CN113240321A - Oil field analogy evaluation method and system based on deep learning - Google Patents
Oil field analogy evaluation method and system based on deep learning Download PDFInfo
- Publication number
- CN113240321A CN113240321A CN202110605795.9A CN202110605795A CN113240321A CN 113240321 A CN113240321 A CN 113240321A CN 202110605795 A CN202110605795 A CN 202110605795A CN 113240321 A CN113240321 A CN 113240321A
- Authority
- CN
- China
- Prior art keywords
- oil field
- oil
- parameters
- type
- evaluated
- 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
- 238000011156 evaluation Methods 0.000 title claims abstract description 51
- 238000013135 deep learning Methods 0.000 title claims abstract description 46
- 230000006870 function Effects 0.000 claims abstract description 48
- 230000003068 static effect Effects 0.000 claims abstract description 36
- 238000011161 development Methods 0.000 claims abstract description 25
- 238000012216 screening Methods 0.000 claims abstract description 23
- 238000004519 manufacturing process Methods 0.000 claims abstract description 15
- 238000012163 sequencing technique Methods 0.000 claims abstract description 7
- 239000003921 oil Substances 0.000 claims description 300
- 239000010779 crude oil Substances 0.000 claims description 10
- 230000035699 permeability Effects 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 230000008030 elimination Effects 0.000 claims description 5
- 238000003379 elimination reaction Methods 0.000 claims description 5
- 238000007637 random forest analysis Methods 0.000 claims description 5
- 230000018109 developmental process Effects 0.000 description 21
- 238000011084 recovery Methods 0.000 description 13
- 238000000034 method Methods 0.000 description 12
- 238000009826 distribution Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 4
- 230000035515 penetration Effects 0.000 description 4
- 239000011435 rock Substances 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000008021 deposition Effects 0.000 description 3
- BVKZGUZCCUSVTD-UHFFFAOYSA-L Carbonate Chemical compound [O-]C([O-])=O BVKZGUZCCUSVTD-UHFFFAOYSA-L 0.000 description 2
- 230000009471 action Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 239000008398 formation water Substances 0.000 description 2
- 239000000295 fuel oil Substances 0.000 description 2
- 239000007789 gas Substances 0.000 description 2
- 239000003345 natural gas Substances 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 230000033558 biomineral tissue development Effects 0.000 description 1
- -1 claystone Substances 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000000556 factor analysis Methods 0.000 description 1
- 239000012530 fluid Substances 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000003209 petroleum derivative Substances 0.000 description 1
- 239000011026 pyroclastic rock Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000013049 sediment Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 239000000341 volatile oil Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- General Business, Economics & Management (AREA)
- Evolutionary Biology (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Tourism & Hospitality (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Mining & Mineral Resources (AREA)
- Marine Sciences & Fisheries (AREA)
- Animal Husbandry (AREA)
- Agronomy & Crop Science (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Edible Oils And Fats (AREA)
Abstract
The invention relates to an oil field analogy evaluation method and system based on deep learning, which comprises the following steps: sl: acquiring parameters of an oil field to be evaluated and an oil field analogy; s2: determining one of static characteristic parameters of the oil field to be evaluated and/or oil field development and production parameters as an analog objective function; s3: screening out N first-type oil fields with the same type as the oil field to be evaluated according to the first-type parameters in the parameters of the step S1; s4: calculating the similarity of the N first-type oil fields and the oil fields to be evaluated according to the respective second-type parameters of the N first-type oil fields and the oil fields to be evaluated, and sequencing the N first-type oil fields according to the size of the similarity; s5: and determining the target value of the oil field to be evaluated in the target value interval of the analog target function applied to the first analog oil field in different sets according to the similarity sequence. The analogy evaluation method and the analogy evaluation system can improve the analogy evaluation working efficiency and can improve the accuracy of the analogy target function recommendation value.
Description
Technical Field
The invention relates to the technical field of oil and gas development, in particular to an oil field analogy evaluation method based on deep learning and an oil field analogy evaluation system based on deep learning.
Background
In exploration investment evaluation and overseas asset co-purchasing, an analogy evaluation method provides conditions for absorbing global experience, knowing the basic conditions of oil fields in an area and quickly framing the development scale and economic value of a target area. In the new oil field development scheme compilation, the analogy evaluation method provides a basis for the optimization of the development strategy and the development mode of the oil field to be evaluated and the reasonable prediction of the development index. But is influenced by the abundant degree of working experience and the complete and detailed degree of mastering oil field data, and in the aspect of similar oil field screening, the influence factor is considered to be single or only a few factors, and the judgment of the main control factor for the development of the target area is lacked; in similarity quantitative calculation, the weight of the influence factors is considered by a few methods, but the method is only limited to the relevance between every two parameters, and the influence of the combined action of multiple factors is not considered.
