CN105069303A - Quantitative evaluation method of low-permeability reservoir production capacity - Google Patents

Abstract

The present invention relates to a quantitative evaluation method of a low-permeability reservoir production capacity. Steps are: performing data transformation based on a construction parameter, a log parameter, and a reservoir parameter, to obtain a derived parameter; performing data cleaning based on the construction parameter, the log parameter, and the derived parameter, and selecting a parameter closely related to the low-permeability reservoir production capacity; establishing a parameter set of a low-permeability reservoir production capacity evaluation model; analyzing a parameter obtained after low-permeability reservoir cleaning, and analyzing a correlation between all parameters and the low-permeability reservoir production capacity; selecting a parameter set has a relatively high correlation with the low-permeability reservoir production capacity, and extracting a feature sub-set for establishing the low-permeability reservoir production capacity evaluation model; establishing multiple low-permeability reservoir production capacity evaluation models; selecting an optimal low-permeability reservoir production capacity evaluation model; and integrating actual information in all aspects of a low-permeability reservoir into the optimal low-permeability reservoir production capacity evaluation model, to implement quantitative evaluation on the low-permeability reservoir production capacity. The quantitative evaluation method of a low-permeability reservoir production capacity can be widely applied to the field of oil gas well production capacity engineering technologies.

Description

A kind of low permeability reservoir production capacity method for quantitatively evaluating

Technical field

The present invention relates to a kind of oil--gas reservoir performance evaluation method, particularly about a kind of low permeability reservoir production capacity method for quantitatively evaluating.

Background technology

For conventional gas and oil reservoir, its evaluating production capacity Main Basis darcy (Darcy) law and fur coat (Dupuit) formula, be expressed as formula:

$Q=\frac{2{\mathrm{\πKH}}_e(P_e-P_{wf})}{\mathrm{\μ}(ln\frac{R_e}{R_w}+S)}---\left(1\right)$

In formula: Q is flow, unit is m ³/ s; π is circular constant, dimensionless; K is effective permeability, and unit is mD; H _efor net thickness, unit is m; P _efor terminal pressure, unit is Pa; P _wffor sand face pressure, unit is Pa; μ is fluid viscosity, and unit is Pas; R _efor effective supply radius, unit is m; R _wfor well radius, unit is m; S is skin factor, dimensionless.

From formula (1), the factor affecting Oil & Gas Productivity mainly contains: effective permeability, net thickness, drawdown pressure, effectively supply oil radius and skin factor.But under low-velocity seepage condition, percolation law no longer meets Darcy's law, therefore the influence factor of its production capacity be not limited solely to above-mentioned factor.Due to the existence of free-boundary problem and stress sensitive, Porous Flow Mechanics for Low Permeability Reservoirs is comparatively complicated, making to describe the Darcy's law of fluid steady seepage in porous medium and fur coat formula can not the seepage flow characteristics of its fluid of accurate characterization, thus make to be applicable to Darcy's law that conventional gas and oil hides and fur coat formula when the seepage flow behavior of portraying fluid in low-permeability oil deposit, have some limitations.Therefore, formula (1) can not think poorly of the production capacity of permeable reservoir strata effectively.

Summary of the invention

For the problems referred to above, the object of this invention is to provide a kind of low permeability reservoir production capacity method for quantitatively evaluating, the method energy accurate evaluation low permeability reservoir production capacity, evaluation result is accurate, highly versatile, and can be effectively cost-saving.

