Abstract

Based on the fluid visualization loop device, design and process a simulation device for reducing drag of heavy oil mixed with gas. Conduct experimental research on the flow resistance characteristics of two types of heavy oil mixed with gas in a horizontal pipe, capture the pipe flow patterns under different gas-liquid flow ratios, analyze the drag reduction effect of gas relative to heavy oil under different experimental conditions, and establish corresponding pressure drop prediction models.The results showed that six flow patterns were observed within the gas-liquid ratio range of 0-15, namely bubbly flow, slug flow, stratified flow, slug flow, annular flow, and misty flow. The pipeline drag reduction rates corresponding to simulated oils 220 # and 440 # reached their maximum values of 48.19% and 33.76% at gas-liquid ratios of 1.17 and 0.96, respectively. When the blending ratio was between 0.9 and 1.2, the drag reduction rates could be maintained above 20%.The mechanism can be attributed to the fact that air transforms oil oil contact into oil gas oil contact, reducing the interlayer shear stress of the mixed phase. The Dukler method is not applicable to high viscosity gas-liquid two-phase flow. The predicted values of the heavy oil gas two-phase pressure drop model established are in good agreement with the measured values, with an average relative error of less than 20%.

With the continuous consumption of conventional crude oil, the development of unconventional crude oil has gradually become the focus of people's attention.As a type of unconventional crude oil with huge reserves, heavy oil is highly attractive to the oil and gas industry, but its high density and viscosity pose significant challenges in production and transportation processes.As viscosity is an important indicator for measuring the flowability of heavy oil, one of the important research topics for oil and gas workers today is how to economically, safely, and effectively reduce its apparent viscosity.At present, there are many conventional methods to improve the flowability of heavy oil, which are widely and maturely applied in oilfield sites. Although these methods can effectively reduce its viscosity, they generally have problems such as uneconomical and environmentally unfriendly. For example, using heating to reduce crude oil viscosity, improve crude oil flowability, reduce pipeline friction loss, high energy consumption, poor adaptability to low throughput conditions, complex processes and operations, and large investment.Adding diluents or oil soluble viscosity reducers can reduce the concentration of wax or asphaltene in the mixed oil, weaken the interaction between asphaltene, but with a large dilution amount, it is often difficult to implement due to the lack of dilute oil. At the same time, adding a large amount of chemical agents can corrode pipelines and equipment, increasing maintenance costs.Emulsifiers can evenly disperse heavy oil in water to form O/W emulsions, thereby reducing viscosity and drag. However, this method requires high stability of the emulsion, which is not conducive to oil dehydration and results in high processing loads.In response to the shortcomings of the above methods, many scholars have proposed a new method of reducing pipeline friction by blending heavy oil with gas. Research has shown that heavy oil can effectively reduce pipeline pressure drop at an appropriate blending ratio. The gas source can be both oilfield associated gas and inert gases such as nitrogen. This method has the characteristics of wide gas source, convenient preparation, flexible operation, safety and environmental protection, and simple separation.

At present, research on gas-liquid two-phase flow patterns is relatively in-depth, and the widely recognized flow patterns are divided into bubbly flow, slug flow, stratified flow, slug flow, annular flow, and misty flow.In addition, many scholars have also studied the factors and trends of flow pattern changes in the transition zone of horizontal pipes, and proposed flow pattern prediction models.However, there is little research on the gas-liquid two-phase flow patterns of high viscosity liquids in horizontal pipelines, and the applicability of related studies to heavy oil still needs to be verified.Chawla established a pressure drop prediction model for gas-liquid two-phase flow based on the principle of momentum exchange between the two phases and between the two phases and the pipe wall, but this model is only applicable to low viscosity liquids.The pressure drop prediction model established by Beggs et al. through experimental research provides good predictions for gas-liquid two-phase flow, but the predicted values deviate significantly from the actual values as the viscosity of the liquid phase increases.Foletti et al. validated the Petalas and Orell models through loop experiments, and the results showed that the prediction accuracy of the low flow rate model was high, but gradually decreased with increasing flow rate.Brito et al. proposed a typical pressure drop prediction model and used flow pattern recognition method as the basis to predict pressure changes. However, this model also has low accuracy in predicting pressure drop for gas-liquid two-phase flow of high viscosity liquids.From this, it can be seen that although scholars at home and abroad have proposed many gas-liquid two-phase flow pressure drop prediction models, they are generally only applicable to low viscosity liquids and have significant limitations. Therefore, a gas-liquid two-phase flow pressure drop model suitable for high viscosity liquids such as heavy oil needs to be established.

Therefore, this article intends to use two relatively viscous white oils as simulated oils, design a loop simulation device, simulate and study the flow pattern changes and pipe flow characteristics of heavy oil mixed with gas in horizontal pipes, explore the drag reduction mechanism of heavy oil under gas mixing conditions, and establish a gas-liquid two-phase flow pressure drop prediction model suitable for heavy oil.