Disclosure of Invention
In view of the above problems, an object of the present invention is to provide an oil field analogy evaluation method and system based on deep learning, which can improve the working efficiency of oil field geological reservoir analogy evaluation and improve the accuracy of the analogy objective function recommendation.
In order to achieve the purpose, the invention adopts the following technical scheme: an oil field analogy evaluation method based on deep learning comprises the following steps:
s1: acquiring parameters of an oil field to be evaluated and an oil field analogy;
s2: determining one of static characteristic parameters of the oil field to be evaluated and/or oil field development and production parameters as an analog objective function;
s3: screening out N first-type oil fields with the same type as the oil field to be evaluated according to the first-type parameters in the parameters of the step S1;
s4: calculating the similarity of the N first-type oil fields and the oil fields to be evaluated according to the respective second-type parameters of the N first-type oil fields and the oil fields to be evaluated, and sequencing the N first-type oil fields according to the size of the similarity;
s5: and determining the target value of the oil field to be evaluated according to the similarity ranking in the target value interval of the analog target functions of the first analog oil fields in different sets.
In the deep learning-based oilfield analogy evaluation method, preferably, the parameters in the step S1 include: oil field location, geological static parameters, reservoir static parameters, and oil field development production parameters.
In the deep learning-based oilfield analogy evaluation method, preferably, the first type of parameters in the step S3 include any one or more of a depositional type, lithology, reservoir depth, permeability rating, and crude oil property belonging to geological static parameters and reservoir static parameters.
In the deep learning-based oilfield analogy evaluation method, preferably, the second type of parameters in the step S4 include geological static parameters and reservoir static parameters.
Preferably, the oil field analogy evaluation method based on deep learning in step S4 specifically includes the following steps:
s41: quantifying the influence degree of the second type parameter of each first type oil field on the analog target function by using an algorithm combining deep learning random forest and recursive feature elimination to obtain the weight of each parameter in the second type parameter of each first type oil field;
s42: calculating the similarity of each first-type oil field and the oil field to be evaluated:
wherein S is the similarity between the oil field to be evaluated and one of the N first-type oil fields; xiFor the ith parameter, X, of the second parameters of one of the N first-type oil fieldsiBThe parameter is the ith parameter in the second type parameters of the oil field to be evaluated; n is the number of the second kind of parameters, WiThe weight of the ith parameter in the second parameters of one of the N first-type oil fields is set;
s43: and sorting the oil fields to be evaluated and the N first-class oil fields according to the similarity S calculated in the step S42 from large to small.
Preferably, the oil field analogy evaluation method based on deep learning in step S5 specifically includes the following steps:
s51: respectively screening out a first-class oil field with the similarity larger than K% and a first-class oil field with the similarity in different sections as a first set and a first-class oil field of a second set;
s52: calculating the quartile value, the binary value and the three-quarter value of the analog objective function of the first analog oil field screened in the step S51 as the target value interval of the analog objective function;
s53: and determining the target value of the oil field to be evaluated according to the target value interval of the analog target function in the step S52.
Preferably, the oil field analogy evaluation method based on deep learning in step S51 specifically includes the following steps:
s511: respectively screening out first-class oil fields with similarity degrees of more than 35%, more than 50%, more than 55%, more than 60%, more than 65% and more than 70% as a first set of first-class oil fields;
s512, screening the first-class oil fields with the similarity of 35% -50%, 50% -55%, 55% -60%, 60% -65%, 65% -70% and more than 70% respectively to serve as first-class oil fields of a second set.
Preferably, the oil field analogy evaluation method based on deep learning in step S53 specifically includes the following steps:
and taking the interval range of the binary values of the analog objective function of the first analog oilfield with the similarity of more than 60%, more than 65% or more than 70% acquired in the step S511 as the recommended interval value of the target value interval of the oilfield to be evaluated.
Preferably, the oil field analogy evaluation method based on deep learning in step S53 specifically includes the following steps:
and taking the average value of the binary values of the analog objective function of the first analog oil field with the similarity of 60% -65%, 65% -70% or more than 70% acquired in the step S512 as the recommended average value of the target value interval of the oil field to be evaluated.
On the other hand, the invention also provides an oil field analogy evaluation system based on deep learning, which comprises the following steps:
the acquisition module is used for acquiring the oil field to be evaluated and the parameters for simulating the oil field;
the method comprises the steps of obtaining an analogy objective function module, and determining one oil field static characteristic parameter and/or oil field development and production parameter of an oil field to be evaluated as an analogy objective function;
the screening module is used for screening N first-type oil fields with the same type as the oil field to be evaluated according to the first-type parameters in the acquisition module;
the sorting module is used for calculating the similarity of the N first-type oil fields and the oil fields to be evaluated according to the respective second-type parameters of the N first-type oil fields and the oil fields to be evaluated and sorting the N first-type oil fields according to the similarity;
and the target value module is used for determining the target value of the oil field to be evaluated in the target value interval of the analog target function of the first analog oil field in different sets according to the similarity sequence.
