Method for predicting natural gas flow in fluctuating pipe of multiphase flow based on dichotomy

Abstract

The present application relates to a kind of based on dichotomous method's multi-phase flow fluctuation pipeline natural gas flow prediction method, constructs the data set of upper inclined section pressure drop model and lower inclined section pressure drop model;The parameters of this data set are: natural gas flow, natural gas water content, inclination, pressure, temperature, pipe diameter and pressure drop;The purpose fluctuation pipeline is divided into the combination of upper inclined section and lower inclined section, the length and inclination of each upper inclined section / lower inclined section are calculated, and the actual total pressure drop of the purpose fluctuation pipeline;The natural gas flow range value of preset purpose fluctuation pipeline is obtained, and the total pressure drop range value of the purpose fluctuation pipeline is obtained;Judge actual natural gas flow range in the natural gas flow range of preset purpose fluctuation pipeline, then natural gas flow is predicted by dichotomous method;Otherwise, then adjust the natural gas flow range value of preset purpose fluctuation pipeline, calculate the total pressure drop range value of its corresponding purpose fluctuation pipeline, and then natural gas flow is predicted by dichotomous method.The present application method is simple.

Description

Technical Field

[0001] This invention relates to the field of moisture metering technology, and more specifically to the field of natural gas flow metering technology for fluctuating gathering and transmission pipelines. Background Technology

[0002] Currently, online moisture metering for two-phase flow without gas-liquid separation mainly involves solving calculations using mathematical models. There are three main types: First, microwave or radio frequency measurements of water cut are used to establish a virtual high or low model to correct the gas phase flow rate and solve for the gas-liquid two-phase flow rate. Second, two flow meters based on different principles are connected in series to establish a virtual high or low combined metering model for online metering. This includes various combinations such as differential pressure + differential pressure series and differential pressure + non-differential pressure series. Some researchers cleverly combine throttling elements or non-differential pressure elements with different structures to extract both frequency and pressure drop signals, thereby establishing a combined metering model to solve for the gas-liquid two-phase flow rate. Third, regression algorithms such as neural networks and support vector machines are used to establish a nonlinear mapping between the flow rate of each phase and characteristic parameters, including parameters related to the gas wellbore, nozzle, and wellhead piping, thereby achieving the purpose of virtual measurement of the gas-liquid two phases.

[0003] All of the above gas-liquid two-phase testing methods require the installation of throttling elements or other components at the wellhead. Pressure, temperature, frequency, and other signals are measured using throttling or other equipment, and the gas-liquid two-phase flow rate is calculated using a pre-set gas-liquid two-phase metering model. However, in actual production, to save on gas field operating costs and reduce investment, the wellhead process has been gradually simplified, and even the installation of wellhead flow meters has been eliminated. For a given gas well, the available parameters mainly include wellhead pressure, wellhead temperature, pipeline pressure drop, pipeline routing, and average water production in the well area. Based on this, calculating the gas well's production becomes particularly important. Summary of the Invention

[0004] The present invention aims to address the above-mentioned problems by proposing a method for predicting the flow rate of multiphase flow fluctuating pipeline natural gas based on the bisection method.

[0005] The technical solution of this invention is as follows: A method for predicting the flow rate of multiphase flow fluctuating natural gas in pipelines based on the bisection method is as follows: Data sets were constructed for the pressure drop model of the upslope section and the pressure drop model of the downslope section; the parameters of the data sets are: natural gas flow rate, natural gas water content, inclination angle, pressure, temperature, pipe diameter and pressure drop; The target undulating pipeline is divided into combinations of upward-sloping and downward-sloping sections, and the length of each upward-sloping / downward-sloping section is calculated. li and tilt angle θ i And the actual total pressure drop Δ of the target fluctuating pipeline. p 0; Preset target natural gas flow range value for fluctuating pipelines [ q l , q h ];in, q l The low value of natural gas flow prediction for the target fluctuating pipeline. q h The high value of the natural gas flow forecast for the target fluctuating pipeline is used as a basis; based on this, the total pressure drop range of the target fluctuating pipeline is obtained by calling the dataset [△]. p l ,△ p h ]; where △ p l for q l The corresponding pressure drop value, Δ p h for q h The corresponding pressure drop value; If (△ p l -△ p 0)×(△ p h -△ p If 0) < 0, it indicates that the actual natural gas flow rate is within the range of the natural gas flow rate of the pre-defined target fluctuating pipeline; therefore, the natural gas flow rate is predicted using the dichotomy method. If (△ p l -△ p 0)×(△ p h -△ p If 0)≥0, it indicates that the actual natural gas flow range is outside the preset target range for natural gas flow in the fluctuating pipeline. q l , q h Within this range, adjust the preset target natural gas flow range value for the fluctuating pipeline. q l ’ , q h ’ ], calculate the total pressure drop range of the corresponding target fluctuating pipeline [△ p l ’ ,△ p h ’ ], until (△ p l ’ -△ p0)×(△ p h ’ -△ p 0) < 0; then predict the natural gas flow rate using the dichotomy method.

