Key Technology Research On Online Water-cut Detection Of Crude Oil

The petroleum industry has many important evaluation parameters.Water content is one of the most important parameters.It has a prominent role in evaluating the value of the oil wells and developing a suitable mining plan.At present,distillation test method is the common measurement of crude oil water-cut detection in China’s oil industry.Using this method is a lack of high accuracy and a waste of human and material resources.It is difficult to meet the actual production requirements.Measuring water-cut detection rapidly and accurately is necessary of predicting the life of the wells,evaluating the production,controlling the cost,achieving oilfield automation management.However,the current technological level is limited.There are lots of problems in measurement accuracy,security and cost.Based on the research status of water-cut measuring instrument,this subject considers studying an online water-cut measuring instrument.It requires that achieving the whole range measuring,high accuracy,and security.First of all,according to the design requirements of the instrument,nine test conditions that affect the test results should be considered.This subject uses AHP to screen six common measurements.Multi-sensor fusion method is evaluated as the highest scoring method on site.Then,the hardware test platform system based on multi-sensor fusion method is designed.It uses neural networks for data fusion to predict water-cut content.The subject of multi-sensor fusion method using capacitance,conductivity sensor for measurement.Temperature sensor is used for temperature compensation.Laboratory experiment results shows that it can achieve the whole range water-cut detection of combining conductance method and capacitance method.The accuracy of this method is higher than single capacitance or conductivity method.When the water content is lower than 3%,the prediction error is less than 0.1%.When the water content is in a range from 3% to 10%,the prediction error is less than 0.5%.When the water content is in a range from 10% to 100%,the prediction error is less than 1.5%.Comparing multisensor method with capacitance method and conductivity method,it shows that temperature has a significant impact on capacitance method and conductivity method.However,because of using neural networks to compensate,the impact of temperature on multi-sensor fusion method is reduced effectively.Moreover,the accuracy is greatly reduced by capacitance method when the water content is higher than 30%.The accuracy is greatly reduced by conductivity method when the water content is lower than 30%.The multi-sensor fusion method combines the two methods together.It can expand the measurement range and improve the measurement accuracy.Therefore,it can meet the different measurement requirements under different conditions.It makes the instrument more generally applicable.Moreover,the electromagnetic radiation of this method is extremely low.The sensors of this method is priced well.Therefore the instrument is cost-effective and will have a widespread application.