downhole pump, frequency of the electrical current, well head tubing fluid temperature, well head tubing fluid pressure, downhole pump intake pressure, downhole pump discharge pressure, downhole pump intake fluid temperature, downhole pump motor windings temperature, and well head annulus fluid pressure.   
     
     
         11 . The system of  claim 8 , wherein the analysis engine is further configured to:
 obtain a plurality of probability density functions, wherein each of the plurality of probability density functions represents a probability distribution of measurement noise associated with one of the plurality of parameter values,   wherein the classifier is further configured using the plurality of probability density functions.   
     
     
         12 . The system of  claim 8 , wherein analyzing the surveillance data comprises:
 generating, using the classifier and based on the surveillance data, a classification probability associated with each of the plurality of root causes,   wherein identifying the root cause is based on the classification probability associated with the root cause meeting a pre-determined criterion.   
     
     
         13 . The system of  claim 12 ,
 wherein the analysis engine is further configured to:
 obtain previous surveillance data at a previous time; and 
 obtain the surveillance data at a current time subsequent to the previous time, 
   wherein the classifier is further configured to:
 analyze the previous surveillance data to generate a previous classification probability associated with each of the plurality of root causes, wherein the classifier is a Bayesian classifier, and 
   wherein generating the classification probability associated with each of the plurality of root causes comprises updating, based on the surveillance data, the previous classification probability associated with each of the plurality of root causes.   
     
     
         14 . The system of  claim 8 ,
 wherein the plurality of root causes comprise at least one selected from a group consisting of gas failing to flow into a bottom value in a gas lift well, a flowrate to a gas lift well being incorrect, a gas lift valve being stuck in an open position, and an injection through multiple gas lift values.   
     
     
         15 . A non-transitory computer readable medium comprising instructions to perform diagnostic of hydrocarbon production in a field, the instructions when executed by a computer processor comprising functionality for:
 generating a thermal-hydraulic production system model of a wellsite and a surface facility in the field;   simulating, by a computer processor, using the thermal-hydraulic production system model, and based on a plurality of root causes, a hydrocarbon production problem to generate a plurality of feature vectors corresponding to the plurality of root causes,
 wherein each of the plurality of feature vectors comprises a plurality of parameter values corresponding to a plurality of physical parameters associated with the hydrocarbon production; 
   configuring, using the plurality of feature vectors, a classifier of the hydrocarbon production problem,
 wherein the classifier is configured to classify the hydrocarbon production problem according to the plurality of root causes; 
   detecting the hydrocarbon production problem in the field;   analyzing, by the computer processor, using the classifier, and in response to detecting the hydrocarbon production problem, surveillance data from the wellsite and the surface facility to identify a root cause of the plurality of root causes; and   presenting the root cause to a user.   
     
     
         16 . The non-transitory computer readable medium of  claim 15 ,
 wherein the plurality of root causes comprises at least one selected from a group consisting of a change in a reservoir inflow performance, a change in a tubing characteristic, and a change in a surface characteristic.   
     
     
         17 . The non-transitory computer readable medium of  claim 15 ,
 wherein the plurality of root causes comprises at least one selected from a group consisting of a zero flow through a downhole pump, low flow rate through the downhole pump, and operating the downhole pump that is not submerged in liquid, and   wherein the plurality of physical parameters comprises at least one selected from a group consisting of electrical current to the downhole pump, electrical voltage at the downhole pump, frequency of the electrical current, well head tubing fluid temperature, well head tubing fluid pressure, downhole pump intake pressure, downhole pump discharge pressure, downhole pump intake fluid temperature, downhole pump motor windings temperature, and well head annulus fluid pressure.   
     
     
         18 . The non-transitory computer readable medium of  claim 15 , further comprising:
 obtaining a plurality of probability density functions, wherein each of the plurality of probability density functions represents a probability distribution of measurement noise associated with one of the plurality of parameter values,   wherein the classifier is further configured using the plurality of probability density functions.   
     
     
         19 . The non-transitory computer readable medium of  claim 15 , wherein analyzing the surveillance data comprises:
 generating, using the classifier and based on the surveillance data, a classification probability associated with each of the plurality of root causes,   wherein identifying the root cause is based on the classification probability associated with the root cause meeting a pre-determined criterion.   
     
     
         20 . The non-transitory computer readable medium of  claim 19 , further comprising:
 obtaining previous surveillance data at a previous time;   analyzing, using the classifier, the previous surveillance data to generate a previous classification probability associated with each of the plurality of root causes, wherein the classifier is a Bayesian classifier; and   obtaining the surveillance data at a current time subsequent to the previous time,   wherein generating the classification probability associated with each of the plurality of root causes comprises updating, based on the surveillance data, the previous classification probability associated with each of the plurality of root causes.