[0006] 1. (△) p l -△ p 0)×(△ p h -△ p When 0) < 0, the specific process of predicting natural gas flow using the dichotomy method is as follows: Take a natural gas flow rate median value q mid ’ , q mid ’ =( q l + q h ) / 2; Calculate the median value of natural gas flow rate q mid ’ The corresponding pressure drop value △ p ’ mid; if |△ p ’ mid-△ p If 0|≤0.001, then... q mid ’ As the actual natural gas flow rate; Conversely, judge (△) p ’ mid-△ p 0)×(△ p l -△ p The relationship between 0 and 0; if (△ p ’ mid-△ p 0)×(△ p l -△ p If 0) ≥ 0, then the natural gas flow range value of the target fluctuating pipeline is changed to [ q mid ’ , q h Conversely, the natural gas flow range value of the target fluctuating pipeline is changed to [ q l , q mid ’ ]; by[ qmid ’ , q h ]or[ q l , q mid ’ Recalculate the total pressure drop range of the corresponding target fluctuating pipeline until the final flow rate median value is reached. q mid The corresponding pressure drop value △ p mid Satisfy | △ p mid -△ p When 0|≤0.001, with q mid To predict natural gas flow.

[0007] 2. (△) p l -△ p 0)×(△ p h -△ p When 0)≥0, it indicates the range of natural gas flow rate for the preset target fluctuating pipeline. q l , q h If the value is too small, adjust the preset target natural gas flow range for the fluctuating pipeline: Determine △ p l -△ p 0 and △ p h -△ p The size relationship of 0; if △ p l -△ p 0>△ p h -△ p 0, take q h ’ =2 q h Then adjust the preset target fluctuating pipeline natural gas flow range value. q l ’ , q h ’ Specifically, [ q l ,2 q h Conversely, take q l ’ =0.5 ql Then adjust the preset target fluctuating pipeline natural gas flow range value. q l ’ , q h ’ Specifically, [0.5] q l , q h ]; The specific process of predicting natural gas flow using the dichotomy method is as follows: Take a natural gas flow rate median value q mid ’ , (△ p l -△ p 0)×(△ p h -△ p When 0)≥0, q mid ’ =( q l ’ + q h ’ ) / 2; Calculate the median value of natural gas flow rate q mid ’ The corresponding pressure drop value △ p ’ mid If |△ p ’ mid -△ p If 0|≤0.001, then... q mid ’ As the actual natural gas flow rate; Conversely, judge (△) p ’ mid-△ p 0)×(△ p l ’ -△ p The relationship between 0 and 0; if (△ p ’ mid-△ p 0)×(△ p l ’ -△ p If 0)≥0, then the natural gas flow range value of the target fluctuating pipeline is changed to [ q mid’ , q h ’ Conversely, the natural gas flow range value of the target fluctuating pipeline is changed to [ q l ’ , q mid ’ ]; by[ q mid ’ , q h ’ ]or[ q l ’ , q mid ’ Recalculate the total pressure drop range of the corresponding target fluctuating pipeline until the median flow rate is reached. q mid The corresponding pressure drop value △ p mid satisfies |△ p mid-△ p When 0|≤0.001, with q mid To predict natural gas flow.

