A kind of monitoring polymer drives the method for early warning of fluid-channeling channel development between injection-production wellTechnical field
The present invention relates to the method for early warning that a kind of monitoring polymer drives fluid-channeling channel development between injection-production well, belong to oil gas field and openDynamic parameter monitoring technical field during hair.
Background technology
Oil and gas reservoir in forming process by depositional environment, diagenesis and tectonism due to being influenceed, its space pointAll there is anisotropism in cloth, it is influence underground oil and gas water transport with built-in attribute (such as porosity, permeability and hole mechanism)The key factor of the dynamic and rate of oil and gas recovery.And by waterflooding extraction for many years, due to injecting the long-term erosion of water, cause trickleGrain migration and formation sand production, form current fluid-channeling channel between injection well and producing well, exacerbate the anisotropism on stratum.AlterAfter circulation road is formed, injection water is advanced by leaps and bounds readily along fluid-channeling channel occurs channelling phenomenon, and middle less permeable layer but absorbs water seldom, indulgesIt is difficult to improve to sweep efficiency, inefficient is circulated injection water in the earth formation, has a strong impact on water flooding effectiveness and oil field developmentOverall efficiency.For polymer flooding develop oil reservoir there is also identical channelling problem, cause indivedual oil wells to see poly-, output too earlyPolymer residual is high in liquid, not only have impact on the development effectiveness of polymer flooding but also had added the intractability of production fluid.
The developmental state for grasping fluid-channeling channel in reservoir be formulate the steady oil control water measure scheme of profile control, water blockoff etc. it is important according toAccording to.Existing fluid-channeling channel sentences knowledge method, mainly divides production performance observation, well testing identification, tracer monitoring, well-log information identificationEtc. several classes, these methods are required for substantial amounts of test data, obtain cycle length, cost height, and influence oil field is produced, and can only be rightOne time point carries out sentencing knowledge, it is impossible to carry out early warning to Later development trend.
The content of the invention
In view of the above-mentioned problems, developed it is an object of the invention to provide fluid-channeling channel between a kind of monitoring polymer drive injection-production wellMethod for early warning, this method both can be carried out sentencing knowledge to the development position and degree of fluid-channeling channel at this stage, and the later stage can also be alteredThe developmental state of circulation road is predicted, and is that the optimization in the development process of polymer flooding oil field and adjustment provide theoretical direction, fromAnd polymer displacement of reservoir oil efficiency is improved, mining-employed reserves is improved to greatest extent.
To achieve the above object, the present invention uses following technical scheme：Channelling is led between a kind of monitoring polymer drives injection-production wellThe method for early warning of road development, it is characterised in that comprise the following steps：
1) alert source is found
Dynamic static data and monitoring materials based on subject oil field, every ginseng of statistical analysis oil well, injection well and well groupNumerical value, finds the influence factor and specific manifestation feature that fluid-channeling channel is formed in reservoir；
2) alert million are analyzed
For the step 1) in the influence factor and each parameter of specific manifestation feature of the fluid-channeling channel that count, applicationGrey Incidence Analysis determines the Dominated Factors of influence crossflow of polymer flooding, and crossflow of polymer flooding early warning is constituted by Dominated FactorsIndex system；
3) knowledge police's degree is sentenced
Set up to gather using fuzzy synthetic evaluation model and alter Alarm Assessment model；
4) police's degree is predicted
Step 2) in determine crossflow of polymer flooding pre-warning indexes system be made up of Static State Index and dynamic indicator two parts,Wherein, dynamic indicator can change with production process, therefore by being predicted to dynamic indicator, so as to later stage channellingThe development trend of passage is predicted.