Due to the adoption of the technical scheme, the invention has the following advantages:
1. the oil field to be developed is evaluated according to a plurality of parameters of the existing developed oil field, so that the accuracy of the evaluation work of the oil field geological oil deposit is improved;
2. the weights of multiple parameters are considered in the evaluation work, the relevance between every two parameters is not limited, the influence of the combined action of the multiple parameters is considered, the precision of the recommendation value of the analogy objective function is improved, and the application in the field of petroleum and natural gas development analogy evaluation is facilitated.
Drawings
FIG. 1 is a schematic diagram of the steps of the deep learning-based oilfield analogy evaluation method of the present invention;
FIG. 2 is a schematic diagram of the steps of calculating and ranking the similarity between the simulated oil field and the oil field to be evaluated according to the present invention;
FIG. 3 is a schematic diagram of the steps of determining the target values of the oil fields to be evaluated after ranking according to similarity in accordance with the present invention;
FIG. 4 is a schematic representation of the steps of the present invention to screen out the first category of fields as being in a different set;
FIG. 5 is a graph of the mean probability distribution of oil column heights of the present invention;
FIG. 6 is a productivity prediction diagram for different similarity intervals of the first set according to the present invention;
FIG. 7 is a second set capacity prediction graph for different similarity intervals according to the present invention;
FIG. 8 is a first set of different-similarity interval recovery prediction plots of the present invention;
FIG. 9 is a second set of different-similarity interval recovery prediction plots of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the description of the present invention, the terms "first," "second," "third," and the like are used to define steps, methods, or groups, etc. merely to distinguish one element from another, and unless otherwise indicated, the terms have no special meaning and are not intended to indicate or imply relative importance.
The oil field analogy evaluation method and system based on deep learning provided by the embodiment of the invention can not only improve the working efficiency of oil field geological oil deposit analogy evaluation, but also improve the accuracy of the analogy target function recommendation value.
The invention adopts the following technical scheme: an oil field analogy evaluation method based on deep learning comprises the following steps:
s1: acquiring parameters of an oil field to be evaluated and an oil field analogy; the embodiment of the invention can be applied to the technical field of oil and natural gas development. In actual practice, the field to be evaluated and the analogous field include offshore fields and/or onshore fields. Establishing an oil field geological oil reservoir analogy application database: the collected oilfield actual data includes: oil field location, geological static parameters, reservoir static parameters, and oil field development parameters. The geological static parameters include: deposit type, formation type, lithology, displacement pressure type, oil column height, original reservoir pressure, original reservoir temperature, reservoir pressure gradient, reservoir depth, original geological reserve, folded oil-bearing area, average porosity, average permeability, total reservoir thickness, average effective thickness, net-to-gross ratio, original oil saturation; the reservoir static parameters include: bubble point pressure, formation crude oil viscosity, average API degree, original oil-gas ratio, crude oil volume coefficient and formation water mineralization degree; developing production parameters includes: well number, well type, single well productivity, oil field recovery ratio, peak oil production speed, and comprehensive annual rate of decline.
S2: determining one of static characteristic parameters of the oil field to be evaluated and/or oil field development and production parameters as an analog objective function; the analog target function is used for the oil field to be evaluated and is also applied to the analog oil field. The static characteristic parameters of the oil field geological oil deposit comprise: fluid property, fault plugging property and the like, and index parameters of oil field development and production comprise: single well productivity, oil field recovery, peak production rate, etc. The static characteristic parameters of the oil field geological oil deposit and/or the index parameters of oil field development and production are functions of other parameters in deep learning.
S3: screening out N first-type oil fields with the same type as the oil field to be evaluated according to the first-type parameters in the parameters of the step S1; taking first type parameters belonging to oil field positions, geological static parameters, oil reservoir static parameters and oil field development parameters as screening standards, and screening N first type oil fields with the same type as the oil field to be evaluated; the type of the oil field to be evaluated comprises any one or more than two of the deposition type, lithology, reservoir depth, permeability rating and crude oil property. The deposit type classification includes: a terrestrial phase, a marine phase, a transition phase and a carbonate phase; lithology classification includes: sandstone, conglomerate, carbonate rock, claystone, pyroclastic rock, extrusive rock, invaded rock, and metamorphic rock; the oil reservoir depth classification comprises the following steps: a shallow layer (less than 1500m), a medium deep layer (1500 + 2800m), a deep layer (2800 + 4000m) and an ultra deep layer (more than 4000 m); the permeability grading comprises the following steps: high penetration (greater than 500md), medium penetration (50-500md), low penetration (10-50md) and ultra-low penetration (less than 10 md); the crude oil property classifications include volatile oils, black oils, heavy oils, and oil sands. The screened N first-type oil field types also comprise any one or more than two of a sediment type, lithology, reservoir depth, permeability grading and crude oil property. Specific examples include, for example: the method is characterized in that the type of an oil field to be evaluated is defined comprehensively by five factors of deposition type, lithology, oil reservoir depth, permeability grading and crude oil property. And screening out alternative analogy oilfield sets in the step S1 application database by respectively using 1 type, 2 types, 3 types, 4 types and 5 types of standard combinations, wherein the combination mode is shown in Table 1, and 32 alternative analogy oilfield sets with different capacities can be obtained. The capacity of the subset is minimum by adopting 1 and 4 types of constraints, namely the number of contained oil fields is minimum, and the capacity of the summation subset is maximum by adopting 5 types of constraints, so that the number of contained oil fields is maximum. Wherein Sum (1) represents a collection with 1 classification standard terms.
