CN116595881A - Method for predicting temperature field of crude oil storage tank and evaluating oil storage safety - Google Patents
Method for predicting temperature field of crude oil storage tank and evaluating oil storage safety Download PDFInfo
- Publication number
- CN116595881A CN116595881A CN202310581716.4A CN202310581716A CN116595881A CN 116595881 A CN116595881 A CN 116595881A CN 202310581716 A CN202310581716 A CN 202310581716A CN 116595881 A CN116595881 A CN 116595881A
- Authority
- CN
- China
- Prior art keywords
- storage tank
- temperature
- data
- crude oil
- neural network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003860 storage Methods 0.000 title claims abstract description 278
- 239000010779 crude oil Substances 0.000 title claims abstract description 103
- 239000003921 oil Substances 0.000 title claims abstract description 76
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000004088 simulation Methods 0.000 claims abstract description 82
- 238000004519 manufacturing process Methods 0.000 claims abstract description 67
- 239000002245 particle Substances 0.000 claims abstract description 63
- 238000001879 gelation Methods 0.000 claims abstract description 61
- 238000003062 neural network model Methods 0.000 claims abstract description 61
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 230000008859 change Effects 0.000 claims abstract description 15
- 238000012502 risk assessment Methods 0.000 claims abstract description 8
- 238000012216 screening Methods 0.000 claims abstract description 4
- 238000002474 experimental method Methods 0.000 claims description 77
- 238000009826 distribution Methods 0.000 claims description 27
- 239000011159 matrix material Substances 0.000 claims description 24
- 238000010438 heat treatment Methods 0.000 claims description 23
- 238000007667 floating Methods 0.000 claims description 22
- 239000007788 liquid Substances 0.000 claims description 22
- 239000013078 crystal Substances 0.000 claims description 20
- 238000013528 artificial neural network Methods 0.000 claims description 19
- 230000001419 dependent effect Effects 0.000 claims description 18
- 238000005457 optimization Methods 0.000 claims description 17
- 230000015572 biosynthetic process Effects 0.000 claims description 16
- 238000005259 measurement Methods 0.000 claims description 13
- 238000000917 particle-image velocimetry Methods 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 12
- 239000010802 sludge Substances 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000011156 evaluation Methods 0.000 claims description 11
- 230000009467 reduction Effects 0.000 claims description 11
- 238000013507 mapping Methods 0.000 claims description 10
- 238000012360 testing method Methods 0.000 claims description 9
- 230000007613 environmental effect Effects 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 7
- 238000009833 condensation Methods 0.000 claims description 6
- 230000005494 condensation Effects 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000001556 precipitation Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 5
- 238000009825 accumulation Methods 0.000 claims description 4
- 230000006399 behavior Effects 0.000 claims description 4
- 230000000704 physical effect Effects 0.000 claims description 4
- 230000001133 acceleration Effects 0.000 claims description 3
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 239000000700 radioactive tracer Substances 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 2
- 238000011160 research Methods 0.000 claims description 2
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 239000000126 substance Substances 0.000 claims description 2
- 239000012780 transparent material Substances 0.000 claims description 2
- 239000000499 gel Substances 0.000 description 18
- 239000003208 petroleum Substances 0.000 description 7
- 238000000513 principal component analysis Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 3
- UFWIBTONFRDIAS-UHFFFAOYSA-N Naphthalene Chemical compound C1=CC=CC2=CC=CC=C21 UFWIBTONFRDIAS-UHFFFAOYSA-N 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 239000000084 colloidal system Substances 0.000 description 2
- 230000000295 complement effect Effects 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000010355 oscillation Effects 0.000 description 2
- 229930195734 saturated hydrocarbon Natural products 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 238000009529 body temperature measurement Methods 0.000 description 1
- 239000009096 changqing Substances 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000012530 fluid Substances 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000007786 learning performance Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000012847 principal component analysis method Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computer Hardware Design (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Geometry (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a method for predicting a temperature field of a crude oil storage tank and evaluating oil storage safety, which comprises the following steps: according to factors influencing the change of the temperature field of the storage tank in the actual production process, screening and optimizing the storage tank to determine a characteristic element group; acquiring actual storage tank temperature field data under a characteristic element group, forming a data set by the characteristic elements and the storage tank internal temperature data, and establishing a BP neural network model I; acquiring a numerical simulation and an indoor experimental data set, respectively establishing BP neural network models II and III, and optimizing an initial weight and a threshold by adopting a particle swarm algorithm; predicting a temperature field and correcting parameters; establishing a BP neural network model IV of the double-sample data, and optimizing a weight and a threshold by adopting a particle swarm algorithm; carrying out temperature field prediction and correcting parameters to obtain an optimal temperature field prediction model; the risk assessment is carried out on different areas of the storage tank, and the gelation risk level of the storage tank at the current moment and the future moment is assessed according to the temperature field prediction result.
Description
Technical field:
the invention relates to the technical field of prediction of temperature field distribution in a crude oil storage tank and identification of storage tank gelation risk, in particular to a method for predicting the temperature field of the crude oil storage tank and evaluating oil storage safety.
The background technology is as follows:
petroleum is regarded as blood of modern industry, and is of self-evident importance for national development. With the rapid development of national economy, the demand for petroleum is increasing. The country is a large country for producing oil, including Changqing oil field and Daqing oil field, but is also a large country for import of petroleum. Petroleum is taken as non-renewable energy, is a strategic reserve material which cannot be extracted endlessly, and in order to meet the requirements of domestic petroleum, the national energy safety is ensured, and a series of measures are implemented to build a strategic reserve system of petroleum.
As national crude oil strategic reserves increase, so does the number of oil depot tanks. In order to cope with the fluctuation of the petroleum sales markets at home and abroad, the storage time of the oil storage tank is prolonged, and the energy consumption of the storage tank is increased. To prevent the crude oil from gelling, and cause great economic loss, the temperature inside the storage tank is kept to a certain limit, and the temperature inside the storage tank is monitored. Because the large-scale vertical storage tank has large structural size, the oil temperature distribution in the storage tank is uneven, and the accurate characterization of the temperature distribution in the storage tank is difficult to carry out by simply relying on a conventional single-point temperature measurement method. By adopting the multipoint temperature monitoring system, the comprehensive temperature distribution data in the tank can be obtained, but the number of the collection points and the distribution positions determine the accuracy of the temperature distribution to a certain extent.
The invention comprises the following steps:
the invention aims to provide a method for predicting the temperature field of a crude oil storage tank and evaluating the safety of oil storage, which predicts the temperature field of the storage tank based on the existing data, ensures the safety of the storage tank and reduces the running cost.
The technical scheme adopted for solving the technical problems is as follows: the method for predicting the temperature field of the crude oil storage tank and evaluating the oil storage safety comprises the following steps:
s1, acquiring factors and corresponding data influencing an internal temperature field of an oil storage tank in field production, wherein the factors and the corresponding data comprise: production site environment temperature, surface temperature, pressure, storage tank heating mode, heating time, crude oil storage time length, initial temperature, liquid level, crude oil density and crude oil viscosity; measuring temperatures at different positions and at different moments in the storage tank by using a temperature measuring rod, recording temperature data, taking the temperature in the storage tank as an index of the performance of a temperature field of the oil storage tank, taking the temperature data obtained by the temperature measuring rod as a dependent variable and corresponding influence factor data as independent variables, forming a data set together, and preprocessing the data;
s2, screening and optimizing the preprocessed data set by adopting a dimension reduction algorithm, removing repeated or relatively high-correlation data, and determining that characteristic elements affecting the internal temperature of the oil storage tank are combined into factors affecting the internal temperature field of the oil storage tank in screened field production, specifically the environmental temperature, the surface temperature, the crude oil storage duration, the initial temperature and the liquid level of the production field;
S3, based on the independent variables and the dependent variables after dimension reduction, a data set is formed together, a BP neural network model I is established, a storage tank temperature influence model in actual production is described, and the model is continuously enriched through continuous field data acquisition, so that the model truly represents the storage tank temperature change condition in actual production;
s4, based on the S2, performing numerical simulation and an indoor storage tank experiment respectively by taking the data of the optimized characteristic element group as a numerical simulation parameter and an experimental parameter to obtain a simulation data set and an indoor storage tank experiment data set;
s5, respectively carrying out normalization processing on the simulation data set and the indoor storage tank experimental data set;
s6, establishing a BP neural network model II by using the normalized simulation data in the S5, wherein the BP neural network model II is a mapping relation between independent variables and a data set of dependent variables in the numerical simulation process; establishing a BP neural network model III by using the indoor storage tank experimental data set normalized in the step S5; respectively optimizing the initial weight and the threshold of the constructed BP neural network model II and BP neural network model III by adopting a particle swarm optimization algorithm;
s7, training the BP neural network model II and the BP neural network model III optimized by the S6 for a plurality of times, and respectively comparing the obtained results with the measurement result of the S1 actual storage tank until the error of the two results is less than 10%; otherwise, modifying the simulation parameter value of the BP neural network model II through numerical simulation; indoor experiment corrects the indoor simulation parameter value of BP neural network model III;
S8, utilizing the data results obtained by the S7 final experiment and numerical simulation to establish a BP neural network model IV of double-sample data, training the BP neural network model IV for multiple times, and comparing the obtained results with the actual storage tank measurement results until the errors of the two are less than 10%; otherwise, correcting experiment and numerical simulation parameters, repeating the experiment or the numerical simulation to obtain an optimal prediction model, namely a temperature field prediction model, and predicting the temperature field in the storage tank under different conditions by using the temperature field prediction model;
s9, based on the temperature field prediction model of S8, and combining an indoor crude oil rheological experiment, evaluating and predicting the gelation risk level of crude oil in the storage tank in actual production.