[0008] The dataset is based on multiphase flow pipeline network simulation data, using the inclination angle as the dividing line, and establishes multiple upward and downward inclination sections. Natural gas flow rate, natural gas water content, upward / downward inclination angle, pressure, temperature, and pipe diameter are orthogonally designed. Through simulation calculations, the pressure drop of each upward / downward inclination section is extracted, thus forming a multiphase flow pipeline dataset containing both upward and downward inclination sections. The multiphase flow pipeline dataset is then standardized and substituted into a Gaussian process regression model for learning. The model with the highest correlation coefficient is used as the basis to finally determine the dataset for the upward and downward inclination pressure drop models established by the Gaussian process regression model.

[0009] The actual total pressure drop Δ of the target fluctuating pipeline p 0 is the starting pressure of the fluctuating pipeline. p 1. The pressure at the end of the undulating pipeline. p The difference of 2.

[0010] The target is to predict the minimum total pressure drop of the fluctuating pipeline. p l And the highest predicted total pressure drop △ p h The specific solution process is as follows: If the target fluctuating pipeline is the first i The angle of inclination of the segment θ i≥0, retrieve the data set of the pressure drop model for the upward-sloping section; if the target fluctuating pipeline is... i The angle of inclination of the segment θ i <0, retrieve the data set of the downward slope pressure drop model; calculate the target fluctuating pipeline first... i The pressure drop value of the section is updated, and the target fluctuating pipeline section is updated. i +1 stage pressure P i+1 and temperature T i+1 ;from i= 1 to i=n End, obtaining the desired fluctuating pipeline. n pressure drop value of segment P n ; For the first segment of the target undulating pipeline to the second segment of the target undulating pipeline n The pressure drop values ​​of each section are summed to calculate the predicted minimum total pressure drop Δ of the target fluctuating pipeline. p l And the highest predicted total pressure drop △ p h .

[0011] Among them, the target fluctuating pipeline first i +1 stage pressure P i+1 The calculation process is as follows: P i+1 = P 0-∑ P i In the formula: P 0 represents the pressure of the first section of the undulating pipeline, in MPa; Purpose of fluctuating pipeline i +1 segment temperature T i+1 The calculation is performed using the Sukhov temperature formula, as follows: T i+1 = T env +( T 0- T env )×e ﹣Kli ; K =π YOU / ( q m c p ); In the formula: T i+1 For the purpose of undulating pipelines, the first i +1 segment temperature, ℃;T env The ambient temperature is in °C. T 0 represents the temperature of the first section of the undulating pipe, in °C. K The temperature decay coefficient is dimensionless. li For the purpose of undulating pipelines, the first i +1 is the length of segment 1, in meters. D Let the inner diameter of the undulating pipe be in meters (m). U The overall heat transfer coefficient is W / (M). 2. K); q m The mass flow rate of natural gas is kg / s; c p Specific heat capacity of natural gas, J / (kg) . K).

[0012] The data sets of the upslope pressure drop model and the downslope pressure drop model established by the Gaussian process regression model are separately packaged and stored.

[0013] The natural gas flow range value of the preset target fluctuating pipeline [ q l , q h This estimate was obtained based on historical production data from gas wells.

[0014] The technical advantages of this invention are as follows: This invention proposes a method for predicting natural gas flow in multiphase flow undulating pipelines based on the bisection method. By dividing the target undulating pipeline along its inclination angle, and given parameters such as natural gas water content, inclination angle, pressure, and temperature, the method utilizes data sets from established pressure drop models for the upper and lower inclination sections, using the actual total pressure drop as the control variable, and predicts the natural gas flow using the bisection method. This method can estimate gas well production based on measurement data during gas field production without installing wellhead throttling elements, effectively saving investment and providing technical support for gas field production. Detailed Implementation

[0015] A method for predicting the flow rate of multiphase flow fluctuating natural gas in pipelines based on the bisection method is as follows: (1) Construct the datasets for the pressure drop model of the upslope section and the pressure drop model of the downslope section; Table 1. Data set for the pressure drop model of the upward slope section (excerpt) ; Table 2. Data set for the pressure drop model of the downsloping section (excerpt) .

[0016] (2) Taking the following undulating pipeline as an example, the total length is 2.89 km, and it is divided into the following 8 segments according to the inclination angle; the actual total pressure drop Δ of the undulating pipeline is as follows. p 0 is 850 kPa, the pressure of the first section of the undulating pipeline. P 0 is 6.25 MPa, temperature T 0 represents 20°, pipe diameter DN80, and water content of 0.6m³ of natural gas per 10,000 cubic meters. 3 The actual natural gas flow rate is 241 kg / h; the parameters of each section of the undulating pipeline are shown in Table 3. Table 3. Parameters of each section of the undulating pipeline. During the calculation process, each pressure drop segment is obtained by calling the data set, and the pressure and temperature of the next segment are updated synchronously after each pressure drop segment is obtained.