TABLE 1
S4: calculating the similarity of the N first-type oil fields and the oil fields to be evaluated according to the respective second-type parameters of the N first-type oil fields and the oil fields to be evaluated, and sequencing the N first-type oil fields according to the size of the similarity; if the similarity is determined, the number of first-category fields to be evaluated can be further reduced according to the similarity.
S5: and determining the target value of the oil field to be evaluated according to the similarity ranking in the target value interval of the analog target functions of the first analog oil fields in different sets. After the similarity is sorted, only the target values of the analog target functions of the first analog oil fields belonging to the first set and the second set are taken, and the target values of the oil fields to be evaluated are determined more accurately while the calculated amount is reduced.
In the deep learning-based oilfield analogy evaluation method, preferably, the parameters in the step S1 include: oil field location, geological static parameters, reservoir static parameters, and oil field development production parameters.
In the deep learning-based oilfield analogy evaluation method, preferably, the first type of parameters in the step S3 include any one or more of a depositional type, lithology, reservoir depth, permeability rating, and crude oil property belonging to geological static parameters and reservoir static parameters.
In the deep learning-based oilfield analogy evaluation method, preferably, the second type of parameters in the step S4 include geological static parameters and reservoir static parameters.
Preferably, the oil field analogy evaluation method based on deep learning in step S4 specifically includes the following steps:
s41: quantifying the influence degree of the second type parameter of each first type oil field on the analog target function by using an algorithm combining deep learning random forest and recursive feature elimination to obtain the weight of each parameter in the second type parameter of each first type oil field; before an algorithm combining deep learning random forest and recursive feature elimination is used, in order to better apply a deep learning method to develop oil field development main control factor analysis, cleaning second parameter data of a first-class oil field according to the physical significance (such as porosity, original oil saturation and net-to-gross ratio all less than 1) of second parameters, applying a sampling sample distribution rule of the second parameters, and solving the mean value of the second parameters and 95% confidence intervals under T distribution to supplement the vacancy values of the parameter items.
S42: calculating the similarity of each first-type oil field and the oil field to be evaluated:
wherein S is the similarity between the oil field to be evaluated and one of the N first-type oil fields; xiFor the ith parameter, X, of the second parameters of one of the N first-type oil fieldsiBThe parameter is the ith parameter in the second type parameters of the oil field to be evaluated; n is the number of the second kind of parameters, WiThe weight of the ith parameter in the second parameters of one of the N first-type oil fields is set;
s43: and according to the similarity S between the oil field to be evaluated and the N first-class oil fields calculated in the step S42, sorting the oil fields to be evaluated into the first-class oil fields from large to small according to the similarity S. According to actual needs, other sorting methods of the similarity S can be adopted for sorting; the purpose of the different sorting methods is to screen out the required analog fields.
Preferably, the oil field analogy evaluation method based on deep learning in step S5 specifically includes the following steps:
s51: respectively screening out a first-class oil field with the similarity larger than K% and a first-class oil field with the similarity in different sections as a first-class oil field in a first set and a first-class oil field in a second set; the value of K is as follows, and the first-class oil fields which are sorted according to the similarity S from large to small and are greater than K% are screened.
S52: calculating the quartile value, the binary value and the three-quarter value of the analog objective function of the first analog oil field screened in the step S51 as the target value interval of the analog objective function; the quartile, binary, and ternary may also be referred to as: a first quartile, a second quartile, and a third quartile. Can also be expressed as: q25%, Q50%, Q75%.
S53: and determining the target value of the oil field to be evaluated according to the target value interval of the analog target function in the step S52.
Preferably, the oil field analogy evaluation method based on deep learning in step S51 specifically includes the following steps:
s511: respectively screening out first-class oil fields with similarity degrees of more than 35%, more than 50%, more than 55%, more than 60%, more than 65% and more than 70% as a first set of first-class oil fields; the following table 2 is specifically provided:
TABLE 2
S512: the first category oil fields with the similarity of 35% -50%, 50% -55%, 55% -60%, 60% -65%, 65% -70% and more than 70% are respectively screened out to be used as the first category oil fields of the second set, and the first category oil fields are specifically shown in the table 2.