The dimension reduction algorithm in the above scheme S2 is:
step one: calculating a covariance matrix Z from the data set:
for a set of samples X, there are m observations X 1 ,x 2 ,……,x m A total of n groups,
(1) The average value for each column is calculated:
(2) Variance of each column:
(3) And (3) carrying out standardized processing on the data:
obtaining a matrix:
(4) Calculating a correlation coefficient:
obtaining a correlation coefficient matrix R:
r ij is the correlation between the samples of the ith row and the samples of the jth column of the matrix X, and has a value of [ -1,1]And the R matrix should be a symmetric matrix, when R >At 0, positive linear correlation; r=0, linear independent; when r is<0, is a linear negative correlation;
step two: calculating the characteristic value of R and the corresponding characteristic vector
The covariance matrix is a real symmetric matrix, the eigenvalue of which is known to be non-negative, the eigenvalue of which is set,
λ 1 ≥λ 2 ≥λ 3 ≥…≥λ p and (3) not less than 0, and the unit eigenvectors after orthogonalization corresponding to the unit eigenvectors are as follows:
if the index variable represented by each column of original X is a synthesized vector, the synthesized vector is expressed as Var= [ Var ] 1 ,Var 2 ,…,Var p ] T
The ith principal component with X is then:
F i =(a i ) T Var=a 1i *Var 1 +a 2i *Var 2 +…+a pi *Var p
step three: calculating the contribution rate and the accumulated contribution rate of each component
The selection of the number of the main components is determined according to the accumulated contribution rate, the accumulated contribution rate reaches more than 85%, and the main component measurement standard is reserved:
a. the contribution rate of the reserved component variance reaches more than 85 percent;
b. the variance of the principal component is kept to be larger than 1;
and displaying the characteristic values in a graphical mode by adopting a lithotriptic graph, and reserving the number of main components corresponding to the position with the largest change in the graph.
In the above scheme S6, the method for optimizing the initial weight and the threshold of the BP neural network by using the particle swarm optimization algorithm is as follows:
dimension D of the entire search space, the location of the ith particle is given by:
wherein: omega 11 ……ω nl Is W 1 Is an element of (2); omega 1m ……ω lm Is W 2 Is an element of (2);for theta 1 Elements of (2):for theta 2 Is an element of (2); d is the dimension of the search space;
(1) Generating a population, randomly initializing, setting a maximum speed V for each particle max Corresponds to an initial position maximum X max A minimum velocity V is set for each particle min Corresponds to a minimum value X of an initial position min At [ -X max ,X max ]、[-V max ,V max ]Randomly selecting the position and the speed of particles in a section to initialize, wherein each particle in a D-dimensional space in the population is represented by a set of weight and threshold values of BP neural network;
(2) Setting a termination condition, iteration times N, population scale M, an initial value w of inertia weight and initial values C1 and C2 of acceleration factors of PSO optimization;
(3) Obtaining a fitness value F (i) of each particle by a fitness function according to the initial position of each particle:
wherein: n is the number of output nodes; y is i A desired output for an ith node of the neural network; o (o) i A prediction output for the i-th node; k is a coefficient;
(4) Extremum gbest of each particle population i Setting the particle size as the optimal value of the objective function, calculating the fitness value of each particle, and if F (i)>pbest i F (i) replaces pbest i Optimizing population extremum gbest i In each iterationUpdating the velocity and position of the particles according to:
v i =v i +c 1 ×rand()×(pbest i -x i )+c 2 ×rand()×(gbest i -x i )
x i =x i +v i
in the formula, i=1, 2,3 … …, N is the total number of particles in the population;
v i : particle velocity; rand (): a random number between (0, 1); x is x i : the current position of the particle; c 1 And c 2 : is a learning factor, c 1 =c 2 =2;v i Has a maximum value of V max (greater than 0), if v i Greater than V max V is then i =V max ;
(5) And (3) terminating PSO optimization to obtain the optimal weight and threshold of the BP neural network.
The method for establishing the BP neural network model II and the BP neural network model III in the scheme S6 comprises the following steps:
taking the production site environment temperature, the surface temperature, the crude oil storage duration, the initial temperature and the liquid level in the normalized simulation data set in S5 as input data, taking the internal characteristic point temperature of a storage tank in the normalized simulation data set as output data, selecting characteristic points by reference document research and actual engineering, arranging the above data to construct a training set, and establishing a BP neural network model II, wherein the characteristic points show the temperature condition of a certain position and moment of the storage tank, and the BP neural network model II is the mapping relation between independent variables and dependent variable data sets in the numerical simulation construction process;
and S5, taking the production site environmental temperature, the surface temperature, the crude oil storage duration, the initial temperature and the liquid level in the normalized indoor storage tank experimental data set as input data, taking the internal characteristic point temperature of the storage tank in the normalized indoor storage tank experimental data set as output data, finishing the above data to construct a training set, and establishing a BP neural network model III which is a mapping relation between the independent variable and the dependent variable data set in the indoor experimental process.
The specific method of S9 in the scheme is as follows:
s9.1, analyzing physical and chemical properties of a crude oil sample in a storage tank based on a temperature prediction model, and summarizing the gelation rule of crude oil;
s9.2, dividing the storage tank into areas based on a numerical simulation result of the storage tank, and acquiring temperature, pressure and density data of different areas;
s9.3, determining gelation risk evaluation weights of all areas based on storage tank area division results and combining on-site production data;
s9.4, finely dividing grids of each region according to the gelation risk weights of different regions, and identifying gelation regions;
s9.5, according to the high-low temperature distribution in the storage tank, and the working condition result of numerical simulation, adjusting the grid division method, applying the optimized grid division to an actual storage tank, acquiring temperature data of different grid areas by combining the temperature field prediction result, evaluating the gelation state of each area based on the physical property measured by experiments, combining the prediction result and the grid division, and accurately positioning the low-temperature gelation grid areas;
s9.6, carrying out risk grade assessment on the temperature field prediction result based on regional grid division, and carrying out integrated calculation on all grid blocks to obtain the risk grade of the regional blocks; then carrying out integrated calculation on the regional blocks to obtain the overall risk level of the oil storage tank; the calculation method comprises the following steps:
(1) According to the gel volume of the grid blocks, carrying out volume weighted average to obtain the risk level of the region;
(2) And carrying out volume weighted average according to the gel volume of each region to obtain the risk grade of the storage tank.
The method for dividing the storage tank in the above scheme S9.2 includes:
(1) Firstly, uniformly dividing regions with the same gradient according to temperature distribution, and dividing a storage tank into five regions: a tank top area, a tank wall area, a tank bottom area, a receiving and transmitting oil port area and a tank body center area;
(2) Sequencing the importance of the five areas in the storage tank gelation risk assessment, wherein the oil receiving and transmitting port area is a tank bottom area, a tank top area, a tank wall area and a tank body center area;
oil receiving and transmitting port: the area controls oil transceiving, and once the area is gelled, huge damage is caused to the whole storage tank;
can bottom area: the oil sludge is deposited on the bottom of the tank, the bottom oil sludge belongs to a stable accumulation layer, and if the bottom oil sludge is gelled, the cleaning of the tank bottom is extremely difficult and causes huge loss;
tank top area: the floating roof is a floating top cover, and floats up and down along with the input and output of the liquid storage, if gelation occurs, the floating of the floating roof is affected, and the oil loss is caused.