The step 1) in formation fluid-channeling channel influence factor and specific manifestation feature be subject oil field polymerizationThing drives reservoir static parameter and characteristic of production dynamic parameter in region, and it includes：Permeability grade, cumulative injection, accumulation productionLiquid measure, zero dimension see the poly- time, produce poly- concentration, produce poly- change rate of concentration, fluid productivity index rate of change, apparent water injectivity index rate of change,Note adopts connective, crossflow of polymer flooding intensity factor；
Wherein, apparent water injectivity index rate of change refers to the ratio of apparent water injectivity index at initial stage and current apparent water injectivity index；
Crossflow of polymer flooding intensity factor is used for the power for characterizing crossflow of polymer flooding, and symbol represents that it calculates public using MFormula is as follows：
M=Qp×Vp×ω (1)
In formula：QpFor the poly- amount of zero dimension production, gather amount equal to oil well production divided by split the poly- amount of dispensing；VpFor polymer tempo of penetration,Equal to well spacing divided by see the poly- time；ω is the factor that takes effect, and when seeing that the poly- time is earlier than responding time, ω=responding time/(takes effectWhen m- see the poly- time), when seeing that the poly- time is equal to responding time, ω=1, when seeing that the poly- time is later than responding time, ω=(see it is poly- when m- responding time)/see the poly- time.
The step 2) in set up crossflow of polymer flooding pre-warning indexes system detailed process it is as follows：
1. using step 1) the crossflow of polymer flooding intensity factor that calculates is used as reference sequences；
2. using step 1) in remaining dynamic and static state parameters be used as comparative sequences；
3. the degree of association of comparative sequences and reference sequences is calculated using Grey Incidence Analysis, the degree of association is bigger, explanationComparative sequences are more consistent with the situation that reference sequences change, and do normalized to the degree of association on this basis, comparative sequences withThe degree of association of reference sequences then can form crossflow of polymer flooding pre-warning indexes system for more than 0.7 as Dominated Factors.
The step 3) in set up the poly- detailed process for altering Alarm Assessment model as follows：
1. the exemplary value simulation model of different basic permeabilities is set up according to the relevant parameter of subject oil field；
2. the exemplary value simulation model that application is set up, simulates drive liquor capacity condition in different injection of polymer respectivelyUnder, the variation tendency of the moisture content fall of polymer flooding under different permeability grades obtains step 2) polymerization that determinesThing drives the boundary value of each Dominated Factors in channelling pre-warning indexes system：
According in exemplary value simulation model, the change of the moisture content fall of permeability grade and polymer flooding becomesGesture, using permeability grade as transverse axis, moisture content fall is the longitudinal axis, and the polymer flooding set up under different permeability grades containsWater rate fall change curve, corresponding permeability grade is when the moisture content fall for defining polymer flooding is begun to declineThe poly- boundary for altering observation, when moisture content fall is more than 1.2%, corresponding permeability grade is the poly- boundary for altering development；RootAccording to the different poly- boundary definitions for altering degree, each Dominated Factors under the conditions of different injection of polymer drive liquor capacity are obtainedBoundary value；
3. be directed to each Dominated Factors parameter value application control quantity method in crossflow of polymer flooding pre-warning indexes system, i.e., it is singleDominated Factors parameter value changes according to+10% and -10%, and applying equation (1) calculates corresponding crossflow factor, different Dominated Factors pairThe influence degree of polymer channelling is different, and the weight of each index symbol A is drawn by sensitivity analysis and normalizediTableShow：
In formula：|ΔMi| change for single index parameter value after+10% and -10%, it is exhausted that corresponding crossflow factor changesTo value；N is Dominated Factors number in crossflow of polymer flooding pre-warning indexes system；I is crossflow of polymer flooding pre-warning indexes systemAny Dominated Factors；
4. last boundary value and weight by each Dominated Factors in the crossflow of polymer flooding pre-alarming system of determination substitutes into mouldPaste in Model for Comprehensive, set up it is poly- alters Alarm Assessment model, and then the poly- of each well can be altered development, poly- alter observation, nothing and gatherSituation about altering is judged respectively, is drawn to gather on well location map and is altered identifying result figure, the result figure intuitively can be found out on well location mapThe poly- distribution altered.