Preferably, the oil field analogy evaluation method based on deep learning in step S53 specifically includes the following steps:
and taking the interval range of the binary values of the analog objective function of the first analog oilfield with the similarity of more than 60%, more than 65% or more than 70% acquired in the step S511 as the recommended interval value of the target value interval of the oilfield to be evaluated. That is, the recommended value interval range is obtained from the Q50% data with the "first set" similarity being greater than 60%, greater than 65%, or greater than 70%. When the number of the first type is larger than that of the oil fields, the standard with the similarity degree larger than 60% can be adopted, and the standard with the similarity degree larger than that can be adopted, such as 70%, 75% or larger. If the number of first type versus field is small, the criteria can be relaxed appropriately to a small degree of similarity, such as 50%, 40% or less.
Preferably, the oil field analogy evaluation method based on deep learning in step S53 specifically includes the following steps:
and taking the average value of the binary values of the analog objective function of the first analog oil field with the similarity of 60% -65%, 65% -70% or more than 70% acquired in the step S512 as the recommended average value of the target value interval of the oil field to be evaluated. Namely, the average value of Q50% with the similarity of the second set being 60% -65%, 65% -70% or more than 70% is used to obtain the recommended average value.
On the other hand, the invention also provides an oil field analogy evaluation system based on deep learning, which comprises the following steps:
the acquisition module is used for acquiring the oil field to be evaluated and the parameters for simulating the oil field;
the method comprises the steps of obtaining an analogy objective function module, and determining one oil field static characteristic parameter and/or oil field development and production parameter of an oil field to be evaluated as an analogy objective function;
the screening module is used for screening N first-type oil fields with the same type as the oil field to be evaluated according to the first-type parameters in the acquisition module;
the sorting module is used for calculating the similarity of the N first-type oil fields and the oil fields to be evaluated according to the respective second-type parameters of the N first-type oil fields and the oil fields to be evaluated and sorting the N first-type oil fields according to the similarity;
and the target value module is used for determining the target value of the oil field to be evaluated in the target value interval of the analog target function of the first analog oil field in different sets according to the similarity sequence.
The specific embodiment is as follows:
the present invention will be described in detail with reference to examples.
The invention relates to an oil field analogy evaluation method based on deep learning, which comprises the following steps of:
step (1) collecting 1334 actual data of the oil field, and establishing an oil field geological oil reservoir analogy application database: the collected oilfield actual data includes: oil field position (offshore/onshore), geological static parameters 17, reservoir static parameters 6, development parameters 6;
step (2) determination of an analog objective function: the oil field to be evaluated is currently in an early stage research stage, the basic conditions of the oil field are shown in table 3, and the predicted recommended values of the initial productivity of a single well and the recovery ratio of the oil field are determined by the analog evaluation method.
TABLE 3 basic conditions of the oil field to be evaluated
And (3) defining the type of the oil field to be evaluated: the method is characterized in that the type of an oil field to be evaluated is defined comprehensively by five factors of deposition type, lithology, oil reservoir depth, permeability grading and crude oil property. And screening the analogy application database in the step (1) to obtain 32 alternative analogy oil field subsets under the condition of each oil field type combination. The number of fields each subset contains is shown in table 4.
The oil field to be evaluated belongs to: transition phase-shallow water plait river delta, continental phase-meandering river, sandstone, shallow-middle layer, high-permeability, black oil-heavy oil reservoir.
TABLE 4 alternative analogy oilfield set Generation results
And (4) taking the alternative analog oilfield subsets 23 marked in the table 4 as an example, cleaning the data of each parameter in the subsets according to the physical significance of each parameter, and solving the mean value and 95% confidence interval of each parameter to complement the vacancy value of each parameter item by applying the distribution rule of the sampling samples of each parameter. For example, the mean probability distribution of the oil column height (unit: ft) is shown in FIG. 5, and the 95% confidence intervals are [538, 920] ft, respectively.
The data of each parameter is cleaned and supplemented as shown in the table 5; the bold face labeled data in table 5 is the data filled up by applying the method; tests show that when the percentage of missing data in a certain parameter is more than 30%, the analysis result is greatly influenced.
TABLE 5
And (5) sequencing and quantifying the correlation degree of the static parameters of the geological oil deposit in each alternative analog oil field set to the analog objective function by applying an algorithm combining deep learning random forest and recursive characteristic elimination, and calculating to obtain the weight value of the influence of each parameter on the analog objective function. The average bubble point pressure, the formation water salinity, the single well control area, the average volume coefficient and the average Kh/u value are 36.1 percent, 37.8 percent, 31.9 percent, 31.1 percent and 30.3 percent respectively. In order to ensure the reliability of the analysis result, the actual oil field data is taken as the main data, the 5 parameters are specially excluded in the analysis, and when the data is rich enough, the result can be updated by using the method. The sequence and the weight of the influence degree of the 20 parameters on the initial productivity and the recovery ratio of the oil field are respectively calculated and obtained and are shown in tables 6 to 7.