The evaluation method in the above scheme S9.5:
(1) The state of crude oil is defined as three types: hardly gelling; gelling; is easy to gel;
(2) Calculating a gelation risk grade, analyzing a gelation rule according to a condensation point and a wax precipitation amount in combination with the comprehensive consideration of actual storage tank production, classifying the gelation risk grade according to the comprehensive consideration of each influencing factor, and classifying the gelation risk grade into five grades;
(3) Selecting a low-temperature block after regional meshing, and defining the risk level of the overall temperature of the region according to the viscosity-temperature curve of the low-temperature block: the first-level risk level is that 80% of crude oil in the area is formed in a wax crystal network, namely, 80% of crude oil in the area loses flow behavior; the secondary risk grade is 60% of crude oil wax crystal network formation in the area; the third-level risk level is 40% of the crude oil wax crystal network formation in the area, the fourth-level risk level is 20% of the crude oil wax crystal network formation in the area, and the fifth-level risk level is 5% of the crude oil wax crystal network formation in the area.
In above-mentioned scheme indoor storage tank experiments go on through indoor experimental apparatus, and indoor experimental apparatus includes experiment storage tank, square cavity, temperature acquisition system, particle image velocimetry device, heating cabinet, and inside square cavity was placed in to experiment storage tank, experiment medium was full of inside experiment storage tank, temperature acquisition system included: the temperature sensor is reasonably arranged on the temperature measuring rod, and the other end of the temperature sensor is connected with the data display instrument; the experimental storage tank is provided with a floating roof, a plurality of test holes are arranged on the floating roof, a temperature measuring rod extends into the experimental storage tank through the test holes, and a plurality of temperature sensors are arranged on the temperature measuring rod; PIV tracer particles and a thermosensitive developer are added into an experiment storage tank filled with experiment medium, the PIV device is used for shooting, a temperature field and a speed field of the experiment medium in the experiment storage tank are obtained, and the temperature distribution and the movement condition of the medium in the storage tank are monitored; the experimental storage tank, the centrifugal pump and the heating box are connected in series to form a circulating system, and a valve, a temperature display and a speed display are arranged on an experimental pipeline to control the flow and the temperature of input and output experimental media.
The invention has the following beneficial effects:
1. the invention realizes the monitoring of the distribution of the temperature field in the oil storage tank. The system is used for testing the temperature field of the storage tank, grasping the temperature change data of the storage tank, determining the temperature change rule of the storage tank under different working conditions, and forming a crude oil storage tank temperature field prediction method, so that necessary data support is provided for an intelligent algorithm for further optimizing a prediction result.
2. According to the invention, the environmental temperature, the surface temperature, the storage time length, the initial temperature and the liquid level of a production site are used as input variables, the temperatures of different space points are used as output variables, a BP neural network is adopted to establish an oil storage tank temperature field distribution prediction model, the initial weight and the threshold of a particle swarm optimization algorithm model are used for optimization, then training optimization is carried out, and an optimal BP neural network prediction model is established.
3. The particle swarm BP neural network prediction model provided by the invention can be used for making a certain prediction on the internal flow and heat transfer rule of the oil storage tank and guiding the actual production process. The internal temperature of the storage tank is ensured to meet the actual requirement, and an effective solution is provided for the storage tank temperature control requirement facing the actual production.
4. Based on a large number of indoor experimental results and numerical simulation results, in combination with field production data, risk assessment is carried out on different areas of the storage tank, and according to temperature field prediction results, the storage tank gelation risk level at the current moment and the future moment is assessed, so that production safety in actual production is further guaranteed.
5. According to actual measurement data in the actual production process, numerical simulation and indoor experiments are combined, the method is used for predicting the temperature field distribution of the oil storage tank, guiding the actual production and reducing the operation cost and risk of the oil storage tank. In actual production, crude oil storage tanks are different in type and actual working condition, and the monitoring of the temperature field distribution of the oil storage tank requires larger investment. By adopting the method, the temperature field of the storage tank can be predicted based on the existing data, so that the safety of the storage tank is ensured, and the running cost is reduced.
6. The invention solves the problem of crude oil storage tank temperature field distribution by means of computer numerical simulation and experimental device, experimental method and engineering actual measurement data of the evolution process of the indoor storage tank temperature field based on the measurement method of the storage tank temperature field. In addition, an evaluation method for constructing the storage tank gelation risk level is also provided based on the temperature field prediction model, and the storage tank gelation risk is evaluated while the storage tank temperature field distribution is predicted.
Drawings
FIG. 1 is a flow chart of the operation of the present invention.
FIG. 2 is a system for measuring the temperature of an oil storage tank.
Fig. 3 is an enlarged view of the thermometric hose of fig. 2.
FIG. 4 is a schematic view of the temperature measuring point arrangement position of FIG. 2.
Fig. 5 is a frame diagram of a tank gelation risk assessment.
Detailed Description
The invention is further described with reference to the accompanying drawings:
referring to fig. 1, the method for predicting the temperature field of the crude oil storage tank and evaluating the safety of oil storage is used for screening and optimizing according to a plurality of influencing factors influencing the change of the temperature field of the storage tank in the actual production process and then determining a characteristic element group; acquiring actual storage tank temperature field data under a characteristic element group, taking the characteristic element as input data, taking the internal temperature of the storage tank as output data, forming a data set together, and establishing a BP neural network model I; based on the BP neural network model I, respectively carrying out numerical simulation and indoor storage tank experiments by taking the data of the optimized characteristic element group as numerical simulation parameters and experimental parameters to respectively obtain numerical simulation and indoor experimental data sets, and respectively establishing BP neural network models II and III; respectively optimizing the initial weights and the threshold values of the established BP neural network models II and III by adopting a particle swarm optimization algorithm, predicting a temperature field, comparing a predicted result with an actual production measurement result, and continuously correcting parameters until the error is less than 10%; and establishing a BP neural network model IV of double-sample data from data obtained by numerical simulation and indoor experiments, optimizing weights and thresholds by adopting a particle swarm algorithm, predicting a temperature field, comparing a predicted result with an actual production test result, continuously correcting parameters until the error is less than 10%, and finally obtaining an optimal temperature field predicted model. Meanwhile, based on a large number of indoor experimental results and numerical simulation results and combined field production data, risk assessment is carried out on different areas of the storage tank, and according to a temperature field prediction result, the storage tank gelation risk level at the current moment and the future moment is assessed, and the method specifically comprises the following steps:
S1, collecting data through consulting and researching, analyzing and identifying factors which possibly influence the temperature change of a crude oil storage tank in the actual production process, namely: ambient temperature of production site, surface temperature, pressure, heating mode of storage tank, heating time, crude oil storage time length, initial temperature, liquid level, crude oil density, crude oil viscosity and the like. In addition, a temperature measuring rod (a limited number of temperature sensors are arranged on the rod) is used for measuring temperatures at different positions and different moments inside the storage tank, recording temperature data and taking the temperature inside the storage tank as a performance index of a temperature field of the storage tank. And taking temperature data obtained by the temperature measuring rod as dependent variables, and corresponding influence factor data (such as production field environment temperature, surface temperature, pressure, storage tank heating mode, heating time, crude oil storage duration, initial temperature, liquid level, crude oil density, crude oil viscosity and the like) as independent variables to jointly form a data set, and carrying out data normalization pretreatment.
S2, as the temperature influence factors of the oil storage tank are more and most factors are repeated, the factors are required to be subjected to dimension reduction, key characteristic elements are extracted, effective and reasonable characteristic elements are constructed, and the prediction capacity of the model is improved. And reducing the dimension of the independent variable by adopting a principal component analysis algorithm, using a small number of variables as new input variables, filtering repeated information, and keeping main influencing factors as far as possible. And determining characteristic element combinations influencing the internal temperature of the oil storage tank, namely, factors influencing the internal temperature field of the oil storage tank in the screened field production, and particularly, the environmental temperature, the surface temperature, the storage duration, the initial temperature and the liquid level of the production field.
