Controlling Speed of Pig during Gas Pipeline Cleaning

1 Factors Affecting PIG Speed

With the exponential growth in global energy demands, petroleum industries’ research and development work has also increased likewise for the enhanced oil and gas recovery through pipelines. During the transportation of oils and gases through pipelines, a pipeline inspection gauge (PIG) is used for pipeline cleaning and inspection, and this process is called Pigging. Sometimes, the PIG stops due to hindrance by the accumulated paraffin. In such a case, the PIG can resume its motion only if sufficient rear pressure is introduced. The resumption of PIG movement by an augmented rear pressure is termed as “speed excursion”. However, this motion is spontaneous and uncontrollable that can consequently damage the PIG itself and the pipelines.

The speed of a pig during pipeline cleaning is an important factor that needs to be controlled and optimized for an efficient operation. Three major factors that affect the PIG speed are

1. The bypass diameter
2. The slope of the pipe
3. Differential pressure.

The bypass flow is integrated into pigging operation to control the speed of the PIG by controlling the pressure gradient across it. The fluid flux through the bypass is adjusted to maintain a specific pressure drop across the PIG, thus optimizing its speed [1]. A study reported an optimum speed of PIG to be 1—5 m/s for a gas flow in a pipeline [2]. The bypass diameter is found to have an inverse effect on PIG speed. The increase in bypass diameter results in a decrease in the average velocity of the PIG in a pipeline. A study reports a PIG speed of 1.22 m/s for a bypass diameter of 1.0″ whereas a pig speed of 0.8 m/s is achieved for a bypass diameter of 5.0″. This indicates a reciprocal relationship between PIG speed and the bypass diameter.

Another important factor that affects the PIG speed is the pipeline inclination/slop. A sharp slope in the pipeline is reported to have a variable effect on the PIG speed. A major factor that affects the PIG speed is the differential pressure across the PIG. Differential pressure is the driving force that sets the PIG in motion. A lower differential pressure results in a lower PIG speed and vice versa. The overall pipeline pressure developed by the fluid flow also affects the PIG speed. Another factor is the geometry of the PIG. The PIG geometry affects the flow behaviour of the fluid across its boundaries. The fluid flow behaviour determines the differential pressure across the PIG that ultimately controls the PIG speed [1].

2 Controlling the PIG speed by Pressure Transducers and Hall Effect Sensor

The motion of a PIG mainly is said to be a function of pressure difference (ΔP), by which the fluid inside the pipeline drives the PIG in the direction of the positive pressure gradient. In this method of controlling the speed of PIG, pressure transducers are installed at the outer surface of the pipelines. The pressure transducers measure the relative pressure inside the pipelines. Many types of pressure transducers are used, among which the piezoelectric pressure sensors are most commonly used because of their simple construction and robustness.

A PLC is also installed to provide an interface between a computer and the pressure transducer in this procedure. The main function of this PLC is to convert the analogue signals into digital signals to retrieve the digital information via the attached computer. Multiple transducers are installed on the pipeline so that if one or a couple of them malfunction, the other units can be used to retrieve the desired digital response. The pigging operation starts, and the time taken by the PIG to move from one transducer to another, in addition to the distance between the two, is used to measure the speed of the PIG as follows.

To have a comparative analysis, the speed of the pig is also determined by the electronic board that receives the speed data from an odometer installed inside the back cover of the PIG. An odometer works on the principle of the Hall effect. The rotary motion of the spinning wheel of an odometer can be determined by fixing a magnet on the spinning surface. The pulses thus produced are read and digitally interpreted by a microcontroller. The speed data received from the supervisory (pressure transducers) and the odometer are compared, and a per cent error is reported.

A study reported a PIG speed of 0.43 m/s and 0.45 m/s while using the pressure transducers and an odometer, respectively, that produced an error of 4.44% within the acceptable limits. It merits mentioning that an odometer provides more reliable data than pressure transducers. However, an error of only 4.44% corroborates the significance of speed data acquired by transducers.  However, this technique of speed control is not applicable for PIGs that are not equipped with an odometer. Such PIGs include spherical, foam, and solid cast PIGs [3]. It can be concluded that the speed of PIG can be controlled by controlling the pressure gradient across it, and the corrective measures can be accomplished keeping in view the digital feedback received from the pressure transducers.