TABLE 6 oil field initial Capacity influencing factor ranking and weighting
TABLE 7 oil field recovery factor ordering and weighting
And (6) calculating the similarity between the alternative analog oil field and the oil field to be evaluated. Limited by the data in the application database, the initial capacity of a single well of the oil field is analogous to 54 alternative analogous oil fields obtained in the alternative analogous oil field subset 23, and the recovery rate of the oil field is analogous to 94 alternative analogous oil fields obtained in the alternative analogous oil field subset 23. Respectively calculating the similarity between each type of oil field and the oil field to be evaluated by using the following calculation formulas:
wherein S is the similarity between the oil field to be evaluated and one of the N first-type oil fields and is a decimal; xiFor the ith parameter, X, of the second parameters of one of the N first-type oil fieldsiBThe parameter is the ith parameter in the second type parameters of the oil field to be evaluated; n is the number of the second kind of parameters, WiFor N first-class oil fieldsThe weight of the ith parameter in the second type parameters of the first type-specific oilfield;
the similarity and the sequencing result of 54 productivity simulation oil fields and oil fields to be evaluated are calculated and obtained and are shown in table 8, and the similarity and the sequencing result of 94 recovery simulation oil fields and oil fields to be evaluated are shown in table 9;
TABLE 8 oil field to be evaluated and alternative analogy oil field similarity and sorting result (oil field single well productivity is used as target)
TABLE 9 oil field to be evaluated and alternative analogy oil field similarity and sorting result (oil field recovery ratio is used as target)
And (7) sorting according to the similarity, and obtaining the oil fields with the similarity of more than 60% with the oil field to be evaluated in the alternative analog subset. The similarity degree of 21 alternative analog oil fields in the oil field single well productivity analog oil field subset to the oil field to be evaluated is more than 60%, wherein the similarity degree of 5 alternative analog oil fields to the oil field to be evaluated is more than 70%. The similarity degree of 20 alternative analogy oil fields in the oil field recovery analogy oil field subset and the oil field to be evaluated is more than 60%, wherein the similarity degree of 1 alternative analogy oil field and the oil field to be evaluated is more than 70%. Q25%, Q50% and Q75% values of the oilfield analog objective function in different similarity degree intervals are calculated, as shown in tables 10 and 11.
TABLE 10 statistics of productivity quantiles between different similarity intervals
TABLE 11 statistical chart of recovery ratio quantiles in different similarity degree intervals
As shown in fig. 6, 7, 8 and 9, step (8) determines the high, medium and low recommended values of the analog objective function according to the connecting areas of the values of the objective function Q25%, Q50% and Q75%. The recommended emphasis is to use the oil field set with the similarity more than 60% as the reference object, and if the number of alternative analog oil fields is less, the limit can be properly relaxed. And obtaining a recommended value interval range by using the Q50% data with the similarity of the first set being more than 60%, and obtaining a recommended mean value by using the Q50% average with the similarity of the second set being more than 60%. In the case, the recommended value range of the oil field single well productivity is [29.3, 55.5] square/day, and the recommended average value is 39.5 square/day; the recommended range of the final recovery rate of the oil field is [19.0, 23.6 ]%, and the recommended average value is 22.0%.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. An oil field analogy evaluation method based on deep learning is characterized by comprising the following steps:
s1: acquiring parameters of an oil field to be evaluated and an oil field analogy;
s2: determining one of static characteristic parameters of the oil field to be evaluated or oil field development and production parameters as an analog objective function;
s3: screening out N first-type oil fields with the same type as the oil field to be evaluated according to the first-type parameters in the parameters of the step S1;
s4: calculating the similarity of the N first-type oil fields and the oil fields to be evaluated according to the respective second-type parameters of the N first-type oil fields and the oil fields to be evaluated, and sequencing the N first-type oil fields according to the size of the similarity;
s5: and determining the target value of the oil field to be evaluated according to the similarity ranking in the target value interval of the analog target functions of the first analog oil fields in different sets.
2. The deep learning-based oilfield analogy evaluation method according to claim 1, wherein the parameters in the step S1 include: oil field location, geological static parameters, reservoir static parameters, and oil field development production parameters.
3. The deep learning-based oilfield analogy evaluation method according to claim 2, wherein the first type parameters in step S3 include any one or more of depositional type, lithology, reservoir depth, permeability rating, and crude oil properties, which depend on geological static parameters and reservoir static parameters.
4. The deep learning-based oilfield analogy evaluation method according to claim 1, wherein the second type of parameters in step S4 comprise geological static parameters and reservoir static parameters.