[0017] Preset the natural gas flow rate range for the target fluctuating pipeline [100 kg / h, 200 kg / h], and calculate the total pressure drop range [944.11 kPa, 880.97 kPa] corresponding to [100 kg / h, 200 kg / h]; △ p l -△ p 0 = 94.11 kPa, △ p h -△ p 0 = 30.97 kPa; (△) p l -△ p 0)×(△ p h -△ p If 0) > 0, it indicates that the preset target undulating pipeline natural gas flow range [100kg / h, 200kg / h] is too small; therefore, the preset target undulating pipeline natural gas flow range needs to be adjusted. The adjustment process is as follows: Determine △ p l -△ p 0 and △ p h -△ p The size relationship of 0; △ p l -△ p 0>△ p h -△ p 0, then q h ’ =2 q hThe natural gas flow range of the preset target fluctuating pipeline is adjusted to [100 kg / h, 400 kg / h]; the total pressure drop range corresponding to [100 kg / h, 400 kg / h] is calculated to be [944.11 kPa, 771.06 kPa]; at this time, (944.11-850)×(771.06-850)<0, indicating that the actual natural gas flow is within [100 kg / h, 400 kg / h]. Iteratively obtain intermediate flow values q mid The specific process is as follows: ; As shown above, the binary search method is used to calculate 17 times, until the median flow rate is reached. q mid When the pressure is 252.57 kg / h, the corresponding pressure drop value is Δ. p mid satisfies |△ p mid-△ p If 0|=0≤0.001, then the median flow rate will be used. q mid = 252.57 kg / h was used as the predicted natural gas flow rate; compared with the actual natural gas flow rate of 241 kg / h, the relative error is 4.8%.

[0018] The method provided by this invention is used to predict the natural gas flow rate of fluctuating pipelines for other purposes: .

[0019] Pipeline 1, total length 4.72km, actual total pressure drop Δ p 0 is 1550 kPa, the pressure of the first stage. P 0 is 7.9 MPa, temperature T 0 represents 20°, pipe diameter DN150, and water content of 0.6m³ of natural gas per 10,000 cubic meters. 3 The actual natural gas flow rate was 534 kg / h. The predicted natural gas flow rate obtained by the method provided in this invention was 561 kg / h, with a relative error of 5.06%.

[0020] Pipeline 2, 2.75km in length, actual total pressure drop Δ p 0 is 1395 kPa, the pressure of the first stage. P 0 is 7.8 MPa, temperature T 0 represents 20°, pipe diameter DN100, and water content of 0.6m³ of natural gas per 10,000 cubic meters. 3 The actual natural gas flow rate was 508 kg / h. The predicted natural gas flow rate obtained by the method provided in this invention was 471 kg / h, with a relative error of -7.28%.

[0021] Pipeline 3, total length 6.84km, actual total pressure drop Δ p 0 is 1990 kPa, the pressure in the first stage. P 0 is 9MPa, temperature T 0 represents 20°, pipe diameter DN100, and water content of 0.6m³ of natural gas per 10,000 cubic meters. 3 The actual natural gas flow rate was 1370 kg / h. The predicted natural gas flow rate obtained by the method provided in this invention was 1276 kg / h, with a relative error of -7.37%.