S3, forming a data set based on the independent variable environment temperature, the surface temperature, the storage duration, the initial temperature, the liquid level and the dependent variable after dimension reduction. And establishing a BP neural network model I, describing a storage tank temperature influence model in actual production, and continuously enriching the model through continuous field data acquisition, so that the model can better represent the storage tank temperature change condition in the actual production.
S4, based on the step S2, performing numerical simulation and indoor storage tank experiments on the data of the optimized characteristic element group as numerical simulation parameters and experimental parameters to obtain respective data sets of the numerical simulation parameters and the experimental parameters.
S5, acquiring data of numerical simulation and indoor storage tank experiments based on the step S4, and respectively performing normalization processing.
S6, taking the data of numerical simulation as sample data, and establishing a BP neural network model II; taking the data of the indoor experiment as sample data, and establishing a BP neural network model III; and simultaneously, respectively optimizing initial weights and thresholds of the neural network models II and III by adopting a particle swarm optimization algorithm. The model II is the mapping relation between the independent variable and the dependent variable data set in the process of constructing the numerical simulation, and the model III is the mapping relation between the independent variable and the dependent variable data set in the process of constructing the indoor experiment.
S7, training the optimized BP neural network models II and III constructed in the step S6 for a plurality of times, and respectively comparing the obtained results with the actual storage tank measurement results in the step S1 until the errors of the two are less than 10%; otherwise, modifying the simulation parameter value of the BP neural network model II through numerical simulation; and (3) correcting the indoor simulation parameter value of the BP neural network model III through an indoor experiment.
S8, establishing a BP neural network model IV of double-sample data according to data results obtained by final experiments and numerical simulation in the step S7, wherein the correlation between the established experiments and the numerical simulation is complementary, so that the two methods of the numerical simulation and the indoor experiments can be utilized to jointly predict field production reality, the BP neural network model IV is trained for multiple times, and the obtained results are compared with actual storage tank measurement results until the errors of the two are less than 10%; otherwise, correcting experiment and numerical simulation parameters, repeating the experiment or the numerical simulation to finally obtain an optimal prediction model, and predicting the distribution of the temperature field of the oil storage tank in actual production.
S9, based on a storage tank temperature prediction model, adopting a crude oil sample in the storage tank to carry out physicochemical property analysis, summarizing the gelation law of crude oil, providing data support for the future prediction and evaluation of the gelation state in the crude oil storage tank, and guaranteeing the safe production of crude oil.
(1) Determining the wax precipitation point of the crude oil by adopting a differential scanning calorimeter according to a determination program of SY/T0545-2012;
(2) Determining the pour point of the crude oil according to the determination procedure of ASTM D5853-11 by adopting a crude oil condensation point and pour point determinator;
(3) Determining the saturated hydrocarbon content, the naphthene content, the colloid content and the asphaltene content of the crude oil according to a determination program of SY/T5119-2016 by adopting a column separation method;
(4) The density of crude oil at different temperatures is determined by a U-shaped tube oscillation method according to the measurement program of GB/T2013-2010.
S9.1, through a large number of indoor storage tank experiments and numerical simulation, the temperature distribution difference of different positions of the storage tank is obvious, and the risk of gelation of the storage tank is evaluated according to the overall average temperature and other parameters of the storage tank, so that the storage tank is divided into areas based on the numerical simulation result of the storage tank, and temperature, pressure and density data of different areas are obtained. The method comprises the following steps:
(1) Firstly, uniformly dividing regions with the same gradient according to temperature distribution, and dividing a storage tank into five regions: a tank top area, a tank wall area, a tank bottom area, a transceiver oil port area and a tank body center area.
(2) The importance of the different zones in the risk assessment of tank gelation is then ordered on the basis of the previous step. The oil receiving and transmitting port area is equal to the tank bottom area, the tank top area, the tank wall area and the tank body center area.
Oil receiving and transmitting port: this zone controls oil transceiving and once this zone gels, it will cause a significant hazard to the whole tank.
Can bottom area: the sludge is deposited on the tank bottom, the bottom sludge belongs to a stable accumulation layer, and if the bottom sludge is gelled, the tank bottom is cleaned with great difficulty and great loss.
Tank top area: the floating roof is a floating top cover, and floats up and down along with the input and output of the liquid storage, if gelation occurs, the floating of the floating roof is affected, and the oil loss is caused.
S9.2, determining gelation risk evaluation weights of all areas based on the storage tank area division of the S9.1 and combining on-site production data.
And S9.3, finely dividing the regional grids according to the gelation risk weights of different regions to obtain more accurate results, thereby facilitating the recognition of the gelation regions.
And S9.4, according to the high-low temperature distribution in the storage tank, and combining a large number of working condition results of numerical simulation, adjusting the dividing method of the S9.3, and improving the applicability of the method. The optimized grid division is applied to an actual storage tank, temperature data of different areas are obtained by combining temperature field prediction results, the gelation state of the areas is estimated based on physical properties measured by experiments, the prediction results and the grid division are combined, a low-temperature gelation grid area is accurately positioned, and technical support is provided for storage tank heating. Wherein the evaluation steps are as follows:
(1) The state of crude oil is first defined as three: hardly gelling; gelling; is easy to gel.
(2) And secondly, calculating the gelation risk grade, and classifying the gelation risk grade into five grades according to the data such as the condensation point, the wax precipitation amount and the like by combining with the comprehensive consideration of the actual storage tank production. The gel degree is judged by simply relying on the gel point, so that the defect is overcome in actual production, and the gel of crude oil is not represented even if the gel point is reached in the actual production process. Therefore, the gelation law is analyzed by combining with the indoor experimental study, and the gelation risk grade is classified by comprehensively considering all influence factors.
(3) Selecting a low-temperature block after regional meshing, and defining the risk level of the overall temperature of the region according to the viscosity-temperature curve of the low-temperature block: the first-level risk level is that 80% of crude oil in the area is formed in a wax crystal network, namely, 80% of crude oil in the area loses flow behavior; the secondary risk grade is 60% of crude oil wax crystal network formation in the area; the third-level risk level is 40% of the crude oil wax crystal network formation in the area, the fourth-level risk level is 20% of the crude oil wax crystal network formation in the area, and the fifth-level risk level is 5% of the crude oil wax crystal network formation in the area.
And S9.5, carrying out risk grade assessment on the temperature field prediction result based on regional grid division. Integrating and calculating the grid blocks to obtain the risk level of the regional blocks; and then carrying out integrated calculation on the regional blocks to obtain the overall risk level of the oil storage tank. The calculation method comprises the following steps:
(1) And carrying out volume weighted average according to the gel volume of the grid blocks to obtain the risk grade of the region.
(2) And carrying out volume weighted average according to the gel volume of each region to obtain the risk grade of the storage tank.
The value range of the variable in the scheme is derived from the variation interval of the equipment parameters in the actual production process.
The indoor experimental device in the invention comprises: experiment storage tank, square cavity, temperature acquisition system, particle Image Velocimetry (PIV), heating cabinet, centrifugal pump. The experiment storage tank and the square cavity are made of transparent materials, the experiment storage tank is arranged in the square cavity, and the experiment storage tank is filled with experiment medium. The heating box provides a guarantee for the initial fluid temperature of the storage tank. The temperature acquisition system includes: the temperature sensor is reasonably arranged on the temperature measuring rod, and the other end of the temperature sensor is connected with the data display instrument to realize real-time data transmission. The experiment storage tank is provided with a floating roof, a plurality of test holes are arranged on the floating roof, the temperature measuring rod stretches into the experiment storage tank through the test holes, and a plurality of temperature sensors can be arranged on the temperature measuring rod. The bottom of the tank wall is provided with an outlet and an inlet, and the output and input of experimental media in the storage tank are realized through a centrifugal pump. PIV tracer particles and a thermosensitive developer are added into an experiment storage tank filled with experiment medium, and shooting is carried out through a PIV device, so that a temperature field and a speed field of the experiment medium in the experiment storage tank are obtained, and further the temperature distribution and the movement condition of the medium in the storage tank are monitored.
The experimental device connects the experimental storage tank, the centrifugal pump and the heating box in series through the experimental pipeline to form a circulating system, and the experimental pipeline is provided with a valve, a temperature display and a speed display, so that the flow and the temperature of input and output experimental media can be controlled. The particle image velocimetry device comprises a laser emission system, an image acquisition system and operation processing software.