5. The deep learning-based oilfield analogy evaluation method according to claim 1, wherein the step S4 specifically comprises the following steps:
s41: quantifying the influence degree of the second type parameter of each first type oil field on the analog target function by using an algorithm combining deep learning random forest and recursive feature elimination to obtain the weight of each parameter in the second type parameter of each first type oil field;
s42: calculating the similarity of each first-type oil field and the oil field to be evaluated:
wherein S is the similarity between the oil field to be evaluated and one of the N first-type oil fields; xiFor the ith parameter, X, of the second parameters of one of the N first-type oil fieldsiBThe parameter is the ith parameter in the second type parameters of the oil field to be evaluated; n is the number of the second kind of parameters, WiThe weight of the ith parameter in the second parameters of one of the N first-type oil fields is set;
s43: and sorting the oil fields to be evaluated and the N first-class oil fields according to the similarity S calculated in the step S42 from large to small.
6. The deep learning-based oilfield analogy evaluation method according to claim 5, wherein the step S5 specifically comprises the following steps:
s51: respectively screening out a first-class oil field with the similarity larger than K% and a first-class oil field with the similarity in different sections as a first set and a first-class oil field of a second set;
s52: calculating the quartile value, the binary value and the three-quarter value of the analog objective function of the first analog oil field screened in the step S51 as the target value interval of the analog objective function;
s53: and determining the target value of the oil field to be evaluated according to the target value interval of the analog target function in the step S52.
7. The deep learning-based oilfield analogy evaluation method according to claim 6, wherein the step S51 specifically comprises the following steps:
s511: respectively screening out first-class oil fields with similarity degrees of more than 35%, more than 50%, more than 55%, more than 60%, more than 65% and more than 70% as a first set of first-class oil fields;
s512: and respectively screening the first-class oil fields with the similarity of 35% -50%, 50% -55%, 55% -60%, 60% -65%, 65% -70% and more than 70% as a second set of first-class oil fields.
8. The deep learning-based oilfield analogy evaluation method according to claim 7, wherein the step S53 specifically comprises the following steps:
and taking the interval range of the binary values of the analog objective function of the first analog oilfield with the similarity of more than 60%, more than 65% or more than 70% acquired in the step S511 as the recommended interval value of the target value interval of the oilfield to be evaluated.
9. The deep learning-based oilfield analogy evaluation method according to claim 7, wherein the step S53 specifically comprises the following steps:
and taking the average value of the binary values of the analog objective function of the first analog oil field with the similarity of 60% -65%, 65% -70% or more than 70% acquired in the step S512 as the recommended average value of the target value interval of the oil field to be evaluated.
10. An oil field analogy evaluation system based on deep learning is characterized by comprising:
the acquisition module is used for acquiring the oil field to be evaluated and the parameters for simulating the oil field;
the analog objective function module is used for determining one oil field static characteristic parameter and/or oil field development and production parameter of the oil field to be evaluated as an analog objective function;
the screening module is used for screening N first-type oil fields with the same type as the oil field to be evaluated according to the first-type parameters in the acquisition module;
the sorting module is used for calculating the similarity of the N first-type oil fields and the oil fields to be evaluated according to the respective second-type parameters of the N first-type oil fields and the oil fields to be evaluated and sorting the N first-type oil fields according to the similarity;
and the target value module is used for determining the target value of the oil field to be evaluated in the target value interval of the analog target function of the first analog oil field in different sets according to the similarity sequence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110605795.9A CN113240321A (en) | 2021-05-31 | 2021-05-31 | Oil field analogy evaluation method and system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110605795.9A CN113240321A (en) | 2021-05-31 | 2021-05-31 | Oil field analogy evaluation method and system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113240321A true CN113240321A (en) | 2021-08-10 |
Family
ID=77135983
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110605795.9A Pending CN113240321A (en) | 2021-05-31 | 2021-05-31 | Oil field analogy evaluation method and system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113240321A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103218750A (en) * | 2013-04-03 | 2013-07-24 | 中国海洋石油总公司 | Geology evaluation method for offshore sedimentary rock oil field |
| CN106530109A (en) * | 2016-12-21 | 2017-03-22 | 中海石油(中国)有限公司 | Oilfield development appraisal well decision method based on information value |