4. the poly- specific calculating process for altering Alarm Assessment model of middle foundation is as follows for the step：
I) Comment gathers are determined, it is poly- alter in Early-warning Model it is general Comment gathers are set to poly- alter development, poly- alter observation and without gatheringThree ranks are altered, Comment gathers are designated as：
V={ v1,v2,v3} (3)
In formula：V is Comment gathers；v1Represent oil well and alter developmental condition in poly-；v2Represent oil well and alter observation state in poly-；v3Represent oil well and poly- alter does not occur；
Ii the set of factors of influence Comment gathers, i.e. step 2) are determined) the middle crossflow of polymer flooding pre-warning indexes system determined, becauseElement collection is designated as：
U={ u1,u2,…,un} (4)
In formula：U representation polymers drive channelling pre-warning indexes system；u1,u2,…,unCorrespond crossflow of polymer flooding early warningEach Dominated Factors in index system；N is master control factor number；
Iii the boundary value of each Dominated Factors in the crossflow of polymer flooding pre-warning indexes system) 2. determined according to step, andAfter being handled by normalization method, Evaluations matrix is obtained：
In formula, ri=(ri1,ri2,ri3) it is the corresponding Evaluations matrix of any Dominated Factors；I is any Dominated Factors, i=1,2 ... ..., n；
Iv the weight of each Dominated Factors in the crossflow of polymer flooding pre-warning indexes system) 3. determined according to step, sets upWeight sets, weight sets is designated as：
A={ a1,a2,…,an} (6)
In formula：A represents weight sets；aiFor the weight of any index, i=1,2 ... ..., n；
V) the synthesis computing of application blurring mapping, obtains Comprehensive Evaluation result：
B=A × R={ b1,b2,b3} (7)
In formula：B is Comprehensive Evaluation result；b1Correspond to v in Comment gathers1Evaluation result；b2Correspond to the v in Comment gathers2Evaluation result；b3Correspond to v in Comment gathers3Evaluation result；
In Comprehensive Evaluation result, b1、b2Or b3Which numerical value is big, then oil well, which is in, corresponding in Comment gathers poly- alters stateUnder, so as to be decision making to poly- situation about altering.
The step 4) in the detailed process that is predicted to the development trend of later stage fluid-channeling channel it is as follows：
1. method for numerical simulation, support vector machine method and one or more mutually knots in gray prediction method are chosen firstClose, to step 2) in the numerical value at dynamic indicator T+n time points in crossflow of polymer flooding pre-warning indexes system be predicted；
2. will predict that obtained numerical value substitutes into step 3) set up it is poly- alter in Alarm Assessment model, obtain T+n time pointsIt is poly- to alter developmental state, draw to gather on well location map and alter the figure that predicts the outcome, the result figure can intuitively find out the poly- developmental state altered.
The present invention uses above technical scheme, and it has the following advantages that：The method for early warning of the present invention can be to polymer flooding mistakeChannelling situation in journey is monitored in real time, and to the later stage it is poly- alter development trend and be predicted, the strong operability of this method, predictionPrecision is high, and result of calculation accurately and reliably, can provide theoretical direction for the implementation of polymer flooding development phase Optimized Adjustment Measures.
Brief description of the drawings
Fig. 1 is the overall flow schematic diagram of the present invention；
Fig. 2 is J oilfield polymer floodings region well location distribution map；
Fig. 3 is the moisture content fall change curve of the polymer flooding under different permeability grades；
Fig. 4, which is that J oilfield injections are poly- after 4 years, alters identifying result；
Fig. 5, which is that J oilfield injections are poly- after 5 years half, alters early warning result；
Fig. 6, which is that J oilfield injections are poly- after 5 years half, alters actual identifying result.
Embodiment
The present invention is described in detail with reference to the accompanying drawings and examples.
The invention provides the method for early warning that a kind of monitoring polymer drives fluid-channeling channel development between injection-production well, it includes followingStep：
1) alert source is found
Dynamic static data and monitoring materials based on subject oil field, every ginseng of statistical analysis oil well, injection well and well groupNumerical value, finds the influence factor and specific manifestation feature that fluid-channeling channel is formed in reservoir.
Reservoir static parameter and characteristic of production dynamic parameter in the polymer flooding region of subject oil field are obtained, it includes：OozeSaturating rate is differential, cumulative injection, cumulative liquid production, zero dimension are shown in the poly- time, produces poly- concentration, produce poly- change rate of concentration, adopt liquid refers toNumber rate of change, apparent water injectivity index rate of change, note adopt connective, crossflow of polymer flooding intensity factor.