The experimental method for observing the temperature field and the speed field in the storage tank by the indoor experimental device comprises the following steps:
(1) Cleaning an experimental device to prevent dirt from interfering experimental shooting;
(2) Reasonably arranging the temperature measuring rod and the temperature measuring point according to an experimental scheme, and starting a temperature collector;
(3) Heating the experimental medium to a preset temperature by using a heating box;
(4) Using a centrifugal pump to pump a test medium to an experiment storage tank through pipe transportation;
(5) Starting a particle image velocimetry system;
(6) PIV trace particles and a thermosensitive developer are added into the storage tank to capture the movement track of the particles, and the temperature sensor is combined with the thermosensitive developer, so that the change of the temperature field of the storage tank can be reflected better;
(7) And combining the temperature acquisition data, and visualizing through software. Coupling and analyzing the temperature field and the speed field of the storage tank,
And analyzing the evolution law of the temperature field of the storage tank.
According to the particle swarm optimization algorithm, the calculation accuracy is greatly improved, after engineering measured data and prediction data are compared, the BP neural network model is enriched through iteration of numerical simulation data and indoor experimental data, and the feasibility of the prediction model is greatly guaranteed.
Examples:
according to factors influencing the change of the internal temperature field of the oil storage tank in the actual engineering production process, a principal component analysis method is adopted to screen and optimize a data set, remove repeated or relatively high-correlation data, and determine characteristic element combinations influencing the internal temperature of the oil storage tank, namely: ambient temperature at production site, surface temperature, storage duration, initial temperature and liquid level. And respectively carrying out numerical simulation and indoor storage tank experiments by taking the optimized data of the characteristic element group as numerical simulation parameters and experimental parameters.
The numerical simulation scheme is shown in table 1:
table 1 numerical simulation scheme
The indoor protocol is shown in table 2 below:
table 2 table of laboratory protocols
S1, collecting data through consulting and researching, analyzing and identifying factors which possibly influence the temperature change of a crude oil storage tank in the actual production process, namely: ambient temperature of production site, surface temperature, pressure, heating mode of storage tank, heating time, crude oil storage time length, initial temperature, liquid level, crude oil density, crude oil viscosity and the like. In addition, a temperature measuring rod (a limited number of temperature sensors are arranged on the rod) is used for measuring temperatures at different positions and different moments inside the storage tank, recording temperature data and taking the temperature inside the storage tank as a performance index of a temperature field of the storage tank. And taking temperature data obtained by the temperature measuring rod as dependent variables, and corresponding influence factor data (such as production field environment temperature, surface temperature, pressure, storage tank heating mode, heating time, crude oil storage duration, initial temperature, liquid level, crude oil density, crude oil viscosity and the like) as independent variables to jointly form a data set, and carrying out data preprocessing.
S2, as the temperature influence factors of the oil storage tank are more and most factors are repeated, the factors are required to be subjected to dimension reduction, key characteristic elements are extracted, effective and reasonable characteristic elements are constructed, and the prediction capacity of the model is improved. And reducing the dimension of the independent variable by adopting a principal component analysis algorithm, using a small number of variables as new input variables, filtering repeated information, and keeping main influencing factors as far as possible. And determining characteristic element combinations influencing the internal temperature of the oil storage tank, namely, factors influencing the internal temperature field of the oil storage tank in the screened field production, and particularly, the environmental temperature, the surface temperature, the storage duration, the initial temperature and the liquid level of the production field.
Principle of Principal Component Analysis (PCA): a multi-element statistical method for changing multiple variables into few variables by using the idea of dimension reduction. The essence is that the comprehensive substitute object of the related variable is sought through the relativity of the original variable, and the minimum information loss in the conversion process is ensured.
The application steps are as follows:
step one: calculating a covariance matrix Z according to the standardized data set:
for a set of samples X, there are m observations X 1 ,x 2 ,……,x m There are n groups.
(5) The average value for each column is calculated:
(6) Variance of each column:
(7) And (3) carrying out standardized processing on the data:
obtaining a matrix:
(8) Calculating a correlation coefficient:
obtaining a correlation coefficient matrix R:
r ij is the correlation between the samples of the ith row and the samples of the jth column of the matrix X, and has a value of [ -1,1]And the R matrix should be a symmetric matrix. When r is>At 0, positive linear correlation; r=0, linear independent; when r is<0, is a linear negative correlation.
Step two: calculating the characteristic value of R and the corresponding characteristic vector
The covariance matrix is a real symmetric matrix, and its eigenvalue is known to be non-negative, so that its eigenvalue may be set.
λ 1 ≥λ 2 ≥λ 3 ≥…≥λ p And (3) not less than 0, and the unit eigenvectors after orthogonalization corresponding to the unit eigenvectors are as follows:
if the index variable represented by each column of original X is a synthesized vector, the synthesized vector is expressed as Var= [ Var ] 1 ,Var 2 ,…,Var p ] T The ith principal component with X is then:
F i =(a i ) T Var=a 1i *Var 1 +a 2i *Var 2 +…+a pi *Var p
step three: calculating the contribution rate and the accumulated contribution rate of each component
The number of principal components is determined according to the cumulative contribution rate, and the cumulative contribution rate is generally required to reach more than 85%, so that the new variable can be ensured to comprise most of information of the original variable.
Principal component metrics are retained:
c. the contribution rate of the reserved component variance reaches more than 85 percent;
d. the variance of the principal component is kept to be larger than 1;
e. and displaying the characteristic values in a graphical mode by adopting a lithotriptic graph, and reserving the number of main components corresponding to the position with the largest change in the graph.
S3, forming a data set based on the independent variable environment temperature, the surface temperature, the storage duration, the initial temperature, the liquid level and the dependent variable after dimension reduction. And establishing a BP neural network model I, describing a storage tank temperature influence model in actual production, and continuously enriching the model through continuous field data acquisition, so that the model can better represent the storage tank temperature change condition in the actual production.
S4, based on the step S2, performing numerical simulation and indoor storage tank experiments on the data of the optimized characteristic element group as numerical simulation parameters and experimental parameters to obtain respective data sets of the numerical simulation parameters and the experimental parameters.
S5, acquiring data of numerical simulation and indoor storage tank experiments based on the step S4, and respectively performing normalization processing.
S6, taking the data of numerical simulation as sample data, and establishing a BP neural network model II; taking the data of the indoor experiment as sample data, and establishing a BP neural network model III; and simultaneously, respectively optimizing initial weights and thresholds of the neural network models II and III by adopting a particle swarm optimization algorithm. The model II is the mapping relation between the independent variable and the dependent variable data set in the process of constructing the numerical simulation, and the model III is the mapping relation between the independent variable and the dependent variable data set in the process of constructing the indoor experiment.
Because the weight and threshold initial values of the neural network are random numbers in the interval of [ -0.5,0.5], the initial values have great influence on network training but cannot be accurately obtained, the weight and threshold values are optimized by adopting a particle swarm algorithm, and the prediction precision is improved. The global searching capability of the particle swarm algorithm is utilized to optimize the topological structure, the connection weight and the threshold value of the neural network, the global optimizing capability of the particle swarm algorithm is combined with the local optimizing capability of the BP neural network, the problem of local minimum points is solved, the learning performance and the generalization capability of the neural network are improved, and the prediction precision of the neural network is optimized.
In PSO, the dimension D of the entire search space, the location of the ith particle is given by:
wherein: omega 11 ……ω nl Is W 1 Is an element of (2); omega 1m ……ω lm Is W 2 Is an element of (2);for theta 1 Elements of (2):for theta 2 Is an element of (2); d is the dimension of the search space.
(6) Generating a population, randomly initializing, setting a maximum speed V for each particle max Corresponds to an initial position maximum X max A minimum velocity V is set for each particle min Corresponds to a minimum value X of an initial position min At [ -X max ,X max ]、[-V max ,V max ]The positions and speeds of the particles are randomly selected in the interval to initialize, and each particle in the D-dimensional space in the population is represented by a set of weights and thresholds of the BP neural network.
(7) Setting a termination condition, the iteration number N, the population scale M, an initial value w of inertia weight, initial values C1 and C2 of acceleration factors and the like of PSO optimization.
(8) The fitness value F (i) of each particle is obtained from the initial position of each particle by a fitness function as shown in the following formula.
