Experimental research on the reflection and response characteristics of pressure pulse waves for different gas pipeline blockage materials

Natural gas pipelines are exposed to complex production and transportation environments, leading to blockages of varying compositions and forms due to various factors. This paper proposes a pressure pulsing wave‐based pipe blockage detection method, and a series of pipeline blockage detection experiments of different materials are carried out, including hydrate blockage, ice blockage, paraffin blockage, and water blockage. Through the blockage detection calculation theory and the reflection and transmission characteristics during pressure wave emission, the blockage position, length, and section blockage rate under different blockage materials are obtained, and the influence of different blockage materials on the reflection and transmission signal characteristics of the pressure pulse wave is analyzed. The results show that the solid blocking material does not noticeably impact the reflectivity and reflective properties of pressure waves, and different cross‐sectional shapes have a noticeable impact on the reflectivity of the pressure wave. The reflected wave waveform of the liquid blockage material is visible but irregular. The average error of blockage location is 0.78%.


| INTRODUCTION
2][3][4] Under normal circumstances, natural gas resources extracted from underground or underwater experience drying, filtering, and purification processes to remove various impurities before being transported through pipelines for utilization. 5However, the natural gas extracted underground or underwater is often not completely purified and contains impurities such as water, condensate oil, paraffin, and sediment.As the pipeline length increases, these impurities continuously settle and accumulate, eventually leading to pipeline blockages. 6Additionally, under special circumstances featuring low-temperature and high-pressure conditions can form gas hydrates, which lead to blockages. 7cross some low-lying pipeline segments, water can also accumulate continuously, resulting in pipeline blockages. 8n a word, various blockages pose significant risks to the safe operation of natural gas pipelines. 9Once blockages occur, they can easily lead to production stoppage, equipment abandonment, and other major engineering disasters, resulting in severe economic losses.][12][13] Through the efforts of researchers, there has been some progress in blockage detection technology.Rogers 14 and Hiebert 15 studied using stress-strain measurements to detect blockages in pipelines.When a fluid flows through a section of pipe that changes in cross-sectional area, the pressures inside the pipeline change, resulting in variations in wall strain.Measuring the change in wall strain can effectively detect pipeline blockages.However, this method requires the installation of large numbers of strain sensors along the pipeline wall, and the detection accuracy is limited.Therefore, both feasibility and costeffectiveness need to be improved.Koyama, 16 Parker, 17 and Watanabe 18 researched the use of acoustic methods to detect pipeline blockages.This method is theoretically feasible, but the acoustic signals are susceptible to environmental noise interference.Enhancing the antiinterference capability is a pressing issue that needs to be addressed.Li demonstrated for the first time the existence of a significant pressurization phenomenon in a pipe filled with water, successfully determined the optimal pulse frequency to produce this phenomenon and the mechanism behind it, and further analyzed the square-wave pulse effect and its relationship with the pulse frequency, amplitude, pipe size, and wave velocity, a discovery that provides an important theoretical basis for understanding and utilizing the pulse pressurization phenomenon in pipes. 19In addition, the dynamic behavior during pulse pressurization can be better observed and analyzed in a two-phase flow system where water containing a small amount of gas flows inside a pipe. 202][23] When the crosssectional area of pipes changes due to blockages, valves, or restricted areas, pressure waves occur and reflect and transmit at these locations.By monitoring the time, amplitude, and frequency information of the incident and reflected waves using sensors, the position of abnormal points in the pipeline and other relevant information can be calculated, achieving fast and accurate detection of pipeline anomalies.Contractor et al. 24 found that pressure waves undergo reflection when encountering blockages, and the arrival time of the reflected wave can provide information about the blockage location, while the vibration of reflected waves can determine the severity of the blockage.Zhou and Adewumi 25,26 used high-order numerical simulation methods with high resolution to simulate upstream and downstream propagation and attenuation processes of pressure waves in natural gas pipelines, studying and simulating the characteristics of pressure wave propagation, reflection, and attenuation in natural gas pipelines.Stewart et al. 27 proposed the use of rapid on and off valves to create pressure waves and recorded the changes in pressure wave data for all types of blockage detection in actual pipelines.Tian et al. 28,29 artificially created pressure pulse waves at the pipeline inlet and used the time-splitting and mixed format TVD/Godunov algorithm to simulate partially blocked pipelines, analyzing the blockage situation.However, this method involves complex computational processes, significantly increasing the computational workload with no guaranteed accuracy in practical applications.For the discontinuous pulsating wave problem, Li proposed a novel flux vector splitting (FVS) formulation based on the weakly compressible Navier-Stokes equations, and demonstrated the reliability and robustness of this FVS formulation by validating it with engineering examples of water hammer and pulsating fracturing. 30A new algorithm based on the MacCormack method is also developed to study and validate the pulsation pressurization phenomenon in two-phase flow.The reliability of the algorithm is tested through Poiseuille's theory and available experimental data on two-phase flow. 31Adekeke et al. 32 proposed a numerical model to accurately describe the features of blockages in natural gas pipelines.However, due to the high complexity of the calculations, the same issues of increased computational effort and uncertainty of accuracy exist in practical pipe applications.
In summary, the existing blockage detection methods have various issues, including short detection distance, limited accuracy, high cost, and low efficiency.Existing research on blockage detection in natural gas pipelines based on pressure wave methods is predominantly focused on numerical simulations and lacks sufficient experimental research and data.As mentioned earlier, blockage materials in natural gas pipelines vary, and their distribution forms within the pipeline are also diverse.The use of pressure pulse waves for detection and the resulting reflection signal patterns can vary for different blockage materials and forms.Therefore, understanding the reflectance and response characteristics of pressure pulses with various types of blockage materials in natural gas pipelines is essential for precise localization and detection of blockage segments.Conducting experiments on the propagation and response characteristics of pressure pulse wave signals with different blockage materials is of great significance for the engineering application of pressure pulses-based blockage detection technology in natural gas pipelines.
To address this, this study designs and constructs an experimental system for detecting blockages in gas pipelines using pressure pulse waves.Based on the different compositions of blockage materials and the physical and chemical differences that cause blockage formation, the study classifies the reasons for blockage generation in natural gas pipelines.
It investigates the reflectance and response characteristics of pressure pulses for various blockage components, including hydrate blockages, wax blockages, and water blockages.With the experimental system for detecting blockages in gas pipelines using pressure pulse waves, which includes controllable temperature sections and U-shaped bends, various blockage detection experiments are conducted at different crosssectional blockage ratios for different blockage materials.The study analyzes the impact of different blockage materials on the reflection and transmission signal characteristics of pressure pulses and achieves highprecision detection of blockage position, length, and cross-sectional blockage ratio for different blockage materials.This provides feasible solutions for identifying different blockage material compositions in natural gas pipelines in engineering practice.