For achieving the above object, the present invention takes following technical scheme: a kind of low permeability reservoir production capacity method for quantitatively evaluating, it is characterized in that, the method step is as follows: 1) data prediction, comprise data transformation and data cleansing: (1) is based on the construction parameter of low permeability reservoir, log parameter and reservoir parameter, carry out data transformation to these parameters, obtaining can the derivative parameter of more accurate characterization low permeability reservoir production capacity; (2) carry out data cleansing based on the construction parameter of low permeability reservoir, log parameter and derivative parameter, optimize and low permeability reservoir relationship between productivity series of parameters closely, the series of parameters that removing is little with low permeability reservoir relationship between productivity; (3), after carrying out data cleansing and data transformation to low permeability reservoir data set, select the parameter after cleaning as the parameter sets setting up low permeability reservoir evaluating production capacity model; 2) extract character subset: the parameter after low permeability reservoir cleaning is analyzed, adopt cluster analysis, correlation rule and feature selecting algorithm carrying out pretreated data centralization, analyze the correlativity between all parameters and low permeability reservoir production capacity; Based on the analysis result of each algorithm, optimize the parameter sets higher with low permeability reservoir production capacity correlativity, and then extract the character subset intended for setting up low permeability reservoir evaluating production capacity model; 3) set up evaluation model: for extracted character subset, adopt sorted generalization algorithm, in conjunction with parameter each in character subset physical meaning and and low permeability reservoir production capacity between relation, set up multiple low permeability reservoir evaluating production capacity model; 4) assess evaluation model: from multi-solution, completeness, practicality and accuracy rate four aspects, each low permeability reservoir evaluating production capacity model is assessed, select best low permeability reservoir evaluating production capacity model; 5) evaluation model is revised: based on the low permeability reservoir evaluating production capacity model of low permeability reservoir the best, the actual information of low permeability reservoir each side is incorporated best low permeability reservoir evaluating production capacity model, low permeability reservoir evaluating production capacity model is revised, improve the applicability of low permeability reservoir evaluating production capacity model, realize carrying out quantitative evaluation to low permeability reservoir production capacity.

Described step 1) in, construction parameter comprises sand amount, sand ratio, broken pressure and discharge capacity; Log parameter comprises: hole diameter CAL, natural gamma relative value Δ GR, spontaneous potential SP, volume density log value DEN, neutron porosity log value CNL, compressional wave time difference log value AC and deep resistivity log value R _t; Derivative parameter comprises: sandstone gross thickness H, sandstone net thickness H _e, net porosity effective permeability K, oil phase effective permeability K _o, water saturation S _w, producing water ratio F _w, three porosity factor TPI, reservoir quality factor RQI, the volume of voids factor thickness factor H _e/ H, resistance enhancement factor the lithology factor saturation degree factor Δ R _t× Δ AC, two differential divisors and lg|dR _t/ dS _w|.

Described step 2) in, the extracting method of character subset is: to think poorly of permeable reservoir strata production capacity for target, the analysis result of incorporating parametric susceptibility and the physical meaning of each parameter, carry out permutation and combination to parameter, forms the character subset thinking poorly of permeable reservoir strata production capacity.

Described step 3) in, sorted generalization algorithm comprises decision tree, support vector machine, Bayesian network and artificial neural network.

Described step 4) in, multi-solution refers to whether low permeability reservoir evaluating production capacity model only provides unique solution to low permeability reservoir production capacity; Completeness refers to whether low permeability reservoir evaluating production capacity model can cover the different production capacity ranks of low permeability reservoir completely, and all can provide optimum production capacity rank to each class low permeability reservoir; Practicality refers to whether formed low permeability reservoir evaluating production capacity model can predict the production capacity rank of low permeability reservoir effectively; Accuracy rate refers to whether the precision of built low permeability reservoir evaluating production capacity model reaches the precision of Accurate Prediction low permeability reservoir production capacity.

The present invention is owing to taking above technical scheme, it has the following advantages: 1, the present invention fully excavates and the information merging low permeability reservoir each side carrys out the production capacity of quantitative evaluation low permeability reservoir, avoid carrying out other a large amount of core experiments, can be effectively cost-saving, there is stronger economy.2, when Darcy's law and fur coat formula accurately can not calculate low permeability reservoir production capacity, the invention provides the method for a set of energy accurate evaluation low permeability reservoir production capacity, and effectively can ensure the accuracy of low permeability reservoir production capacity quantitative evaluation result.The present invention can extensively Oil & Gas Productivity field of engineering technology application.

Accompanying drawing explanation

Fig. 1 is overall flow schematic diagram of the present invention;

Fig. 2 is the present invention at the logging data processing of A well one class payzone and interpretation results figure;

Fig. 3 is the present invention at the logging data processing of B well two class payzone and interpretation results figure;

Fig. 4 is the present invention at the logging data processing of C well three class payzone and interpretation results figure;

Fig. 5 is the present invention at the logging data processing of D well four class payzone and interpretation results figure;

Fig. 6 is the present invention at the logging data processing of E well five class payzone and interpretation results figure.

