In the petroleum industry, sensor data and information are valuable. It can detect, predict and help to understand processes during oil production. Offshore wells require more attention. Once workovers, maintenance, a...In the petroleum industry, sensor data and information are valuable. It can detect, predict and help to understand processes during oil production. Offshore wells require more attention. Once workovers, maintenance, and intervention are more costly than onshore wells. Coupling data-driven methods for well-monitoring applications, two unsupervised classification methods, one statistical and one machine learning-based, are proposed to detect anomalies in well data. The novelty is presented by applying a Control Chart us</span><span style="font-family:Verdana;">ing a 3 standard deviations window for the Permanent Downhole Gauge Pr</span><span style="font-family:Verdana;">es</span><span style="font-family:Verdana;">sure sensor (P-PDG), and a Fuzzy C-means algorithm to classify data from pr</span><span style="font-family:Verdana;">essure and temperature sensors in an offshore field. The main goal in structuring a classified data set is using it to train machine learning models to monitor and manage petroleum production. Modeling applications for early fault detection systems in offshore production, based on real-time data from production sensors, require classified data sets. Then, labeling two target classes</span></span><span style="font-family:Verdana;">:</span><span style="font-family:""><span style="font-family:Verdana;"> “normal” and “fault” is a key step to be implemented in order to train the machine learning models. Therefore, this paper applies two methodologies to classify a real-time data set to create a training data set divided into “normal” </span><span style="font-family:Verdana;">and “fault” classes. Thus, it is possible to visualize the abnormal events poi</span><span style="font-family:Verdana;">nted out by the methodologies and compare how sensible is each method. In addition, </span></span><span style="font-family:Verdana;">it </span><span style="font-family:""><span style="font-family:Verdana;">is proposed a random forest application to test the performance of the classified data sets from both methods. The results have shown that the con</span><span style="font-family:Verdana;">trol chart method presents higher sensibility than fuzzy c-means, however, th</span><span style="font-family:Verdana;">e </span><span style="font-family:Verdana;">differences between are insignificant. The random forest performance displ</span><span style="font-family:Verdana;">ayed sensitivity and specificity values of 99.91% and 100% for the data set classified by the control chart method and 94.01% and 99.98% for the data set classified by fuzzy c-means algorithm.展开更多
基金supported by National Natural Science Foundation of China(61403244,61304031)Key Project of Science and Technology Commission of Shanghai Municipality(14JC1402200)+3 种基金the Shanghai Municipal Commission of Economy and Informatization under Shanghai Industry-University-Research Collaboration(CXY-2013-71)the Science and Technology Commission of Shanghai Municipality under’Yangfan Program’(14YF1408600)National Key Scientific Instrument and Equipment Development Project(2012YQ15008703)Innovation Program of Shanghai Municipal Education Commission(14YZ007)
文摘In the petroleum industry, sensor data and information are valuable. It can detect, predict and help to understand processes during oil production. Offshore wells require more attention. Once workovers, maintenance, and intervention are more costly than onshore wells. Coupling data-driven methods for well-monitoring applications, two unsupervised classification methods, one statistical and one machine learning-based, are proposed to detect anomalies in well data. The novelty is presented by applying a Control Chart us</span><span style="font-family:Verdana;">ing a 3 standard deviations window for the Permanent Downhole Gauge Pr</span><span style="font-family:Verdana;">es</span><span style="font-family:Verdana;">sure sensor (P-PDG), and a Fuzzy C-means algorithm to classify data from pr</span><span style="font-family:Verdana;">essure and temperature sensors in an offshore field. The main goal in structuring a classified data set is using it to train machine learning models to monitor and manage petroleum production. Modeling applications for early fault detection systems in offshore production, based on real-time data from production sensors, require classified data sets. Then, labeling two target classes</span></span><span style="font-family:Verdana;">:</span><span style="font-family:""><span style="font-family:Verdana;"> “normal” and “fault” is a key step to be implemented in order to train the machine learning models. Therefore, this paper applies two methodologies to classify a real-time data set to create a training data set divided into “normal” </span><span style="font-family:Verdana;">and “fault” classes. Thus, it is possible to visualize the abnormal events poi</span><span style="font-family:Verdana;">nted out by the methodologies and compare how sensible is each method. In addition, </span></span><span style="font-family:Verdana;">it </span><span style="font-family:""><span style="font-family:Verdana;">is proposed a random forest application to test the performance of the classified data sets from both methods. The results have shown that the con</span><span style="font-family:Verdana;">trol chart method presents higher sensibility than fuzzy c-means, however, th</span><span style="font-family:Verdana;">e </span><span style="font-family:Verdana;">differences between are insignificant. The random forest performance displ</span><span style="font-family:Verdana;">ayed sensitivity and specificity values of 99.91% and 100% for the data set classified by the control chart method and 94.01% and 99.98% for the data set classified by fuzzy c-means algorithm.