Wherein, apparent water injectivity index rate of change refers to the ratio of apparent water injectivity index at initial stage and current apparent water injectivity index.
Crossflow of polymer flooding intensity factor is used for the power for characterizing crossflow of polymer flooding, and symbol represents that it calculates public using MFormula is as follows：
M=Qp×Vp×ω (1)
In formula：QpFor the poly- amount of zero dimension production, gather amount equal to oil well production divided by split the poly- amount of dispensing；VpFor polymer tempo of penetration,Equal to well spacing divided by see the poly- time；ω is the factor that takes effect, and when seeing that the poly- time is earlier than responding time, ω=responding time/(takes effectWhen m- see the poly- time), when seeing that the poly- time is equal to responding time, ω=1, when seeing that the poly- time is later than responding time, ω=(see it is poly- when m- responding time)/see the poly- time.
2) alert million are analyzed
For the step 1) in the influence factor and each parameter of specific manifestation feature of the fluid-channeling channel that count, applicationGrey Incidence Analysis determines the Dominated Factors of influence crossflow of polymer flooding, and crossflow of polymer flooding early warning is constituted by Dominated FactorsIndex system.
The detailed process for setting up crossflow of polymer flooding pre-warning indexes system is as follows：
1. using step 1) the crossflow of polymer flooding intensity factor that calculates is used as reference sequences；
2. using step 1) in remaining dynamic and static state parameters be used as comparative sequences；
3. the degree of association of comparative sequences and reference sequences is calculated using Grey Incidence Analysis, the degree of association is bigger, explanationComparative sequences are more consistent with the situation that reference sequences change, and do normalized to the degree of association on this basis, it is considered that ratioCompared with sequence and the degree of association of reference sequences then corresponding tune can also be done for more than 0.7 according to actual conditions as Dominated FactorsIt is whole, form crossflow of polymer flooding pre-warning indexes system.
3) knowledge police's degree is sentenced
Using fuzzy synthetic evaluation model set up it is poly- alter Alarm Assessment model, can alter Alarm Assessment model according to poly- and quantitatively commentThe fluid-channeling channel formed during valency polymer flooding, detailed process is as follows：
1. exemplary value simulation model is set up according to the relevant parameter of subject oil field.Generally, for an actual oilField, Permeability Distribution scope is wider, in different permeability ranges, judges that the standard of fluid-channeling channel is different, therefore can be withAccording to different permeability ranges, the exemplary value simulation model of different basic permeabilities is set up；
2. the exemplary value simulation model that application is set up, simulates drive liquor capacity condition in different injection of polymer respectivelyUnder, the variation tendency of the moisture content fall of polymer flooding under different permeability grades obtains step 2) in each master controlThe boundary value of factor：
According in exemplary value simulation model, the change of the moisture content fall of permeability grade and polymer flooding becomesGesture, using permeability grade as transverse axis, moisture content fall is the longitudinal axis, and the polymer flooding set up under different permeability grades containsWater rate fall change curve, corresponding permeability grade is when the moisture content fall for defining polymer flooding is begun to declineThe poly- boundary for altering observation, corresponding permeability grade is during moisture content fall rapid decrease (decrease speed is more than 1.2%)The poly- boundary for altering development；According to the different poly- boundary definitions for altering degree, obtain driving liquor capacity condition in different injection of polymerLower correspondence step 2) in crossflow of polymer flooding pre-warning indexes system each Dominated Factors boundary value；
3. be directed to the parameter value application control quantity method of each Dominated Factors in crossflow of polymer flooding pre-warning indexes system, i.e., it is singleOne Dominated Factors parameter value changes according to+10% and -10%, and applying equation (1) calculates corresponding crossflow factor, different Dominated FactorsInfluence degree to polymer channelling is different, and the weight of each index symbol A is drawn by sensitivity analysis and normalizediRepresent：