Claims

1. A method for predicting the flow rate of natural gas in a multiphase flow fluctuating pipeline based on the bisection method, characterized in that, The method is as follows: Data sets were constructed for the pressure drop model of the upslope section and the pressure drop model of the downslope section; the parameters of this data set are: natural gas flow rate, natural gas water content, inclination angle, pressure, temperature, pipe diameter, and pressure drop; The target undulating pipeline is divided into combinations of upward-sloping and downward-sloping sections, and the length of each upward-sloping and downward-sloping section is calculated. li and tilt angle θ i And the actual total pressure drop Δ of the target fluctuating pipeline. p 0; Preset target natural gas flow range value for fluctuating pipelines [ q l , q h ];in, q l The low value of natural gas flow prediction for the target fluctuating pipeline. q h The high value of the natural gas flow forecast for the target fluctuating pipeline is used as a basis; based on this, the total pressure drop range of the target fluctuating pipeline is obtained by calling the dataset [△]. p l ,△ p h ]; where △ p l for q l The corresponding pressure drop value, Δ p h for q h The corresponding pressure drop value; If (△ p l -△ p 0)×(△ p h -△ p If 0) < 0, it indicates that the actual natural gas flow rate is within the range of the natural gas flow rate of the pre-defined target fluctuating pipeline; therefore, the natural gas flow rate is predicted using the dichotomy method. If (△ p l -△ p 0)×(△ p h -△ p If 0)≥0, it indicates that the actual natural gas flow range is outside the preset target range for natural gas flow in the fluctuating pipeline. q l , q h Within this range, adjust the natural gas flow rate range value of the preset target fluctuating pipeline. q l ’ , q h ’ ], calculate the total pressure drop range of the corresponding target fluctuating pipeline [△ p l ’ ,△ p h ’ ], until (△ p l ’ -△ p 0)×(△ p h ’ -△ p 0) < 0; then predict the natural gas flow rate using the dichotomy method; The specific process of predicting natural gas flow using the dichotomy method is as follows: (△ p l -△ p 0)×(△ p h -△ p When 0) < 0, Take a natural gas flow rate median value q mid ’ , q mid ’ =( q l + q h ) / 2; Calculate the median value of natural gas flow rate q mid ’ The corresponding pressure drop value △ p ’ mid; if |△ p ’ mid-△ p If 0|≤0.001, then... q mid ’ As the actual natural gas flow rate; Conversely, judge (△) p ’ mid-△ p 0)×(△ p l -△ p The relationship between 0 and 0; if (△ p ’ mid-△ p 0)×(△ p l -△ p If 0) ≥ 0, then the natural gas flow range value of the target fluctuating pipeline is changed to [ q mid ’ , q h Conversely, the natural gas flow range value of the target fluctuating pipeline is changed to [ q l , q mid ’ ]; by[ q mid ’ , q h ]or[ q l , q mid ’ Recalculate the total pressure drop range of the corresponding target fluctuating pipeline until the final flow rate median value is reached. q mid The corresponding pressure drop value △ p mid Satisfy | △ p mid -△ p When 0|≤0.001, with q mid To predict natural gas flow.

2. The method for predicting the flow rate of multiphase flow fluctuating pipeline natural gas based on the bisection method according to claim 1, characterized in that, (△ p l -△ p 0)×(△ p h -△ p When 0)≥0, it indicates the range of natural gas flow rate for the preset target fluctuating pipeline. q l , q h If the value is too small, adjust the preset target natural gas flow range for the fluctuating pipeline: Determine △ p l -△ p 0 and △ p h -△ p The size relationship of 0; If △ p l -△ p 0>△ p h -△ p 0, take q h ’ =2 q h Then adjust the preset target fluctuating pipeline natural gas flow range value. q l ’ , q h ’ Specifically, [ q l ,2 q h Conversely, take q l ’ =0.5 q l Then adjust the preset target fluctuating pipeline natural gas flow range value. q l ’ , q h ’ Specifically, [0.5] q l , q h ].

3. The method for predicting the flow rate of multiphase flow fluctuating pipeline natural gas based on the bisection method according to claim 2, characterized in that, The specific process of predicting natural gas flow using the dichotomy method is as follows: Take a natural gas flow rate median value q mid ’ ; (△ p l -△ p 0)×(△ p h -△ p When 0)≥0, q mid ’ =( q l ’ + q h ’ ) / 2; Calculate the median value of natural gas flow q mid ’ The corresponding pressure drop value △ p ’ mid If |△ p ’ mid -△ p If 0|≤0.001, then... q mid ’ As the actual natural gas flow rate; Conversely, judge (△) p ’ mid-△ p 0)×(△ p l ’ -△ p The relationship between 0 and 0; If (△ p ’ mid-△ p 0)×(△ p l ’ -△ p If 0)≥0, then the natural gas flow range value of the target fluctuating pipeline is changed to [ q mid ’ , q h ’ Conversely, the natural gas flow range value of the target fluctuating pipeline is changed to [ q l ’ , q mid ’ ]; by[ q mid ’ , q h ’ ]or[ q l ’ , q mid ’ Recalculate the total pressure drop range of the corresponding target fluctuating pipeline until the median flow rate is reached. q mid The corresponding pressure drop value △ p mid satisfies |△ p mid-△ p When 0|≤0.001, with q mid To predict natural gas flow.