| Experimental equipment
To examine the response characteristics of pipelines blocked by different substances, an experimental system for checking the blockage of pipelines using pressure pulses was constructed as shown in Figure 1.
The experimental station for detecting pipeline blockages using pressure pulses has a total length of 220 m and adopts a spiral layout, which effectively reduces the interference of pipe bends on the propagation of pressure waves.This system consists of the following five modules: the gas injection module, the F I G U R E 1 Conceptual plan of the pressure pulse blockage-detection experimental station.main channel, the simulated blockage section, the data collection module, and the pressure pulse wave control module.
The gas injection module consists of a gas cylinder assembly, pressure regulator valve, and intermediate vessel.High-purity nitrogen gas is selected as a supply of air for the whole system, with a pressure of 14 MPa and a volume of 40 L, to maintain stable pressure inside the pipeline.The main channel is made of 316 L stainless steel to effectively reduce the interference of pipe bends on the propagation of pressure pulses.All bends are quarter-circle bends with a radius of 800 meters.The simulated blockage section is used to simulate blockages of different substances in the pipeline, including hydrate blockage, ice blockage, and paraffin blockage.It also includes a U-shaped bend segment to simulate water blockages of different degrees.The data collection module consists of a resistance temperature sensor, a static pressure sensor, three dynamic pressure sensors, and an industrial computer to monitor the attenuation changes of the pressure pulse wave.The pressure pulse control module consists of electromagnetic valves, branch pipes, needle valves, and fine pipe segments.When the solenoid valve is controlled to open and close quickly, the high-pressure fluid in the pipe will be released in a short time, and at the same time, a pressure pulse wave is generated, which propagates through the branch pipe to the main pipe.The pressure wave propagates, decays, and reflects along the pipe until it finally decays to zero.