Embodiment

Below in conjunction with drawings and Examples, the present invention is described in detail.

As shown in Figure 1, the invention provides a kind of low permeability reservoir production capacity method for quantitatively evaluating, it comprises the following steps:

1) data prediction, comprises data transformation and data cleansing:

(1) based on the construction parameter of low permeability reservoir, log parameter and reservoir parameter, data transformation is carried out to these parameters, obtain following a series of can the derivative parameter of more accurate characterization low permeability reservoir production capacity;

Natural gamma relative value Δ GR in log parameter is:

$\mathrm{\Δ}GR=\frac{GR-{\mathrm{GR}}_{min}}{{\mathrm{GR}}_{max}-{\mathrm{GR}}_{min}},---\left(2\right)$

In formula, Δ GR is natural gamma relative value, dimensionless; GR is natural gamma ray log value, and unit is API; GR _maxfor natural gamma maximal value, unit is API; GR _minfor natural gamma minimum value, unit is API.

Volume density relative value Δ DEN is obtained according to volume density log value in log parameter:

$\mathrm{\Δ}DEN=\frac{DEN-{\mathrm{DEN}}_{min}}{{\mathrm{DEN}}_{max}-{\mathrm{DEN}}_{min}},---\left(3\right)$

In formula, Δ DEN is volume density relative value, dimensionless; DEN is volume density log value, and unit is g/cm ³; DEN _maxfor volume density maximal value, unit is g/cm ³; DEN _minfor volume density minimum value, unit is g/cm ³.

Density porosity

In formula, for density porosity, unit is decimal; DEN _mafor skeleton volume density, unit is g/cm ³; DEN _ffor fluid volume density, unit is g/cm ³.

Neutron porosity relative value Δ CNL:

$\mathrm{\Δ}CNL=\frac{CNL-{\mathrm{CNL}}_{min}}{{\mathrm{CNL}}_{max}-{\mathrm{CNL}}_{min}},---\left(5\right)$

In formula, Δ CNL is neutron porosity relative value, dimensionless; CNL is neutron porosity log value, and unit is percentage %; CNL _maxfor neutron porosity maximal value, unit is percentage %; CNL _minfor neutron porosity minimum value, unit is percentage %.

Neutron porosity

In formula, for neutron porosity, unit is decimal; CNL _mafor skeleton hydrogen index, unit is percentage %; CNL _ffor fluid hydrogen index, unit is percentage %.

Compressional wave time difference relative value Δ AC:

$\mathrm{\Δ}AC=\frac{AC-{\mathrm{AC}}_{min}}{{\mathrm{AC}}_{max}-{\mathrm{AC}}_{min}},---\left(7\right)$

In formula, Δ AC is compressional wave time difference relative value, dimensionless; AC is compressional wave time difference log value, and unit is μ s/m; AC _maxfor compressional wave time difference maximal value, unit is μ s/m; AC _minfor compressional wave time difference minimum value, unit is μ s/m.

Dark relative resistivity Δ R _t:

${\mathrm{\ΔR}}_t=\frac{{\mathrm{lgR}}_t-{\mathrm{lgR}}_{tmin}}{{\mathrm{lgR}}_{tmax}-{\mathrm{lgR}}_{tmin}},---\left(8\right)$

In formula, Δ R _tfor dark relative resistivity, dimensionless; R _tfor deep resistivity log value, unit is Ω m; R _tmaxfor dark resistivity maximal value, unit is Ω m; R _tminfor dark resistivity minimum value, unit is Ω m.

Three porosity factor TPI is:

$TPI=1-\sqrt{\frac{{(\mathrm{\Δ}CNL-\mathrm{\Δ}AC)}^2+{(\mathrm{\Δ}CNL-(1-\mathrm{\Δ}DEN\left)\right)}^2+{(\mathrm{\Δ}AC-(1-\mathrm{\Δ}DEN\left)\right)}^2}{3}},---\left(9\right)$

In formula, TPI is the three porosity factor, dimensionless.

Net porosity for:

In formula, for net porosity, unit is decimal.

Effective permeability K is:

In formula, K is effective permeability, and unit is mD.