Wherein: n is the number of output nodes; y is i A desired output for an ith node of the neural network; o (o) i A prediction output for the i-th node; k is a coefficient.
(9) Extremum gbest of each particle population i Setting the particle size as the optimal value of the objective function, calculating the fitness value of each particle, and if F (i)>pbest i F (i) replaces pbest i Optimizing population extremum gbest i The velocity and position of the particles are updated in each iteration according to the following equation.
v i =v i +c 1 ×rand()×(pbest i -x i )+c 2 ×rand()×(gbest i -x i )
x i =x i +v i
In the formula, i=1, 2,3 … …, N is the total number of particles in the population.
v i : particle velocity;
rand (): a random number between (0, 1);
x i : current position of particle
c 1 And c 2 : is a learning factor, generally c 1 =c 2 =2;
v i Has a maximum value of V max (greater than 0), if v i Greater than V max V is then i =V max 。
(10) PSO optimization is terminated to obtain the optimal weight and threshold of BP neural network
S7, training the optimized BP neural network models II and III constructed in the step S6 for a plurality of times, and respectively comparing the obtained results with the actual storage tank measurement results in the step S1 until the errors of the two are less than 10%; otherwise, modifying the simulation parameter value of the BP neural network model II through numerical simulation; and (3) correcting the indoor simulation parameter value of the BP neural network model III through an indoor experiment.
S8, establishing a BP neural network model IV of double-sample data according to data results obtained by final experiments and numerical simulation in the step S7, wherein the correlation between the established experiments and the numerical simulation is complementary, so that the two methods of the numerical simulation and the indoor experiments can be utilized to jointly predict field production reality, the BP neural network model IV is trained for multiple times, and the obtained results are compared with actual storage tank measurement results until the errors of the two are less than 10%; otherwise, correcting experiment and numerical simulation parameters, repeating the experiment or the numerical simulation to finally obtain an optimal prediction model, and predicting the distribution of the temperature field of the oil storage tank in actual production.
S9, based on a storage tank temperature prediction model, adopting a crude oil sample in the storage tank to carry out physicochemical property analysis, summarizing the gelation law of crude oil, providing data support for the future prediction and evaluation of the gelation state in the crude oil storage tank, and guaranteeing the safe production of crude oil.
(1) Determining the wax precipitation point of the crude oil by adopting a differential scanning calorimeter according to a determination program of SY/T0545-2012;
(2) Determining the pour point of the crude oil according to the determination procedure of ASTM D5853-11 by adopting a crude oil condensation point and pour point determinator;
(3) Determining the saturated hydrocarbon content, the naphthene content, the colloid content and the asphaltene content of the crude oil according to a determination program of SY/T5119-2016 by adopting a column separation method;
(4) The density of crude oil at different temperatures is determined by a U-shaped tube oscillation method according to the measurement program of GB/T2013-2010.
S10, through a large number of indoor storage tank experiments and numerical simulation, the temperature distribution difference of different positions of the storage tank is obvious, and the risk of gelation of the storage tank is estimated according to the overall average temperature and other parameters of the storage tank, so that the storage tank is divided into areas based on the numerical simulation result of the storage tank, and temperature, pressure and density data of different areas are obtained. The method comprises the following steps:
(1) Firstly, uniformly dividing regions with the same gradient according to temperature distribution, and dividing a storage tank into five regions: a tank top area, a tank wall area, a tank bottom area, a transceiver oil port area and a tank body center area.
(2) The importance of the different zones in the risk assessment of tank gelation is then ordered on the basis of the previous step. The oil receiving and transmitting port area is equal to the tank bottom area, the tank top area, the tank wall area and the tank body center area.
Oil receiving and transmitting port: this zone controls oil transceiving and once this zone gels, it will cause a significant hazard to the whole tank.
Can bottom area: the sludge is deposited on the tank bottom, the bottom sludge belongs to a stable accumulation layer, and if the bottom sludge is gelled, the tank bottom is cleaned with great difficulty and great loss.
Tank top area: the floating roof is a floating top cover, and floats up and down along with the input and output of the liquid storage, if gelation occurs, the floating of the floating roof is affected, and the oil loss is caused.
S11, determining gelation risk evaluation weights of all areas based on storage tank area division of the S10 and combining on-site production data.
S12, finely dividing the area grid according to the gelation risk weights of different areas to obtain a more accurate result, so that the gelation areas are conveniently identified.
S13, according to the high-low temperature distribution in the storage tank, and combining a large number of working condition results of numerical simulation, the dividing method of S12 is adjusted, and applicability of the method is improved. The optimized grid division is applied to an actual storage tank, temperature data of different areas are obtained by combining temperature field prediction results, the gelation state of the areas is estimated based on physical properties measured through experiments in S9, the prediction results and the grid division are combined, a low-temperature gelation grid area is accurately positioned, and technical support is provided for storage tank heating. Wherein the evaluation steps are as follows:
(1) The state of crude oil is first defined as three: hardly gelling; gelling; is easy to gel.
(2) And secondly, calculating the gelation risk grade, and classifying the gelation risk grade into five grades according to the data such as the condensation point, the wax precipitation amount and the like by combining with the comprehensive consideration of the actual storage tank production. The gel degree is judged by simply relying on the gel point, so that the defect is overcome in actual production, and the gel of crude oil is not represented even if the gel point is reached in the actual production process. Therefore, the gelation law is analyzed by combining with the indoor experimental study, and the gelation risk grade is classified by comprehensively considering all influence factors.
(3) Selecting a low-temperature block after regional meshing, and defining the risk level of the overall temperature of the region according to the viscosity-temperature curve of the low-temperature block: the first-level risk level is that 80% of crude oil in the area is formed in a wax crystal network, namely, 80% of crude oil in the area loses flow behavior; the secondary risk grade is 60% of crude oil wax crystal network formation in the area; the third-level risk level is 40% of the crude oil wax crystal network formation in the area, the fourth-level risk level is 20% of the crude oil wax crystal network formation in the area, and the fifth-level risk level is 5% of the crude oil wax crystal network formation in the area.
S14, carrying out risk grade assessment on the temperature field prediction result based on regional grid division. Integrating and calculating the grid blocks to obtain the risk level of the regional blocks; and then carrying out integrated calculation on the regional blocks to obtain the overall risk level of the oil storage tank. The calculation method comprises the following steps:
(1) And carrying out volume weighted average according to the gel volume of the grid blocks to obtain the risk grade of the region.
(2) And carrying out volume weighted summation according to the gelled volume of each region to obtain the risk grade of the storage tank.
Claims (8)
1. A method for predicting a temperature field of a crude oil storage tank and evaluating oil storage safety, which is characterized by comprising the following steps:
s1, acquiring factors and corresponding data influencing an internal temperature field of an oil storage tank in field production, wherein the factors and the corresponding data comprise: production site environment temperature, surface temperature, pressure, storage tank heating mode, heating time, crude oil storage time length, initial temperature, liquid level, crude oil density and crude oil viscosity; measuring temperatures at different positions and at different moments in the storage tank by using a temperature measuring rod, recording temperature data, taking the temperature in the storage tank as an index of the performance of a temperature field of the oil storage tank, taking the temperature data obtained by the temperature measuring rod as a dependent variable and corresponding influence factor data as independent variables, forming a data set together, and preprocessing the data;
S2, screening and optimizing the preprocessed data set by adopting a dimension reduction algorithm, removing repeated or relatively high-correlation data, and determining that characteristic elements affecting the internal temperature of the oil storage tank are combined into factors affecting the internal temperature field of the oil storage tank in screened field production, specifically the environmental temperature, the surface temperature, the crude oil storage duration, the initial temperature and the liquid level of the production field;
s3, based on the independent variables and the dependent variables after dimension reduction, a data set is formed together, a BP neural network model I is established, a storage tank temperature influence model in actual production is described, and the model is continuously enriched through continuous field data acquisition, so that the model truly represents the storage tank temperature change condition in actual production;
s4, based on the S2, performing numerical simulation and an indoor storage tank experiment respectively by taking the data of the optimized characteristic element group as a numerical simulation parameter and an experimental parameter to obtain a simulation data set and an indoor storage tank experiment data set;
s5, respectively carrying out normalization processing on the simulation data set and the indoor storage tank experimental data set;
s6, establishing a BP neural network model II by using the normalized simulation data in the S5, wherein the BP neural network model II is a mapping relation between independent variables and a data set of dependent variables in the numerical simulation process; establishing a BP neural network model III by using the indoor storage tank experimental data set normalized in the step S5; respectively optimizing the initial weight and the threshold of the constructed BP neural network model II and BP neural network model III by adopting a particle swarm optimization algorithm;
S7, training the BP neural network model II and the BP neural network model III optimized by the S6 for a plurality of times, and respectively comparing the obtained results with the measurement result of the S1 actual storage tank until the error of the two results is less than 10%; otherwise, modifying the simulation parameter value of the BP neural network model II through numerical simulation; indoor experiment corrects the indoor simulation parameter value of BP neural network model III;
s8, utilizing the data results obtained by the S7 final experiment and numerical simulation to establish a BP neural network model IV of double-sample data, training the BP neural network model IV for multiple times, and comparing the obtained results with the actual storage tank measurement results until the errors of the two are less than 10%; otherwise, correcting experiment and numerical simulation parameters, repeating the experiment or the numerical simulation to obtain an optimal prediction model, namely a temperature field prediction model, and predicting the temperature field in the storage tank under different conditions by using the temperature field prediction model;
s9, based on the temperature field prediction model of S8, and combining an indoor crude oil rheological experiment, evaluating and predicting the gelation risk level of crude oil in the storage tank in actual production.