| Blockage calculation and analysis
When the pressure wave starts propagating along the pipe, its signal is first evaluated at D 1 and denoted by p 1 .When it reaches the position of D 2 , it is denoted by p 2 , and when it propagates to the front edge of the blockage section, it is denoted by p 3 .At this point, the cross-sectional area significantly decreases, and the pressure pulse wave undergoes partial reflection, denoted by p′ 3 .Simultaneously, another portion of the pressure wave passes through the blockage portion, propagating in the same direction as the incident wave, and is denoted by p″ 3 .p′ 3 propagates in other direction and reaches the position of D 2 , denoted by p 4 .p″ 3 continues to be spread and reaches the rear edge of the blockage section, denoted by p 5 .As the pipe diameter significantly increases, p 5 undergoes reflection and transmission.The positive pressure wave propagating in the same direction is denoted by p′ 5 , while a necessary negative pressure wave travels through the blockage section in another direction, arriving at the position of D 2 and denoted by p 7 .The length between D 1 and D 2 is denoted by L 1 , the length between D 2 and the leading edge of the partially blocked section is denoted by L 2 , the length of the partial blockage section is denoted by L 3 , and the length between the rear edge of the partial blockage section and D 3 is denoted by L 4 .The pressure pulse wave propagation path is shown in Figure 2.
By analyzing the arrival time and amplitude of both reflective and transmissive waves, it is possible to effectively calculate the blockage position, length, and cross-sectional blockage ratio.The position of pipeline blockage detected by the pressure pulse wave method is calculated as (1 where C is the traveling speed of the pressure wave.The time off feature of p 4 is T 4 .
The distance to the portion of the blockage is given by the formula (2) where the time of feature of p 7 is T 7 .Then, the crosssectional blocking rate is defined as the ratio of the crosssectional area of the blocking front to the cross-sectional Pressure wave propagation in the pipelines.
area of the normal pipe, which can be calculated from the ratio of the amplitude of the reflected wave p 4 to the amplitude of the incident wave p 1 at the blocking front.
According to the theory of exponential decay of sound wave amplitude with propagation distance in linear acoustics 33 where α is the attenuation coefficient during the propagation of the pressure pulse wave, P 1 is the amplitude of the incident wave p 1 , and P 2 is the amplitude of the pressure wave p 2 .P 3 is the amplitude of the pressure wave p 3 , which is calculated as follows X s is the discontinuity surface distance, which can be represented as where β is a nonlinearity parameter, the Mach number (Ma) is a dimensionless quantity that expresses the ratio of an object's velocity to the speed of sound in the surrounding medium, k is the wave count, ρ 0 is the rest gas density, and γ is the specific heat of the gas, δT is the pulse width, and σ 3 is a dimensionless number expressed as The amplitude P′ 3 of p′ 3 is calculated as follows σ 3 and σ 4 are dimensionless numbers, which are calculated as follows Then, based on the theoretical relationship between P 3 and P″ 3 s and the cross-section clogging rate z, 34 the cross-section clogging rate can be calculated as follows The cross-sectional area of the blocking portion is calculated by the formula where z is the percentage blockage of the partially blocked cross-sectional area.P 1 and P 4 are the magnitudes of p 1 and p 4 .Detailed analysis and calculation of blockage location, length, and cross-sectional blockage rate of gas pipes based on the pulse pressure wave method can be found in the author's previous study. 33

| RESULTS AND DISCUSSION
Six groups of experimental tests were conducted to detect hydrate blockage, ice blockage, paraffin blockage, and water blockage at different cross-sectional blockage ratios.Each group of experiments included four sets of repeated tests.The data obtained from the blockage detection experiments are presented in Table 1.

