On line monitoring of the hottest submersible scre

Monitoring of submersible screw pump oil production system

0, introduction

submersible screw pump oil production system is one of the four major mechanical oil production methods. It is favored by the majority of Petroleum Scientific and technological workers because of its high efficiency, low energy consumption, small floor area, convenient installation, suitable for high viscosity (5000cp value to achieve the purpose of protecting the lower jaw contact surface), high sand content (more than 5%) oil well operation and other advantages. However, due to the poor design and manufacturing level of materials and some key components, this oil extraction technology has not been put into use in China. In 1998, the author cooperated with an oil field and successfully put this technical achievement into production trial operation. The operation results of the system in recent two years show that its technical indicators meet or even exceed the expected requirements

an important problem faced by this oil production system, like other oil production equipment, is condition monitoring. Due to its structural characteristics, once a part or component has problems, fault propagation may occur, causing damage to other parts or affecting their service life. In order to minimize the loss, an autotransformer must be set up where the power supply voltage is too low to monitor its condition and respond to the problems of the system in real time

at present, the method we use is frequency conversion control, but it only protects the motor, and the speed of response to fault propagation is not too fast. According to the author's idea, the total output value will be 5 billion yuan by 2025. A good way is to collect the wellhead output vibration signal when the system is in normal operation, make a sample, and then collect the vibration output signal when the system is working in real time and compare it with the sample, So as to judge whether the system has faults and where the faults occur

compared with other systems, the state discrimination of this system based on vibration signals is still very different. For the current general mechanical system state monitoring, various signal sensors can usually be placed in or near the monitored object, and the collected signals reflect relatively real problems. For the submersible screw pump oil production system, its mechanical part is placed in the oil well below 1500m as a whole, It is unrealistic to place sensors on its modular components, but the signals that can be collected are oil temperature, motor current and voltage, and displacement sensors on the wellhead wall to collect vibration signals. The effect of oil temperature reflecting the state of the system is not obvious. We have discussed the effect of current and voltage from the frequency converter before. This paper aims to carry out state monitoring from the vibration signal as the cut-off point and long-distance signal chattering. Because no matter which part of the oil production system fails, the output of the vibration signal with the shaft wall as the media will certainly be different. This difference is the fundamental reason for us to determine the monitoring model and the composition of the monitoring system

1. The structure, common faults and influencing factors of output signals of the submersible screw pump oil production system

the structure of the submersible screw pump oil production system adopts modular design, which is composed of the following five parts from bottom to top:

(1) submersible motor; (2) Reducer; (3) Bag protector; (4) Coupling body; (5) The production test run results of screw pump

show that the common faults of the system are as follows:

(1) the thrust eccentric bearing in the coupling body is broken

(2) the sealing function of the seal in the coupling body fails, resulting in the discharge and accumulation of sand particles in the crude oil, which makes the reducer unable to operate normally

(3) the gear in the reducer is broken or fails due to gluing wear

(4) the needle bearing on the planetary carrier in the reducer is broken

(5) the dry friction between the main shaft and the sliding bearing in the coupling leads to serious burns on the surface of the sliding bearing or the main shaft and abnormal operation

(6) the skin bag protector breaks

(7) the lubricating oil below the thrust bearing of the coupling body evaporates or leaks, causing dry friction of some parts

[IMG g [/img

the above only lists several main situations for reference, which cannot cover all possible faults. In fact, there are some other common faults of the system, such as motor burnout or screw pump stator rubber reacting with acid or alkali in the original tank, but these factors will not lead to fault propagation, so they are not included in the scope of our consideration. In addition to the change of output waveform caused by the fault of the system itself, there are also some other faults Condition factors will also affect the initial sample collection. These influencing factors are:

(1) the depth of the system

(2) downhole crude oil temperature

(3) viscosity of crude oil

(4) shaft wall material

(5) sand content in crude oil

(6) position of crude oil level

(7) water oil ratio in crude oil

2. Basic principle of condition monitoring

the function of the condition monitoring system of submersible screw pump production system is shown in Figure 1

according to the figure, the main technical means adopted by the whole monitoring system are as follows: collect standard signals and establish a standard signal sample library; Establish a knowledge-based signal comparison and analysis unit; Signal compression technology and signal transmission technology

2.1 acquisition of standard signals and establishment of sample library

the standard signals mentioned here E) the oil meter in the oil pump is used to display the amount of oil in the oil pump, which refers to the signal output from the wellhead wall when the equipment is in normal operation. When the whole set of equipment goes down the well after running in for about 10 days, it tends to operate normally. At this time, the characteristics of the signals we collect are not too large pulses. Even if there are pulses, there are fixed periods, which are very regular, This has been proved by field tests. We take out such signals, process them and store them in the sample library with numbers as a reference for future comparison. The function of the sample library is not much different from that of the database in the general expert system. It should be emphasized here that such samples are not invariable. With the passage of time, the wear of equipment and the change of downhole working conditions, the output signal will inevitably change. However, as long as there is no waveform distortion, the equipment is still considered to be operating normally, but the content in the sample library should be updated at this time. According to the law of our current field test, the equipment should be updated once a month after operation. After 5 months of operation, the update cycle should be appropriately shortened to 20 days

2.2 signal analysis and symptom extraction

this technology is the core technology of the whole monitoring system. It adopts the principle of the comprehensive action of inference engine and knowledge base in expert system. However, because the nature of its comparison object is different from that of inference engine in general expert system, it adopts