In formula：|ΔMi| change for single index parameter value after+10% and -10%, it is exhausted that corresponding crossflow factor changesTo value；N is Dominated Factors number in crossflow of polymer flooding pre-warning indexes system；I is crossflow of polymer flooding pre-warning indexes systemAny Dominated Factors；
4. finally by the boundary value and each Dominated Factors of each Dominated Factors in the crossflow of polymer flooding pre-alarming system of determinationWeight substitute into fuzzy synthetic evaluation model, set up it is poly- alters Alarm Assessment model, and then the poly- of each well can be altered development, gatheredAlter observation, judged respectively without poly- situation about altering, draw to gather on well location map and alter identifying result figure, the result figure can intuitively be seenGo out the distribution for gathering on well location map and altering, specific calculation procedure is as follows：
First, Comment gathers are determined, it is poly- alter in Early-warning Model it is general Comment gathers are set to poly- alter development, poly- alter and observe and nothingPoly- to alter three ranks, Comment gathers can be designated as：
V={ v1,v2,v3} (3)
In formula：V is Comment gathers；v1Represent oil well and alter developmental condition in poly-；v2Represent oil well and alter observation state in poly-；v3Represent oil well and poly- alter does not occur；
Secondly, it is determined that the set of factors of influence Comment gathers, i.e. step 2) the middle crossflow of polymer flooding pre-warning indexes system determined,Set of factors can be designated as：
U={ u1,u2,…,un} (4)
In formula；U representation polymers drive channelling pre-warning indexes system；u1,u2,…,unCorrespond crossflow of polymer flooding early warningEach Dominated Factors in index system；N is master control factor number；
The boundary value of each Dominated Factors in the crossflow of polymer flooding pre-warning indexes system 2. determined according to step, and pass throughAfter normalization method processing, Evaluations matrix is obtained：
In formula, ri=(ri1,ri2,ri3) it is the corresponding Evaluations matrix of any Dominated Factors；I is any Dominated Factors, i=1,2 ... ..., n；
The weight of each Dominated Factors in the crossflow of polymer flooding pre-warning indexes system 3. determined according to step, sets up weightCollection, weight sets can be designated as：
A={ a1,a2,…,anIn (6) formula, A represents weight sets；aiFor any master controlThe weight of factor, i=1,2 ... ..., n；
Finally, using the synthesis computing of blurring mapping, Comprehensive Evaluation result can be obtained：
B=A × R={ b1,b2,b3} (7)
In formula：B is Comprehensive Evaluation result；b1Correspond to v in Comment gathers1Evaluation result；b2Correspond to the v in Comment gathers2Evaluation result；b3Correspond to v in Comment gathers3Evaluation result；
In Comprehensive Evaluation result, b1、b2Or b3Which numerical value is big, then oil well, which is in, corresponding in Comment gathers poly- alters stateUnder, so as to be decision making to poly- situation about altering.
4) police's degree is predicted
Step 2) in determine crossflow of polymer flooding pre-warning indexes system be made up of Static State Index and dynamic indicator two parts,Wherein, dynamic indicator can change with production process, therefore by being predicted to dynamic indicator, so as to later stage channellingThe development trend of passage is predicted, and detailed process is as follows：
1. method for numerical simulation, support vector machine method and one or more mutually knots in gray prediction method are chosen firstClose, the numerical value at the dynamic indicator T+n time points in crossflow of polymer flooding pre-warning indexes system is predicted；
2. will predict that obtained numerical value substitutes into step 3) set up it is poly- alter in Alarm Assessment model, obtain T+n time pointsIt is poly- to alter developmental state, draw to gather on well location map and alter the figure that predicts the outcome, the result figure can intuitively find out the poly- developmental state altered.
Embodiment：
By taking the method for early warning that fluid-channeling channel between the J oilfield polymer flooding injection-production wells of the Bohai Sea is developed as an example, the injection region domain in J oil fieldsWell location map (as shown in Figure 2), totally 28 mouthfuls of wells in region, wherein 8 mouthfuls of injection wells, 20 mouthfuls of producing wells.Shown in comprising the following steps that：
1) the characteristic of production dynamic parameter after the static parameter of statistics injection region domain reservoir and 4 years notes of implementation are poly-, calculates oilThe channelling intensity factor of well, analysis channelling intensity factor and reservoir static parameter and the relation of every characteristic of production dynamic parameter,Because static parameter and characteristic of production dynamic parameter are more, only exposition analysis result below.