| Hydrate blockage
In each experiment, a certain amount of tetra-nbutylammonium bromide (TBAB) with a mass concentration of 30 wt% was injected into the controllable temperature section of the experimental system.The temperature inside the controllable temperature section was maintained at 1°C (above 0°C to ensure the formation of hydrates instead of ice) using a water bath.This was done to simulate hydrate blockage at different cross-sectional blockage ratios.Photographs of hydrate blockages at six different cross-sectional blockage ratios are shown in Figure 3.The cross-sectional blockage ratios for the six groups of experiments were 0.18, 0.28, 0.37, 0.46, 0.55, and 0.64.One case of experimental results was selected from each group of experiments and plotted in Figure 4, which shows the results for all six groups of experiments.By observing the results of the six experimental cases, it can be seen that the shape of the bounce signal produced by the hydrate blockage is similar to that produced by the long partial blockage (for a detailed description of the long partial blockage, please refer to the author's previous study 33,35 ).They are composed of a reflected negative pressure wave produced in front and a reflected positive pressure wave produced behind.The waveform is also similar to the incoming wave, as the length of the blockage section is relatively short; this leads to a short interval between the negative and positive pressure waves.Based on the calculations from the equation, the blockage section positions, lengths, and cross-sectional blockage ratios were obtained for the 24 experiments.The average results for each crosssectional blockage ratio were calculated by taking the average of the six rounds in each group, as indicated in Table 2.
It can be observed that for the blockage location detection, the minimum and maximum detection error are experiments is 0.06% and 0.34%, respectively.The overall average error for the 24 experiments is 0.22%.As for the blockage length detection, due to the size limitation of the controlled temperature pipe section, the generated length of the hydrate blockage is only 2.36 m, which is approaching or slightly smaller than the pressure pulse wave wavelength.Therefore, the blockage length is subject to relatively high detection errors, with the minimum average detection error being 8.90%, the maximum average detection error being 22.46%, and the overall average error for the 24 experiments being 13.99%.For the detection of cross-sectional blockage ratio, the calculated results for hydrate blockage significantly overestimate the actual blockage ratio, and there are notable discrepancies compared to the results of partial extended blockage.Several factors could contribute to these results, including the consideration of different substances and different cross-sectional The hydrate blockage cross-section pictures with six different blockage rates.
shapes that may cause variations in the reflection characteristics.
First, let us discuss the differences in reflection caused by different substances.Since the physical nature of the pressure pulse wave is similar to that of a finiteamplitude sound wave, we can borrow concepts from acoustics theory to discuss the differences in reflection caused by different substances.According to acoustic theory, 34 when a sound wave transitions from one medium to another, reflection and transmission occur at the interface between the two media.The magnitude of reflection and transmission is determined by the characteristic impedance of the media.The calculation formula for characteristic impedance can be expressed as In the equation, R represents the characteristic impedance of the medium, ρ represents the density of the medium, and c is representative of the velocity of propagation of sound waves in a medium.For nitrogen gas, at a pressure of 2.0 MPa and a temperature of around 20°C, its density is approximately 22.6 kg/m 3 , and the speed of sound is about 350 m/s.For stainless steel, its density is around 7.8 × 10 3 kg/m 3 , and the speed of sound is about 5200 m/s.For TBAB hydrate, its density is usually greater than 1.0 × 10 3 kg/m 3 , and the speed of sound is about 3000 m/s. 36It can be observed that although the characteristic impedance of TBAB hydrate is slightly smaller than that of stainless steel, both are solid materials with characteristic impedances much greater than that of nitrogen gas.Therefore, when a sound wave or pressure pulse wave propagates in nitrogen gas and encounters solid materials such as stainless steel or TBAB hydrate, it can be approximated that total reflection occurs, and there is no transmitted wave entering the solid medium.Hence, we can conclude that the differences in reflection caused by different substances can be neglected.
Apart from the differences in the blocking materials, there are also variations in the cross-sectional shapes of the hydrate blockages.Two kinds of blockage crosssection shapes are shown in Figure 5.