4. The method for predicting the flow rate of multiphase flow fluctuating pipeline natural gas based on the bisection method according to claim 1, characterized in that, The dataset is based on multiphase flow pipeline network simulation data and uses the inclination angle as the dividing basis to establish multiple upward and downward inclination sections. The natural gas flow rate, natural gas water content, upward inclination angle, downward inclination angle, pressure, temperature and pipe diameter are orthogonally designed. Through simulation calculation, the pressure drop of each upward and downward inclination section is extracted, thereby forming a multiphase flow pipeline dataset containing upward and downward inclination sections. The multiphase flow pipeline dataset is standardized and then substituted into the Gaussian process regression model for learning. The dataset with the highest correlation coefficient is used as the basis to finally determine the data sets of the pressure drop model for the upsloping section and the pressure drop model for the downsloping section established by the Gaussian process regression model.

5. The method for predicting the flow rate of multiphase flow fluctuating pipeline natural gas based on the bisection method according to claim 1, characterized in that, The actual total pressure drop Δ of the target fluctuating pipeline p 0 is the starting pressure of the fluctuating pipeline. p 1. The pressure at the end of the undulating pipeline. p The difference of 2.

6. The method for predicting the flow rate of multiphase flow fluctuating pipeline natural gas based on the bisection method according to claim 1, characterized in that, The predicted minimum total pressure drop value △ of the target fluctuating pipeline p l And the highest predicted total pressure drop △ p h The specific solution process is as follows: If the target fluctuating pipeline is the first i The angle of inclination of the segment θ i ≥0, retrieve the data set of the pressure drop model for the upward slope section; If the target fluctuating pipeline is the first i The angle of inclination of the segment θ i <0, retrieve the data set of the downslope pressure drop model; Calculation of the purpose of fluctuating pipeline i The pressure drop value of the section is updated, and the target fluctuating pipeline section is updated. i +1 stage pressure P i+1 and temperature T i+1 ;from i= 1 to i=n End, obtaining the desired fluctuating pipeline. n pressure drop value of segment P n ; For the first section of the target undulating pipeline to the second section of the target undulating pipeline n The pressure drop values ​​of each section are summed to calculate the predicted minimum total pressure drop Δ of the target fluctuating pipeline. p l And the highest predicted total pressure drop △ p h .

7. The method for predicting the flow rate of multiphase flow fluctuating pipeline natural gas based on the bisection method according to claim 6, characterized in that, Purpose of fluctuating pipeline i +1 stage pressure P i+1 The calculation process is as follows: P i+1 = P 0－∑ P i In the formula: P 0 represents the pressure in the first section of the undulating pipeline, in MPa. Purpose of fluctuating pipeline i +1 segment temperature T i+1 The calculation is performed using the Sukhov temperature formula, as follows: T i+1 = T env ＋( T 0－ T env )×e ﹣Kli ； K =π DU / ( q m c p )； In the formula: T i+1 For the purpose of undulating pipelines, the first i +1 segment temperature, ℃; T env The ambient temperature is in °C. T 0 represents the temperature of the first section of the undulating pipeline, in °C. K The temperature decay coefficient is dimensionless. li For the purpose of undulating pipelines, the first i +1 is the length of segment 1, in meters. D Let the inner diameter of the undulating pipe be in meters (m). U The overall heat transfer coefficient is W / (M).

2. K); q m The mass flow rate of natural gas is kg / s; c p Specific heat capacity of natural gas, J / (kg) . K).

8. The method for predicting the flow rate of multiphase flow fluctuating pipeline natural gas based on the bisection method according to claim 4, characterized in that, The data sets of the upslope pressure drop model and the downslope pressure drop model established by the Gaussian process regression model are packaged and saved separately.

9. The method for predicting the flow rate of multiphase flow fluctuating pipeline natural gas based on the bisection method according to claim 1, characterized in that, The natural gas flow range value of the preset target fluctuating pipeline [ q l , q h This estimate was obtained based on historical production data from gas wells.

Images