The J oil fields part injection well characteristic of production dynamic parametric statistics result of table 1
| Injection well | Cumulative injection (all places) | Apparent water injectivity index rate of change (decimal) |
| J9 | 154.19 | 0.58 |
| J7 | 221.45 | 0.31 |
| J5 | 92.64 | 0.75 |
| J3 | 154.82 | 0.52 |
| J1 | 206.22 | 0.42 |
In the case where reservoir property does not change, the injectivity index change of injection well is steady, after fluid-channeling channel is formed,Bottom pressure declines when daily water-injection rate is constant, apparent water injectivity index increase.Suction is regarded by contrasting current apparent water injectivity index relative infancyThe change of aqua index gathers the possibility for altering generation to track, and gathers after altering, and apparent water injectivity index rate of change is less than 1, and it is worth smaller explanationChannelling is more severe.The result counted from table 1 can be seen that the bigger well of cumulative injection, and apparent water injectivity index rate of change is smaller, saysThe high place of bright water drive multiple, washes away more serious, also easier formation fluid-channeling channel, water absorbing capacity increase.
Table 2 is the injection water retaining in reservoir and the statistical result of connected ratio for noting poly- initial stage and note part well group after poly- 4 years.
The J oil fields part well group characteristic of production dynamic parametric statistics result of table 2
| Well group | Initial stage injection water retaining in reservoir | Injection water retaining in reservoir after 4 years | Initial stage connected ratio | Connected ratio after 4 years |
| J1 | 0.68 | 0.57 | 0.58 | 0.76 |
| J3 | 0.77 | 0.62 | 0.52 | 0.68 |
| J5 | 0.85 | 0.71 | 0.75 | 0.79 |
| J7 | 0.53 | 0.46 | 0.63 | 0.68 |
| J9 | 0.79 | 0.72 | 0.61 | 0.67 |
Can be seen that from the statistical result of table 2 it is poly- by 4 years notes, it is straight due to great amount of injection water after fluid-channeling channel is formedPractice midwifery out, cause the reduction of stratum injection water retaining in reservoir, while fluid-channeling channel directly connects oil-water well, it is very strong that note adopts relevance,Note adopts connectedness clearly, with obvious channelling feature.
Table 3 is the characteristic of production dynamic parametric statistics result of J oil fields part producing well.
The characteristic of production dynamic parametric statistics result of the J oil fields part producing well of table 3
| Pound sign | Channelling intensity factor | Moisture content (%) | Currently produce poly- concentration (mg/L) |
| J8 | 1.49 | 70.55 | 221.8 |
| J10 | 1.00 | 65.95 | 232.8 |
| J6 | 0.89 | 66.99 | 180.3 |
| J4 | 0.42 | 46.22 | 103.4 |
| J12 | 1.66 | 82.51 | 602.4 |
| J14 | 1.78 | 86.72 | 600.4 |
| J19 | 1.81 | 82.43 | 498.4 |
| J24 | 2.83 | 72.81 | 484.7 |
| J27 | 0.23 | 65.89 | 169.6 |
| J16 | 0.87 | 74.99 | 203.8 |
| J23 | 2.33 | 70.94 | 551.8 |
| J18 | 1.56 | 82.14 | 277.0 |
| J22 | 0.52 | 50.78 | 202.3 |
| J26 | 1.14 | 56.44 | 516.7 |
| J28 | 0.46 | 69.05 | 123.0 |
| J2 | 0.81 | 40.82 | 247.5 |
| J21 | 0.68 | 83.02 | 384.1 |
| J25 | 0.13 | 20.48 | 132.6 |
The moisture content of the higher well of crossflow of polymer flooding intensity factor is can be seen that from the statistical result of table 3 and production gathers denseDegree is all higher, with obvious channelling feature, illustrates the important ginseng that the crossflow of polymer flooding factor can gather situation of altering as characterizingNumber.