| Ice and paraffin blockage
In the ice blockage experiments, a certain amount of pure water was injected into the controllable temperature section of the experimental system.The temperature inside the controllable temperature section was brought at −10°C using a water bath to freeze the water into ice, simulating ice blockages of different cross-sectional blockage ratios.In the paraffin blockage detection experiments, a certain amount of liquid paraffin oil (n-tridecane, molecular formula C 13 H 28 ) was injected into the controllable temperature section.The temperature inside the controllable temperature section was brought at -15°C using a water bath to solidify the

F I G U R E 6
The ice blockage cross-section pictures with six different blockage rates.

F I G U R E 7
The paraffin blockage cross-section pictures with six different blockage rates.
n-tridecane into a solid state, simulating paraffin blockages of different cross-sectional blockage ratios.
Figure 6 shows the cross-sectional photos of ice blockages with six different cross-sectional blockage ratios.Figure 7 shows the cross-sectional photos of paraffin blockages with six different cross-sectional blockage ratios.The cross-sectional blockage ratios for the six ice blockage experiments were 0.17, 0.26, 0.34, 0.44, 0.55, and 0.67, while the cross-sectional blockage ratios for the six paraffin blockage experiments were 0.16, 0.24, 0.32, 0.40, 0.48, and 0.59.
The ice blockage and wax blockage experiments selected two groups each, and the results are shown in Figure 8.By observing these four groups of experimental results, the reflected signals generated by ice blockage and wax blockage are similar in shape to those generated by hydrate blockage and extended blockage.Therefore, only two groups of experiments for each type of blockage are selected for presentation here.
According to the calculations, the blockage position, length, and cross-sectional blockage rate detection results were obtained for 12 groups and 48 experiments for ice blockage and wax blockage.The average detection results for each cross-sectional blockage rate are listed in Table 3 and Table 4.The calculated results in the table can be observed that for the blockage position detection results, the smallest average detection error after each group of experiments was 0.14%, the largest mean detection error was 0.37%, and the overall average error for the 48 experimental cases was 0.26%.Regarding the length detection results, similar to the detection results for hydrate blockage, there is a higher detection error due to the size limitation of the controllable temperature pipe segment.The minimum average detection error after each group of experiments is 7.63%, the largest average detection error is 23.31%, and the overall average error for the 48 cases of experiments is 15.71%.
For the cross-sectional blockage rate detection results, the smallest average error after each group of experiments is 3.33%, the largest average error is 6.67%, and the overall average error for the 60 experiments is 5.09%.In the experiments with three different types of blockage materials in this section, during the process of gas injection and the propagation of pressure waves in F I G U R E 8 Ice blockage with blockage rates of 0.56 and 0.67; paraffin blockage with blockage rates of 0.48 and 0.59. the pipeline, the gas can cause scouring on the blocked section, leading to larger errors in interface blockage rate detection.Further improvements in the experimental methods are needed to investigate the impact of crosssectional shape on the reflection characteristics of pressure waves.