| CN106897821A (en) * | 2017-01-24 | 2017-06-27 | 中国电力科学研究院 | A kind of transient state assesses feature selection approach and device |
| CN107133879A (en) * | 2016-02-26 | 2017-09-05 | 中国石油化工股份有限公司 | A kind of method for screening similar oil field |
| CN110992200A (en) * | 2019-12-11 | 2020-04-10 | 长江大学 | Shale gas well staged fracturing effect evaluation and yield prediction method based on random forest |
| CN112069567A (en) * | 2020-08-07 | 2020-12-11 | 湖北交投十巫高速公路有限公司 | Method for predicting compressive strength of concrete based on random forest and intelligent algorithm |
| CN112070356A (en) * | 2020-08-07 | 2020-12-11 | 湖北交投十巫高速公路有限公司 | Method for predicting anti-carbonization performance of concrete based on RF-LSSVM model |
| CN112819207A (en) * | 2021-01-19 | 2021-05-18 | 武汉中地云申科技有限公司 | Geological disaster space prediction method, system and storage medium based on similarity measurement |
-
2021
- 2021-05-31 CN CN202110605795.9A patent/CN113240321A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103218750A (en) * | 2013-04-03 | 2013-07-24 | 中国海洋石油总公司 | Geology evaluation method for offshore sedimentary rock oil field |
| CN107133879A (en) * | 2016-02-26 | 2017-09-05 | 中国石油化工股份有限公司 | A kind of method for screening similar oil field |
| CN106530109A (en) * | 2016-12-21 | 2017-03-22 | 中海石油(中国)有限公司 | Oilfield development appraisal well decision method based on information value |
| CN106897821A (en) * | 2017-01-24 | 2017-06-27 | 中国电力科学研究院 | A kind of transient state assesses feature selection approach and device |
| CN110992200A (en) * | 2019-12-11 | 2020-04-10 | 长江大学 | Shale gas well staged fracturing effect evaluation and yield prediction method based on random forest |
| CN112069567A (en) * | 2020-08-07 | 2020-12-11 | 湖北交投十巫高速公路有限公司 | Method for predicting compressive strength of concrete based on random forest and intelligent algorithm |
| CN112070356A (en) * | 2020-08-07 | 2020-12-11 | 湖北交投十巫高速公路有限公司 | Method for predicting anti-carbonization performance of concrete based on RF-LSSVM model |
| CN112819207A (en) * | 2021-01-19 | 2021-05-18 | 武汉中地云申科技有限公司 | Geological disaster space prediction method, system and storage medium based on similarity measurement |
Non-Patent Citations (3)
| Title |
|---|
| 丁祖鹏等: ""稠油油藏类比方法研究"", 《特种油气藏》, vol. 20, no. 4, pages 1 * |
| 冯晓荣等: ""基于深度学习与随机森林的高维数据特征选择"", 《计算机工程与设计》, vol. 40, no. 9, pages 5 * |
| 王诚等: ""基于特征约简的随机森林改进算法研究"", 《计算机技术与发展》, vol. 30, no. 3 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110276827B (en) | Effectiveness evaluation method based on shale reservoir | |
| CN108843286B (en) | Technical method for layered oil production well selection | |
| CN105134191B (en) | Method for evaluating reserves of tight oil well | |
| CN108009716A (en) | A kind of horizontal well volume fracturing influential effect factor mutiple-stage model method | |
| CN105134195A (en) | Shale gas reservoir quality evaluation method based on logging information | |
| CN112487582A (en) | Oil-gas drilling machinery drilling speed prediction and optimization method based on CART algorithm | |
| CN108846540A (en) | Recovery ratio calibration method and device for tight sandstone gas field | |
| Mirzavand et al. | Saltwater intrusion vulnerability assessment using AHP-GALDIT model in Kashan plain aquifer as critical aquifer in a semi-arid region | |
| CN107165621B (en) | Method for identifying multi-well sedimentary microfacies by using clustering method | |
| Male et al. | Forecasting production from Bakken and three forks wells using a segregated flow model | |
| CN109403962A (en) | Oil reservoir block Monitoring Indexes association analysis method | |
| CN113033648A (en) | Method for realizing logging interpretation by using machine learning algorithm | |
| CN114254960A (en) | Evaluation method of oil-gas resources in low-exploration-degree area | |
| Ikpeka et al. | Application of machine learning models in predicting initial gas production rate from tight gas reservoirs | |
| Negahdari et al. | Application of discriminant analysis for studying the source rock potential of probable formations in the Lorestan Basin, Iran | |
| Al-Khazraji et al. | Development of heterogeneous immature Brownfield with Waterdrive using dynamic opportunity index: a case study from Iraqi oilfields | |
| Feng et al. | Interpretable Lost Circulation Analysis: Labeled, Identified, and Analyzed Lost Circulation in Drilling Operations | |
| CN105718720A (en) | Complex gas reservoir reserve quality classification comprehensive evaluation method | |
| CN113240321A (en) | Oil field analogy evaluation method and system based on deep learning | |
| Ariadji et al. | Optimization of vertical well placement for oil field development based on basic reservoir rock properties using a genetic algorithm | |
| CN116562428A (en) | Fracturing construction parameter optimization method based on machine learning | |
| CN110930020B (en) | Method for determining economic recoverable resource amount of unconventional oil and gas resources | |
| CN111287739B (en) | Residual oil distribution prediction method based on stratum crude oil viscosity | |
| Amer et al. | Permeability Prediction and Facies Distribution for Yamama Reservoir in Faihaa Oil Field: Role of Machine Learning and Cluster Analysis Approach | |
| CN115964961A (en) | Horizontal well water plugging and well selection decision-making method |
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 |