2) the associating to reservoir static parameter and dynamic production target and channelling intensity factor using gray relative analysis methodDegree is calculated, and show that it is permeability grade, cumulative injection, accumulation production liquid to be compared close index with channelling intensity factorAmount, zero dimension are shown in the poly- time, produce poly- concentration, fluid productivity index rate of change, apparent water injectivity index rate of change, the poly- change rate of concentration of production and noteAdopt connectedness.
3) by the statistical analysis to J oil fields permeability data, permeability is divided into 250 × 10-3μm2、500×10-3μm2、1000×10-3μm2Three ranks.It is 250 × 10 to establish basic permeability-3μm2、500×10-3μm2、1000×10-3μm2Exemplary value simulation model, poly- under the conditions of 1,2,5,8,10,12,15 of permeability grade is simulated respectively and alters situation, is obtainedTo the moisture content fall (as shown in Figure 3) of the polymer flooding under different permeability grades.
From figure 3, it can be seen that it is same it is differential under, with the rise of basic permeability, the precipitation range of polymer is notIt is disconnected to decline；For same foundation permeability, with differential increase, the precipitation range of polymer first rises to be declined afterwards.According to containingWater fall, alters boundary and is demarcated to poly-, moisture content fall begin to decline it is differential alter the area of observation coverage to be poly-, it is aqueousThe differential of rate fall rapid decrease alters serious area to be poly-, and the boundary of specific each index is shown in Table 4.
The channelling index boundary demarcation of the different differential models of table 4
On the basis of exemplary value simulation model, the poly- weight for altering index have studied by control variate method.SingleUnder desired value situation of change, i.e., according to+10% and -10% value for changing each parameter, the variable quantity of channelling intensity factor is calculated, is obtainedGo out different indexs to the poly- influence degree altered, so as to show that the weight of each index is shown in Table 5.
The poly- weight for altering evaluation index of table 5
| Sequence number | Index | Weight |
| 1 | Permeability grade | 0.19 |
| 2 | Cumulative injection | 0.08 |
| 3 | Cumulative liquid production | 0.08 |
| 4 | Produce poly- concentration | 0.06 |
| 5 | Apparent water injectivity index rate of change | 0.15 |
| 6 | Zero dimension is shown in the poly- time | 0.14 |
| 7 | Liquid production index rate of change | 0.05 |
| 8 | Produce poly- change rate of concentration | 0.17 |
| 9 | Note adopts connectedness | 0.08 |
Poly- alter is set up using fuzzy comprehensive evaluation method and sentences knowledge model, and the channelling situation to each well in region is sentencedKnow, identifying result is as shown in Figure 4.
4) the dynamic production target for obtaining grey forecasting model and supporting vector machine model prediction carries out arithmetic average, reallyNumerical value of the fixed items evaluation index after noting poly- 5 years half, substitutes into poly- alter by obtained every evaluation index predicted value and sentences knowledge model,So as to predict that obtain after 5 years half poly- alters developmental state, predict the outcome as shown in Figure 5.By the reality of every evaluation index after 5 years halfActual value substitutes into poly- alter and sentences knowledge model, calculates obtain after 5 years half poly- and alters actual developmental state, result of calculation is as shown in Figure 6.
Contrasted and found by Fig. 4 and Fig. 5, with the extension for poly- time of showwing loving care for, channelling region has expanded, the poly- well for altering development areaIncrease to 8 mouthfuls of poly- wells for altering observation area by 6 mouthfuls increases to 11 mouthfuls by 4 mouthfuls.Contrasted and found by Fig. 5 and Fig. 6 figures, early warning is tied with actualThere are 28 mouthfuls of producing wells in really different a total of 4 mouthfuls of wells, the whole district, and precision of prediction is 86%.
The present invention is only illustrated with above-described embodiment, and the choosing method of each step and each parameter is all to have becomeChange, it is all according to the principle of the invention separate step or individual parameters value to be carried out on the basis of technical solution of the present inventionImprove and equivalents, should not exclude outside protection scope of the present invention.