Oil phase effective permeability K _o:

K _o＝f(K)，(12)

In formula, K _ofor oil phase effective permeability, unit is mD.

Reservoir quality factor RQI:

In formula, RQI is reservoir quality factor, and unit is μm.

Based on A Erqi (Archie) formula, water saturation S _wfor:

In formula, S _wfor water saturation, unit is decimal; A, b are the scale-up factor relevant with lithology, dimensionless; M is cementation exponent, dimensionless; N is saturation index, dimensionless; R _wfor stratum resistivity of water, unit is Ω m; R _tfor dark resistivity, unit is Ω m; Wherein, a, b, m and n are referred to as Archie's parameters.

Oil saturation S _ofor:

S _o＝1-S _w，(15)

In formula, S _ofor oil saturation, unit is decimal.

Producing water ratio F _wfor:

F _w＝f(S _w)，(16)

In formula, F _wfor producing water ratio, unit is decimal.

The computing formula of two differential divisors is as follows respectively:

In addition, based on the concrete degree of depth of every mouthful of well, sandstone gross thickness H and sandstone net thickness H can also be obtained _e, thickness factor H can be obtained through further data transformation _e/ H (dimensionless), the volume of voids factor (dimensionless), resistance enhancement factor (dimensionless) and saturation degree factor Δ R _t× Δ AC (dimensionless).

(2) data cleansing is carried out based on the construction parameter of low permeability reservoir, log parameter and derivative parameter, optimize and low permeability reservoir relationship between productivity series of parameters closely, the series of parameters that removing is little with low permeability reservoir relationship between productivity, as shown in table 1.

The parameter that the data cleansing of table 1 low permeability reservoir is forward and backward

In table 1, CAL is hole diameter, and unit is cm; SP is spontaneous potential, and unit is mV; for the lithology factor, dimensionless.

(3), after carrying out data cleansing and data transformation to low permeability reservoir data set, select the parameter after cleaning as the parameter sets setting up low permeability reservoir evaluating production capacity model: sand amount, Δ GR, DEN, CNL, AC, R _t, H _e, k, K _o, S _w, F _w, TPI, RQI, h _e/ H, Δ R _t× Δ AC, and lg|dR _t/ dS _w| totally 21 parameters.

2) character subset is extracted: 21 parameters after low permeability reservoir cleaning are analyzed, adopt cluster analysis, correlation rule and feature selecting scheduling algorithm carrying out pretreated data centralization, analyze the correlativity between all parameters and low permeability reservoir production capacity.Based on the analysis result of each algorithm, thus optimize the parameter sets higher with low permeability reservoir production capacity correlativity, remove the nuisance parameter that all the other and low permeability reservoir relationship between productivity are little, and then extract the character subset intended for setting up low permeability reservoir evaluating production capacity model.

Wherein, the extracting method of character subset is: to think poorly of permeable reservoir strata production capacity for target, and the analysis result of incorporating parametric susceptibility and the physical meaning of each parameter, carry out permutation and combination to parameter, forms the character subset thinking poorly of permeable reservoir strata production capacity.

Cluster analysis, correlation rule and feature selecting algorithm mainly comprise following 8 kinds of algorithm: Relief-F (feature weight), InformationGain (IG) (information gain), InformationGainRatio (IGR) (information gain-ratio), GiniIndex (Gini) (Gini coefficient), SupportVectorMachine-RecursiveFeatureElimination (SVM-RFE) (support vector machine recursive feature about subtracts), PrincipalComponentAnalysis (PCA) (principal component analysis (PCA)), Deviation (deviation) and Correlation (being correlated with).

3) set up evaluation model: for extracted character subset, adopt sorted generalization algorithm, in conjunction with parameter each in character subset physical meaning and and low permeability reservoir production capacity between relation, and then set up multiple low permeability reservoir evaluating production capacity model.Wherein, sorted generalization algorithm comprises decision tree, support vector machine, Bayesian network and artificial neural network.