2. The method for predicting the temperature field of a crude oil storage tank and evaluating the safety of oil storage according to claim 1, wherein: the dimension reduction algorithm in the S2 is as follows:
Step one: calculating a covariance matrix Z from the data set:
for a set of samples X, there are m observations X 1 ,x 2 ,……,x m A total of n groups,
(1) The average value for each column is calculated:
(2) Variance of each column:
(3) And (3) carrying out standardized processing on the data:
obtaining a matrix:
(4) Calculating a correlation coefficient:
obtaining a correlation coefficient matrix R:
r ij is the correlation between the samples of the ith row and the samples of the jth column of the matrix X, and has a value of [ -1,1]And the R matrix should be a symmetric matrix, when R>At 0, positive linear correlation; r=0, linear independent; when r is<0, is a linear negative correlation;
step two: calculating the characteristic value of R and the corresponding characteristic vector
The covariance matrix is a real symmetric matrix, the eigenvalue of which is known to be non-negative, the eigenvalue of which is set,
λ 1 ≥λ 2 ≥λ 3 ≥…≥λ p and (3) not less than 0, and the unit eigenvectors after orthogonalization corresponding to the unit eigenvectors are as follows:
if the index variable represented by each column of original X is a synthesized vector, the synthesized vector is expressed as Var= [ Var ] 1 ,Var 2 ,…,Var p ] T The ith principal component with X is then:
F i =(a i ) T Var=a 1i *Var 1 +a 2i *Var 2 +…+a pi *Var p
step three: calculating the contribution rate and the accumulated contribution rate of each component
The selection of the number of the main components is determined according to the accumulated contribution rate, the accumulated contribution rate reaches more than 85%, and the main component measurement standard is reserved:
a. the contribution rate of the reserved component variance reaches more than 85 percent;
b. The variance of the principal component is kept to be larger than 1;
and displaying the characteristic values in a graphical mode by adopting a lithotriptic graph, and reserving the number of main components corresponding to the position with the largest change in the graph.
3. The method for predicting the temperature field of a crude oil storage tank and evaluating the safety of oil storage according to claim 2, wherein: in the step S6, a particle swarm optimization algorithm is adopted to optimize the initial weight and the threshold value of the BP neural network:
dimension D of the entire search space, the location of the ith particle is given by:
wherein: omega 11 ……ω nl Is W 1 Is an element of (2); omega 1m ……ω lm Is W 2 Is an element of (2);for theta 1 Elements of (2): />For theta 2 Is an element of (2); d is the dimension of the search space;
(1) Generating a population, randomly initializing, setting a maximum speed V for each particle max Corresponds to an initial position maximum X max A minimum velocity V is set for each particle min Corresponds to a minimum value X of an initial position min At [ -X max ,X max ]、[-V max ,V max ]Randomly selecting the position and the speed of particles in a section to initialize, wherein each particle in a D-dimensional space in the population is represented by a set of weight and threshold values of BP neural network;
(2) Setting a termination condition, iteration times N, population scale M, an initial value w of inertia weight and initial values C1 and C2 of acceleration factors of PSO optimization;
(3) Obtaining a fitness value F (i) of each particle by a fitness function according to the initial position of each particle:
Wherein: n is the number of output nodes; y is i A desired output for an ith node of the neural network; o (o) i A prediction output for the i-th node; k is a coefficient;
(4) Extremum gbest of each particle population i Setting the particle size as the optimal value of the objective function, calculating the fitness value of each particle, and if F (i)>pbest i F (i) replaces pbest i Optimizing population extremum gbest i The velocity and position of the particles are updated in each iteration according to the following formula:
v i =v i +c 1 ×rand()×(pbest i -x i )+c 2 ×rand()×(gbest i -x i )
x i =x i +v i
in the formula, i=1, 2,3 … …, N is the total number of particles in the population;
v i : particle velocity; rand (): a random number between (0, 1); x is x i : the current position of the particle; c 1 And c 2 : is a learning factor, c 1 =c 2 =2;v i Has a maximum value of V max (greater than 0), if v i Greater than V max V is then i =V max ;
And (3) terminating PSO optimization to obtain the optimal weight and threshold of the BP neural network.
4. A method of predicting the temperature field of a crude oil storage tank and evaluating the safety of oil storage according to claim 3, wherein: the method for establishing the BP neural network model II and the BP neural network model III in the S6 comprises the following steps:
taking the production site environment temperature, the surface temperature, the crude oil storage duration, the initial temperature and the liquid level in the normalized simulation data set in S5 as input data, taking the internal characteristic point temperature of a storage tank in the normalized simulation data set as output data, selecting characteristic points by reference document research and actual engineering, arranging the above data to construct a training set, and establishing a BP neural network model II, wherein the characteristic points show the temperature condition of a certain position and moment of the storage tank, and the BP neural network model II is the mapping relation between independent variables and dependent variable data sets in the numerical simulation construction process;
And S5, taking the production site environmental temperature, the surface temperature, the crude oil storage duration, the initial temperature and the liquid level in the normalized indoor storage tank experimental data set as input data, taking the internal characteristic point temperature of the storage tank in the normalized indoor storage tank experimental data set as output data, finishing the above data to construct a training set, and establishing a BP neural network model III which is a mapping relation between the independent variable and the dependent variable data set in the indoor experimental process.
5. The method for predicting the temperature field of a crude oil storage tank and evaluating the safety of oil storage according to claim 4, wherein: the specific method of S9 is as follows:
s9.1, analyzing physical and chemical properties of a crude oil sample in a storage tank based on a temperature prediction model, and summarizing the gelation rule of crude oil;
s9.2, dividing the storage tank into areas based on a numerical simulation result of the storage tank, and acquiring temperature, pressure and density data of different areas;
s9.3, determining gelation risk evaluation weights of all areas based on storage tank area division results and combining on-site production data;
s9.4, finely dividing grids of each region according to the gelation risk weights of different regions, and identifying gelation regions;
S9.5, according to the high-low temperature distribution in the storage tank, and the working condition result of numerical simulation, adjusting the grid division method, applying the optimized grid division to an actual storage tank, acquiring temperature data of different grid areas by combining the temperature field prediction result, evaluating the gelation state of each area based on the physical property measured by experiments, combining the prediction result and the grid division, and accurately positioning the low-temperature gelation grid areas;
s9.6, carrying out risk grade assessment on the temperature field prediction result based on regional grid division, and carrying out integrated calculation on all grid blocks to obtain the risk grade of the regional blocks; then carrying out integrated calculation on the regional blocks to obtain the overall risk level of the oil storage tank; the calculation method comprises the following steps:
(1) According to the gel volume of the grid blocks, carrying out volume weighted average to obtain the risk level of the region;
(2) And carrying out volume weighted average according to the gel volume of each region to obtain the risk grade of the storage tank.
