_ This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 224634, “Study on the Application of Intelligent Risk-Monitoring Technology for Workover Operations, ” by Shaohui Zhang, SPE, Weihe Huang, SPE, and Zehao Lv, PetroChina, et al. The paper has not been peer-reviewed. _ Risk monitoring is critical for ensuring workover-operation quality, efficiency, safety, and economic benefit. At present, the risks of workover operations mainly are judged by on-site workers; automatic acquisition of operation data, intelligent risk identification, and warning have not been realized. To improve workover operations, it is necessary to conduct research on intelligent risk-monitoring technology. With this aim, a study of key technologies for intelligent risk monitoring was performed from the standpoints of hardware, algorithms, and software. Introduction Workover efficiency can be improved and the occurrence of operation accidents reduced with the deep integration of intelligent risk-monitoring technology and traditional workover operations. Currently, digital supervision and intelligent decision-making of workover operations are still in the initial stage. Data-acquisition methods in workover operations are insufficient, the data lake has not been established, and data sharing has not been realized. On-site risks are judged mainly by worker experience, resulting in high accident rates and low operational efficiency. Data Acquisition and Standard Storage of Workover Operations Design and Development of Data-Acquisition Device. A data-acquisition device for workover operations was designed and developed, including basic sensors, central control unit, and management terminal. Data-transmission paths are determined, allowing real-time key data including hook height, string depth, string weight, string speed, wellhead pressure, and hazardous gas to be collected. The circuits of sensor interface, data processing, central control, communication, and other modules were designed. Multiple commonly used acquisition interfaces were designed for the sensors. The appearance of the data-acquisition device was designed based on explosion-proof requirements and installation environments at wellsites. The circuit board and related components were soldered and installed. The device then passed laboratory testing and functional verification. Database of Workover Operations. The data contents, types, and formats of real-time data, historical data, and risk-monitoring data in workover operations were sorted and determined. The business logic relationships of the database and 32 data tables with over 680 data fields were designed. The database of workover operations was thus established. Risk-Monitoring and Warning Models of Workover Operations Risk-Monitoring Model of String Weight Based on Historical Data. Data Processing. Data Cleaning. In the data-processing stage, raw data was first cleaned to remove duplicates, outliers, and anomalies, ensuring data integrity and accuracy. On this basis, the Z-SCORE standardization method was used to normalize the data, making sure that the value distribution of different features tended to be consistent, thus providing a standardized data basis for subsequent analysis. In addition, statistical analysis based on Pearson correlation coefficients was conducted to deeply explore the correlations between feature variables, reveal the inherent connections among features, and provide scientific basis for subsequent modeling.
Chris Carpenter (Fri,) studied this question.