3 Controlling the PIG Speed by Artificial Neural Network (ANN) Technique

In this technique, the Artificial Neural Network (ANN) controls the speed considering the differential pressure across the PIG. The general procedure of PIG speed control by an odometer is sometimes unreliable due to the vulnerability of the odometer to failure and subsequent receipt of poor results. One of the reasons for odometer failure is the loss of its contact with the inner wall of the pipeline when it comes across a rough spot in a faulty pipeline. The ANN technique is reported to be more reliable and robust. The step-wise procedural details of the ANN method include:

1. Installation of the electronic board in the body of PIG
2. Launching the pig in the gas-filled pipeline
3. Retrieval of velocity data and pressure data respectively from the odometer and pressure transducers (installed on the outer pipeline surface) at the end of the pipeline after depressurization
4. Sorting out the retrieved data sets for use in ANN
5. Training and validation of the ANNs
6. Analysis of the results obtained.

The ANN works on the principle of training and validation of the neural network chosen. Based on the initial values, neurons’ bias and weight matrix are adjusted to receive a minimal error in the training step. The validation step follows this by applying the input signals different from those used in the training step. This is accomplished to establish that the trained neural network is well generalized.  A study reports using Nonlinear Autoregressive eXogenous (NARX) input and Multilayer Perceptron (MLP) neural networks to find the PIG velocity using pressure difference at various points along the length of the pipeline. The differential pressure is measured at 10 different locations, including the PIG launching and receiving points. The pressure is determined using manometers and by the pressure transducers at the rest of the 8 locations. The pressure transducers send the data signals to an LPC connected to a supervisory system to process the data into quantitative measurements. The data thus received is trained and validated using MATLAB® (or similar) by keying in the pressure data as an input and velocity data as an output. With MLP neural network, a mean square error (MSE) of 13.1% was observed in the validation step with input and output neurons one each and five neurons on the hidden layer. To decrease the MSE, the number of neurons in the hidden layer was increased to 6 (MSE = 5.0%) and then 7 to achieve an MSE of 3.6%. The procedure was repeated with NARX that produced more promising results with an MSE of just 1.9% compared to 3.6% in the case of MLP. The study indicates that ANN can be reliably employed to determine and control PIG speed, thus opening new vistas of development by the embodiment of an ANN in a PIG [4].

4 Controlling the PIG Speed Using QFT Method

The Quantitative Feedback Theory (QFT) is a robust control strategy used to control the PIG speed using a bypass flow line in a pipeline filled with fluid. DEPENDING ON THE SITUATION, the QFT control strategy sends signals to the bypass valve installed inside the PIG body to open or close. The opening and closing of the bypass valve manipulate the fluid flux across the PIG that consequently moves in a controlled manner. This technique includes the conversion of the nonlinear dynamic equation of PIG motion to a group of uncertain linear systems using Sobhani-Rafeeyan’s (SR) method followed by the synthesis of a QFT controller for nonlinear systems. This method is valid for two-dimensional pipelines. The numerical simulation of both the controllers is accomplished using MATLAB (Simulink tool). The study indicates that PIG reaches its desired speed very quickly. This study provides the basis for developing a PIG speed controller when natural gas flows through the pipeline [5].

5 Controlling the PIG Speed Using Bypass Flow

A PIG’s dynamic behaviour depends on the prevailing differential pressure across its ends and the fluid flux through the bypass valve. The overall dynamics of a pigged pipeline depends on the behaviour of:

1. Fluid flowing inside the pipeline
2. Expelled gas on the front side of PIG, the PIG speed, and the bypass flow.

In this method, a nonlinear controller is synthesized based on PIG speed, the velocity of bypass flow across this PIG, and the PIG position. The simulation is performed on three positions: the PIG launching point, the PIG receiver point, and the point of PIG motion resumption after being jammed at a certain position in the pipeline. The simulation study reports that the PIG speed is well controlled such that the PIG follows the reference velocity without oscillation, particularly when the natural gas pipeline is pigged. The first simulation is accomplished taking into account the static friction force that inhibits the initial motion of the PIG at the launching position. The second simulation is carried out based on the assumption that the PIG stops at 1/3^rd length of the pipeline due to some reason, for example, the drag applied by accumulated paraffin. When the PIG stops, the differential pressure across it escalates, and a point is reached where the differential pressure overcomes the static resistance. The PIG starts moving at a high and uncontrollable speed. At this stage, the controller opens the bypass valve accordingly to bring the PIG velocity to reference value [6].

6 Maximum Speed of PIG Determined by Response Surface Methodology (RSM)

A PIG’s maximum speed in a pipeline is when it is made to resume its motion after going through a blockade in the pipeline. However, as aforementioned, this speed is spontaneous and often uncontrollable, damaging the pipeline and the PIG itself. An unsteady fluid flow persists in the pipeline, particularly after the PIG stoppage midway through the pipeline. In this method, the unsteady flow equations are solved by the Method of Characteristics (MOC), and the MOC results determine the pressure gradient across the PIG. The dynamic behaviour of PIG is studied by the Runge Kutta method, and simulation is performed for the PIG motion resumption after the blockade. The differential pressure, responsible for the PIG release from obstruction, was created by three different methods:

1. By increasing the pressure on the tail side of the PIG
2. by decreasing the pressure on the nose side of the PIG
3. By a combination of 1 and 2.

The study reveals that the maximum PIG velocity achieved by pressurizing the tail side of PIG is approximately the same as the velocity achieved by depressurizing on the nose side. For the 3^rd case, the maximum speed achieved by the combination of tail side pressurization and nose side depressurization is found to be in between the maximum velocity achieved in the individual cases of tail side pressurization and nose side depressurization. Therefore, the average of these three as received by the RSM simulation is represented in the form of an empirical correlation to determine the maximum speed of the PIG after its motion resumption:

Where _Vmax is the PIG maximum speed, p is the gas pressure, and F_obs is the differential pressure across the pipeline caused by the PIG obstruction. The simulation study reveals that the maximum speed of a PIG after the resumption of its motion is a function of pipeline pressure and the pressure gradient across the PIG that escalates due to obstruction. The Response Surface Methodology is employed to derive an empirical correlation to determine the maximum speed of the PIG post its release from the obstruction [7].

7 Commercially Available Technology for PIG Speed Control

Pipelines2Data, UK (P2D) and Inline Services (USA) are two main players for the design and manufacturing of the cleaning and speed control smart PIGs to improve the pipelines’ integrity and seamless transport of the fluids. Studies indicate that pigging at a speed of 8 mph results in ineffective cleaning and fluid transport. However, a velocity greater than 10 mph is observed to bypass the debris and, therefore, proved ineffective. The Inline Services (USA) has introduced integrated Speed Control and Cleaning Pig (SCP) tools with a diversified size range of 30”—48”. The SCP technology is designed for optimal cleaning at a 6—10 mph speed for a gas velocity of 33 mph. The onsite retrieval of pigging data is an added advantage in this technology. In addition, the technology has proven to be effective for pipelines with 1.5D bends too.

From P2D (UK), the Speed Control Tool (SCT) is manufactured and marketed. The SCT monitors the pipeline environment constantly and maintains the speed by adjusting the bypass through the PIG as and when required. In addition to pipeline cleaning and PIG speed control, the SCT records additional information such as pipeline pressure, temperature, differential pressure, and 6 axis acceleration downloaded after the pigging operation and analyzed for pipeline integrity and streamline operations [8].

8 References

1. Talbizadeh, A., & Keshtkar, M. M. (2020). Numerical and experimental study on a bypass pig motion in oil transmission pipeline: a case study. Journal of Petroleum Exploration and Production Technology, 10(7), 3007-3023.
2. Nguyen, T. T., Kim, S. B., Yoo, H. R., & Rho, Y. W. (2001). Modeling and simulation for pig flow control in natural gas pipeline. KSME International Journal, 15(8), 1165-1173.
3. Lima, G. F., Freitas, V. C., Araújo, R. P., Maitelli, A. L., & Salazar, A. O. (2017). PIG’s speed estimated with pressure transducers and hall effect sensor: An industrial application of sensors to validate a testing laboratory. Sensors, 17(9), 2119.
4. De Araújo, R. P., De Freitas, V. C. G., De Lima, G. F., Salazar, A. O., Neto, A. D. D., & Maitelli, A. L. (2018). Pipeline Inspection Gauge’s Velocity Simulation Based on Pressure Differential Using Artificial Neural Networks. Sensors, 18(9), 3072.
5. Mirshamsi, M., & Rafeeyan, M. (2012). Speed control of pipeline pig using the QFT method. Oil & Gas Science and Technology–Revue d’IFP Energies nouvelles, 67(4), 693-701.
6. Nguyen, T. T., Yoo, H. R., Rho, Y. W., & Kim, S. B. (2001, June). Speed control of PIG using bypass flow in natural gas pipeline. In ISIE 2001. 2001 IEEE International Symposium on Industrial Electronics Proceedings (Cat. No. 01TH8570)(Vol. 2, pp. 863-868). IEEE.
7. He, H., & Liang, Z. (2019). Speed simulation of pig restarting from stoppage in gas pipeline. Mathematical Problems in Engineering, 2019.
8. Money, N., Cockfield, D., Mayo, S., & Smith, G. (2012). Dynamic speed control in high velocity pipelines. Pipeline Gas J, 239(8), 30-38.