1. Experimental Materials and Methods

1.1 Materials

Due to the dark black color of ordinary heavy oil, it is not easy to observe the flow pattern. Therefore, 220# and 440# white oil were selected as simulated heavy oil, and then heavy oil gas mixing loop experiments were carried out to explore the flow pattern changes and resistance characteristics of gas heavy oil.The experimental results showed that the densities of 220# and 440# white oil at 20℃ were 843.5 kg·m^-3 and 870.3 kg·m^-3, respectively, and the viscosities at 50℃ were 108.86 mPa·s and 234.06 mPa·s, respectively, both belonging to the category of ordinary heavy oil.Using the Anton Paar Rheolab QC rheometer measurement system, the rheological and viscosity temperature characteristics of 220# and 440# white oils were tested in the range of 10-60℃ at a shear rate of 0100 s^-1, as shown in Figures 1 and 2.As shown in the figure, the rheological curves of the two types of white oil at different temperatures both pass through the origin of the Cartesian coordinate system of shear rate and shear stress, exhibiting Newtonian fluid characteristics.In the testing temperature range, the viscosity of both types of white oil decreases with increasing temperature, and the viscosity decreases significantly with increasing temperature at 10-60℃, which is very similar to the viscosity temperature characteristics of ordinary heavy oil.The experiment was conducted at 20℃, and a simulated oil of ordinary heavy oil was selected based on the viscosity at 20℃. The viscosity of 220# and 440# white oil at 20℃ was 949 and 2286mPa·s, respectively.

1.2 Experimental Equipments

The heavy oil blending transportation simulation device mainly consists of an oil and gas pipeline system, a high-speed photography system, and a data acquisition system, as shown in Figure 3. The material used in the experiment is an organic glass tube, with a total length of 3.7 meters. The test section is 1.7 meters long and has an inner diameter of 24 mm. An absolute pressure sensor is used to collect the pressure at both ends of the test section.

The oil circuit mainly includes oil storage tanks, oil transfer pumps, turbine flow meters, ball valves, etc. After the white oil in the oil storage tank is measured by LWYC turbine flow meters (Bosheng Instrument Co., Ltd., Dalian), it is pumped by YX3-112H-4 high-temperature gear pump (Binhai Pump Manufacturing Co., Ltd., Botou) to the inlet of the Y-shaped oil-gas mixer and mixed with the gas phase;The air circuit is mainly composed of a compressor, an air storage tank, a pressure regulating valve, a glass rotor flowmeter, etc. Compressed air enters the pipeline through a storage tank and is sent to the mixer. The gas pressure is controlled by the pressure regulating valve, and its flow rate is measured by the LZB-10 glass rotor flowmeter.The high-speed photography system used in the experiment is the 2F04M Thousand Eyed Wolf high-speed camera (Fuhuang Junda High tech Information Technology Co., Ltd., Hefei), with a collection period of 10000 μs, exposure time of 300 μs, and a storage length of 5 frames. The flow pattern of oil and gas two-phase flow in the mixed pipeline is observed and recorded.The oil and gas flow rates in the pipeline, as well as the absolute pressure in the testing section, are collected in real-time by the CX-DK-PLC-001 data acquisition system.

1.3 Experimental Methods

Based on the recommended flow rate corresponding to the viscosity of white oil and the pipe diameter, and according to the maximum flow rate that the oil pump can provide, the flow ranges of 220 # and 440 # white oil are determined to be 20-90 ml · s^-1 and 10-30 ml · s^-1, respectively, with a blending ratio of 0-15.

Firstly, collect the absolute pressure of single-phase oil flow in the test section at different flow rates. Secondly, measure the absolute pressure of the pipe flow under the condition of aeration.The specific steps of the experiment are: ① Turn on the oil pump to stabilize its flow rate at the test value; ② Turn on the air compressor to achieve a gas pressure of 0.8 MPa, inject the gas into the mixture with white oil, and control the flow rate through a pressure regulating valve and ball valve; ③ Collect parameters such as absolute pressure transmitter pressure, air and white oil flow rate after stable flow; ④ Use a high-speed camera to record the characteristics of flow pattern changes.

1.4 Data Processing Methods

To study the variation of differential pressure at the testing end with the viscosity of heavy oil and gas-liquid ratio, the relative drag reduction rate Δ p of stable flow of heavy oil air is introduced, and its specific calculation formula is

Due to the influence of experimental conditions and the instrument itself, there is uncertainty in the data obtained by the loop device. In the experimental process, the impact of systematic errors on the drag reduction rate should be minimized as much as possible. For some "outliers" in the drag reduction rate caused by accidental errors that deviate from the normal data variation range, they should be removed by the degree of deviation from the mean variance or standard variance of the data.

2. Results and Discussion

2.1 Analysis of Single-phase Flow Resistance Characteristics

Since both 220 # white oil and 440 # white oil belong to Newtonian fluids, the friction resistance coefficients of the two simulated oils under different flow conditions can be calculated according to the conventional basic hydraulic equation [Equation (4)] (ignoring local friction losses), as shown in Figure 4. As the flow rate of heavy oil increases, the friction resistance coefficient decreases sharply and then tends to flatten out, and the higher the viscosity of the oil, the greater the friction resistance coefficient.