| Experimental study on water blockage detection
In the process of natural gas extraction and transportation, the increase in extraction depth or failure of dehydration mechanisms can often result in an increase in water content in the pipeline.This water tends to accumulate in low-lying areas, forming water blockages that affect the normal transportation and production of natural gas.To simulate these water blockages, a U-shaped pipe section is installed in the experimental system for pressure pulse wave gas pipeline blockage detection.A water injection pipe is used to inject an appropriate amount of clean water into the bottom of the U-shaped bend pipe, simulating the presence of water blockages in real natural gas pipelines.The schematic diagram of the U-shaped bend pipe is presented in Figure 9.
First, 1500 mL of clean water was injected into the bottom of the U-shaped bend pipe, creating a water blockage with a total of 2.76 m and a blockage rate of 36%.The valve was opened to pressurize the pipeline to 2 MPa, and the electromagnetic valve was opened to conduct the blockage detection experiment.The experimental outcome is illustrated in Figure 10.It can be observed that there is still a significant blockage reflection signal when the water blockage rate is 36%.Based on the theoretical calculations for long-distance extended blockages mentioned above, the position, length, and blockage rate of the water blockage were calculated.The results are displayed in Table 5.Then, | 2515 a larger amount of clean water was injected to fill the entire bottom of the U-shaped bend pipe, resulting in a blockage rate of 100%.The experimental findings are illustrated in Figure 11.
From the graph, it can be observed that at a water blockage rate of 100%, there is a more regular blockage reflection signal, and the signals indicating the front and back edges of the blockage can be clearly distinguished.Due to the water oscillation caused by the pressure wave, the reflection signals exhibit significant code wave oscillations.Additionally, before the U-shaped blockage reflection signal, there is a section of continuous irregular reflection signal (p 3 ).This signal is attributed to the water oscillation inside the U-shaped bend caused by the gas injection and pressure wave, resulting in partial water blockage at the front end of the bend.This phenomenon further confirms the effectiveness of the method in water blockage.The theoretical calculations for the water blockage in the U-shaped bend, based on the aforementioned extended blockage calculation theory, are presented in Table 5.
As can be seen from the calculation results, the detection errors for the blockage positions in both groups of experiments are relatively small, with values of 0.94% and 0.61%, respectively.These errors are not markedly dissimilar to the detection errors for long-distance extended blockages.
When calculating the blockage length, at a low blockage rate of 36%, the reflection signals from the water blockage are highly irregular.Therefore, it is challenging to accurately identify the arrival time of the reflection wave at the back edge of the water blockage.If the middle position T 5 is chosen, the calculated result is much higher than the actual value, indicating that T 5 is not the correct point.After the pressure wave reaches the water blockage position, its intense wave fluctuations cause the water to oscillate back and forth.Additionally, when the pressure wave reaches the gas-water interface, it does not fully reflect but refracts partially into the water phase and then refracts back into the gas phase.This overlapping of pressure waves leads to highly complex reflection signals from the water blockage, with a much longer length compared to reflection signals from solid blockages of the same length.Based on the pattern observed in the reflection signals from solid blockages, it is speculated that the first negative pressure wave and the first positive pressure wave of the water blockage reflection signal are close to the reflection signals at the front and back edges of the water blockage.Therefore, T 4 is selected as the arrival time of the reflection wave at the back edge of the water blockage, and the calculated water blockage length based on T 4 and T 3 is 3.26 m, with an error of 18.12%.Through analysis, it is found that besides the inherent characteristics of the reflection signals from water blockages mentioned earlier, the large error is also caused by the short length of the water blockage, which is shorter than the wavelength of the wave (in this experiment, the pulse width of the pressure wave is 9.8 ms, and the wavelength is 3.42 m).This results in a slight superposition of the reflection waves from the front and back edge of the water blockage, leading to an overestimation of the selected value T 4 .When the blockage rate is 100%, the characteristic times of the blockage reflection signals are clear, and the calculation error of the  blockage length is 1.09%, which is similar to the detection error for long-distance extended blockages.The calculated results for the cross-sectional blockage rate in the two experiments are 11.10% and 87.80%, with errors of 74.36% and 12.20%, respectively.Through analysis, it is found that the main reasons for the large errors are as follows: First, the calculation model assumes an ideal interface between the blockage material and the gas, where the pressure wave undergoes total reflection upon encountering the ideal interface without refraction.However, the gas-water interface does not satisfy this assumption.The pressure wave experiences refraction at the gas-water interface, resulting in the consumption of energy.Second, unlike solid blockages, the water phase undergoes movement and oscillation when subjected to intense pressure wave fluctuations.During this process, partial pressure wave energy is converted into kinetic energy of the water, further depleting the energy of the pressure wave.Lastly, due to experimental constraints, the water blockage is placed inside a U-shaped bend.Therefore, when the blockage is partial, the blockage section is not an ideal single plane perpendicular to the pipe diameter, leading to significant errors in blockage rate detection.
In summary, in the water blockage experiments, the reflected signals of the 36% blockage rate are visible but have irregular waveforms.When the blockage rate is 100%, the reflected signals are clear, and the waveforms are regular and complete.Based on this, the detection errors for the blockage position in the two experiments are 0.94% and 0.61%, respectively, the detection errors for the blockage length are 18.12% and 1.09%, respectively, and the detection errors for the blockage rate are 74.36% and 12.20%, respectively.The major reasons for large errors in blockage rates and the blockage length in the partially blocked section include continuous refraction of the pressure wave, movement and oscillation of the water, and the blockage section not being an ideal plane perpendicular to the pipe diameter.Additionally, the reason for the large error in the calculation of the blockage length is that the distance of the water blockage is shorter than the wavelength of the wave.

| CONCLUSIONS
To verify the accuracy of the established pressure pulse wave pipeline blockage signal analysis method and calculation theory, experimental research was conducted using the pressure pulse wave gas pipeline blockage detection system to investigate the reflection and response of different blockage segments, including hydrate blockage, ice blockage, paraffin blockage, and water blockage.The experimental results were analyzed: (1) For different blockage materials, experimental findings show that when the blockage material is solid (such as hydrates, ice blocks, and paraffin), materials do not noticeably influence the reflectivity and reflection characteristics of the wave, while different crosssectional shapes have a noticeable impact on the reflectivity of the wave.When the blockage material is water, the reflected wave in the blocked segment is visible but irregular in shape.The average positioning error for blockage detection is 0.78%.The detection errors for blockage segment length and cross-sectional blockage rate are significantly higher.This is mainly due to the continuous refraction of the pressure pulse wave at the gas-water interface, the movement, and oscillation of water, and the fact that the blockage section is not an ideal plane perpendicular to the pipe diameter direction.
(2) The experimental results indicate that the method for blockage detection has high detection accuracy and applies to complex blockage conditions.Therefore, it exhibits good reliability and practicality.This paper's calculation methods and models demonstrate good detection accuracy for relatively regular extended blockages.However, for irregular and complex blockages, such as water, there is still room for improvement in the accuracy of detecting blockage length and rate.(3) In this study, the calculation of blockage position assumes that the wave velocity of the pressure pulse remains constant during propagation.However, in actual natural gas pipelines, there are variations in temperature, gas composition, and other factors along the pipeline, which lead to differences in the wave velocity of the pressure pulse.These differences can result in increased positioning errors.Therefore, further research is needed to address this issue and establish a more accurate model for the propagation of pressure pulses.This will enable the prediction of wave velocities at different positions and under different conditions, thereby improving the accuracy of blockage localization.

F I G U R E 4
Hydrate blockage detection tests results with different blockage rates.LIANG ET AL. | 2511 R ρc = .

F
I G U R E 5 (A) Shape of the extension blockage.(B) The shape of the hydrate blockage cross-section.

F
I G U R E 10 Water blockage detection (blockage rate 36%).T A B L E 5 Prediction results of water blockage detection.
T A B L E 2 Blockage characterization prediction results on average for hydrate blockage detection.
T A B L E 3 Blockage feature prediction result on average for ice blockage detection.Blockage feature prediction result on average for paraffin blockage detection.
LIANG ET AL.