6. The method for predicting the temperature field of a crude oil storage tank and evaluating the safety of oil storage according to claim 5, wherein: the method for dividing the storage tank in the S9.2 comprises the following steps:
(1) Firstly, uniformly dividing regions with the same gradient according to temperature distribution, and dividing a storage tank into five regions: a tank top area, a tank wall area, a tank bottom area, a receiving and transmitting oil port area and a tank body center area;
(2) Sequencing the importance of the five areas in the storage tank gelation risk assessment, wherein the oil receiving and transmitting port area is a tank bottom area, a tank top area, a tank wall area and a tank body center area;
oil receiving and transmitting port: the area controls oil transceiving, and once the area is gelled, huge damage is caused to the whole storage tank;
can bottom area: the oil sludge is deposited on the bottom of the tank, the bottom oil sludge belongs to a stable accumulation layer, and if the bottom oil sludge is gelled, the cleaning of the tank bottom is extremely difficult and causes huge loss;
tank top area: the floating roof is a floating top cover, and floats up and down along with the input and output of the liquid storage, if gelation occurs, the floating of the floating roof is affected, and the oil loss is caused.
7. The method for predicting the temperature field of a crude oil storage tank and evaluating the safety of oil storage according to claim 6, wherein: the evaluation method in S9.5:
(1) The state of crude oil is defined as three types: hardly gelling; gelling; is easy to gel;
(2) Calculating a gelation risk grade, analyzing a gelation rule according to a condensation point and a wax precipitation amount in combination with the comprehensive consideration of actual storage tank production, classifying the gelation risk grade according to the comprehensive consideration of each influencing factor, and classifying the gelation risk grade into five grades;
(3) Selecting a low-temperature block after regional meshing, and defining the risk level of the overall temperature of the region according to the viscosity-temperature curve of the low-temperature block: the first-level risk level is that 80% of crude oil in the area is formed in a wax crystal network, namely, 80% of crude oil in the area loses flow behavior; the secondary risk grade is 60% of crude oil wax crystal network formation in the area; the third-level risk level is 40% of the crude oil wax crystal network formation in the area, the fourth-level risk level is 20% of the crude oil wax crystal network formation in the area, and the fifth-level risk level is 5% of the crude oil wax crystal network formation in the area.
8. The method for predicting the temperature field of a crude oil storage tank and evaluating the safety of oil storage according to claim 7, wherein: the indoor storage tank experiment is carried out through indoor experimental apparatus, and indoor experimental apparatus includes experiment storage tank, square cavity, temperature acquisition system, particle image velocimetry device, heating cabinet, and experiment storage tank, square cavity all adopt transparent material, and square cavity is placed in to the experiment storage tank inside, and experiment medium is full of inside the experiment storage tank, and temperature acquisition system includes: the temperature sensor is reasonably arranged on the temperature measuring rod, and the other end of the temperature sensor is connected with the data display instrument; the experimental storage tank is provided with a floating roof, a plurality of test holes are arranged on the floating roof, a temperature measuring rod extends into the experimental storage tank through the test holes, and a plurality of temperature sensors are arranged on the temperature measuring rod; PIV tracer particles and a thermosensitive developer are added into an experiment storage tank filled with experiment medium, the PIV device is used for shooting, a temperature field and a speed field of the experiment medium in the experiment storage tank are obtained, and the temperature distribution and the movement condition of the medium in the storage tank are monitored; the experimental storage tank, the centrifugal pump and the heating box are connected in series to form a circulating system, and a valve, a temperature display and a speed display are arranged on an experimental pipeline to control the flow and the temperature of input and output experimental media.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310581716.4A CN116595881A (en) | 2023-05-22 | 2023-05-22 | Method for predicting temperature field of crude oil storage tank and evaluating oil storage safety |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310581716.4A CN116595881A (en) | 2023-05-22 | 2023-05-22 | Method for predicting temperature field of crude oil storage tank and evaluating oil storage safety |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116595881A true CN116595881A (en) | 2023-08-15 |
Family
ID=87589437
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310581716.4A Pending CN116595881A (en) | 2023-05-22 | 2023-05-22 | Method for predicting temperature field of crude oil storage tank and evaluating oil storage safety |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116595881A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20220180202A1 (en) * | 2019-09-12 | 2022-06-09 | Huawei Technologies Co., Ltd. | Text processing model training method, and text processing method and apparatus |
CN117272661A (en) * | 2023-09-28 | 2023-12-22 | 东北石油大学 | Prediction method for cold and hot oil mixing effect of large crude oil storage tank |
CN117870775A (en) * | 2024-03-11 | 2024-04-12 | 山东港源管道物流有限公司 | Storage tank detection system and method based on intelligent oil depot |
CN118070612A (en) * | 2024-03-13 | 2024-05-24 | 东北石油大学 | Method for representing microscopic scale force chain network structure of waxy crude oil based on CFD-DEM model |
-
2023
- 2023-05-22 CN CN202310581716.4A patent/CN116595881A/en active Pending
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20220180202A1 (en) * | 2019-09-12 | 2022-06-09 | Huawei Technologies Co., Ltd. | Text processing model training method, and text processing method and apparatus |
CN117272661A (en) * | 2023-09-28 | 2023-12-22 | 东北石油大学 | Prediction method for cold and hot oil mixing effect of large crude oil storage tank |
CN117272661B (en) * | 2023-09-28 | 2024-05-14 | 东北石油大学 | Prediction method for cold and hot oil mixing effect of large crude oil storage tank |
CN117870775A (en) * | 2024-03-11 | 2024-04-12 | 山东港源管道物流有限公司 | Storage tank detection system and method based on intelligent oil depot |
CN117870775B (en) * | 2024-03-11 | 2024-05-14 | 山东港源管道物流有限公司 | Storage tank detection system and method based on intelligent oil depot |
CN118070612A (en) * | 2024-03-13 | 2024-05-24 | 东北石油大学 | Method for representing microscopic scale force chain network structure of waxy crude oil based on CFD-DEM model |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN116595881A (en) | Method for predicting temperature field of crude oil storage tank and evaluating oil storage safety | |
CN114818363B (en) | Urban deep drainage tunnel siltation risk assessment and prediction early warning method | |
CN103870670B (en) | A kind of tube corrosion degree Forecasting Methodology and device | |
CN105760673A (en) | Fluvial facies reservoir earthquake sensitive parameter template analysis method | |
CN115906675B (en) | Well position and injection and production parameter joint optimization method based on time sequence multi-target prediction model | |
CN116702638B (en) | Double-layer intelligent diagnosis method and system for sedimentation disease of drainage pipeline | |
CN102880905A (en) | Online soft measurement method for normal oil dry point | |
CN107491840A (en) | The prediction of flowing wearing character and lifetime estimation method based on ELM neural network models | |
CN110472689A (en) | The thin-skinned measurement method of Dlagnosis of Sucker Rod Pumping Well hydrodynamic returned based on integrated Gaussian process | |
CN112541571A (en) | Injection-production connectivity determination method based on machine learning of double parallel neural networks | |
CN113297740A (en) | Horizontal well gel profile control parameter optimization method | |
CN111429028A (en) | Power transmission line icing disaster risk assessment method suitable for mountainous terrain | |
CN116821695B (en) | Semi-supervised neural network soft measurement modeling method | |
CN112302594A (en) | Method, device and equipment for determining connection structure of heterogeneous water-drive reservoir | |
CN112733340A (en) | Well selection method and equipment for modifying candidate well based on data-driven reservoir | |
CN116663126A (en) | Bridge temperature effect prediction method based on channel attention BiLSTM model | |
CN116856905A (en) | Crude oil density real-time quantitative prediction method based on multicomponent gas detection system | |
CN116188266A (en) | Spatial interpolation method for geological investigation based on fuzzy theory and ensemble learning | |
CN115758915A (en) | Heat exchange tube overhauling and intelligent operation and maintenance control method of energy pile active bridge deck deicing and snow melting system | |
CN112943215A (en) | Method for selecting horizontal well from water direction by monitoring and judging oil reservoir pressure response | |
CN113627640A (en) | Productivity well testing prediction method and system for fracture-cavity type oil reservoir oil and gas well | |
CN110348123A (en) | A kind of oil-gas pipeline compressional deformation estimation method based on AIC-RBF | |
CN111859280B (en) | Method and device for evaluating physical properties of stratum crude oil | |
CN113756786B (en) | Oil pumping well time sequence indicator diagram prediction method based on deep learning | |
Feder | Machine-learning approach determines spatial variation in shale decline curves |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |