| With the most of the oilfields have entered high water-cut stage,the online detection of water content is becoming more and more important,the traditional method of water content detection is difficult to meet the requirements in the detection range,detection precision,reliability and safety.Microwave method based on the dielectric constant could realize the online detection of water content of crude oil with high water cut,However,the detection accuracy of this method will be affected by many factors,such as temperature,salinity content,gas content etc.In order to improve the detection accuracy of water content sensor,this paper aims at the detection system of water Content in high water cut oil wells,through a large number of experiments,the influence of temperature,salinity and gas content on the detection accuracy of microwave sensor is obtained.By using multivariate regression analysis and BP neural network method,the mathematical correction formulas and models for reducing the effects of temperature,salinity and gas on the detection accuracy of the sensor are obtained respectively,by means of the mathematical formula and model,the detection error caused by the temperature of the sensor is less than 0.2%;the detection error caused by NaCl,MgCl2 and CaCl2 is less than1.26%,0.51%and 0.93%respectively;the detection error caused by gas content?including temperature?is less than 5.2%.In order to realize the remote transmission of moisture content measurement data,a scheme for real-time transmission of water content measurement data is proposed,by using the communication protocol based on Modbus,the data transmission mode and transmission rate are determined,and the hardware condition of realizing the function of Modbus protocol is designed.This research can meet the practical needs of water content detection,and obtain more accurate and reliable water content detection data,which is also accords with the development direction of digital oil fields. |
PDF Full Text Request