4) assess evaluation model: from multi-solution, completeness, practicality and accuracy rate four aspects, each low permeability reservoir evaluating production capacity model is assessed, select best low permeability reservoir evaluating production capacity model; Wherein, multi-solution refers to whether low permeability reservoir evaluating production capacity model only provides unique solution to low permeability reservoir production capacity; Completeness refers to whether low permeability reservoir evaluating production capacity model can cover the different production capacity ranks of low permeability reservoir completely, and all can provide optimum production capacity rank to each class low permeability reservoir; Practicality refers to whether formed low permeability reservoir evaluating production capacity model can predict the production capacity rank of low permeability reservoir effectively; Accuracy rate refers to whether the precision of built low permeability reservoir evaluating production capacity model reaches the precision of Accurate Prediction low permeability reservoir production capacity.

5) evaluation model is revised: based on the low permeability reservoir evaluating production capacity model of low permeability reservoir the best, in conjunction with low permeability reservoir actual reservoir feature and the concrete physical meaning of each parameter, actual information by low permeability reservoir each side incorporates best low permeability reservoir evaluating production capacity model, low permeability reservoir evaluating production capacity model is revised, improve the applicability of low permeability reservoir evaluating production capacity model, to realize carrying out quantitative evaluation to low permeability reservoir production capacity.

Below in conjunction with a specific embodiment, the invention will be described further.

Embodiment: quantitative evaluation is carried out to the production capacity of certain low permeability oil field.

First, the stratum adopting different logging instrumentations to measure low permeability oil field can obtain log parameter GR, DEN, CNL, AC, R _t, to log parameter GR, DEN, CNL, AC, R _tafter carrying out data transformation, Δ GR can be obtained respectively, DEN, CNL, AC, R _t, computing formula is as follows respectively:

$\mathrm{\Δ}GR=\frac{GR-30}{150-30}---\left(19\right)$

$\mathrm{\Δ}DEN=\frac{DEN-2.20}{2.70-2.20}---\left(20\right)$

$\mathrm{\Δ}CNL=\frac{CNL-(-10\%)}{40\%-(-10\%)}---\left(21\right)$

$\mathrm{\Δ}AC=\frac{AC-100}{350-100}---\left(22\right)$

${\mathrm{\ΔR}}_t=\frac{{\mathrm{lgR}}_t-\lg 1}{\lg 100-\lg 1}---\left(23\right)$

Secondly, derivative parameter can also be obtained k,K _o, S _w, F _w, TPI, RQI, computing formula is as follows respectively:

K _o＝0.0729K-0.0074,R ²＝0.9423(26)

$F_w=\frac{0.01}{0.0286\×\exp \left(-0.5319\×\left(100S_w-50\right)\right)+0.0102},R^2=0.9606---\left(28\right)$

$TPI=1-\sqrt{\frac{{(\mathrm{\Δ}CNL-\mathrm{\Δ}AC)}^2+{(\mathrm{\Δ}CNL-(1-\mathrm{\Δ}DEN\left)\right)}^2+{(\mathrm{\Δ}AC-(1-\mathrm{\Δ}DEN\left)\right)}^2}{3}}---\left(29\right)$

In addition, by the statistical study to this low permeability oil field every mouthful well, sandstone gross thickness H and net thickness H can be obtained _e, thus can H be obtained _e/ H; In addition, Δ R _t× Δ AC, lg|dR _t/ dS _w| computing formula respectively as follows:

${\mathrm{\ΔR}}_t\×\mathrm{\Δ}AC=\frac{{\mathrm{lgR}}_t-\lg 1}{\lg 100-\lg 1}\×\frac{AC-100}{350-100}---\left(34\right)$

Adopt Relief-F, InformationGain (IG), InformationGainRatio (IGR), GiniIndex (Gini), SupportVectorMachine-RecursiveFeatureElimination (SVM-RFE), PrincipalComponentAnalysis (PCA), Deviation and Correlation be totally 8 kinds of algorithms, analyze the susceptibility between low permeability reservoir parameter and production capacity, as shown in table 2.

Sensitivity analysis between table 2 low permeability reservoir parameter and production capacity

Based on the analysis result of table 2, known K _obe redundant attributes with RQI, namely to the most insensitive parameter of low permeability reservoir production capacity; Other parameters to the magnitude relationship of production capacity sensitivity are

According to the analysis result of susceptibility between low permeability reservoir parameter and production capacity, by arranging different weight threshold, make decision tree, support vector machine, Bayesian network and artificial neural network four kinds of sorted generalization algorithms in each feature selecting algorithm, obtain the most high-accuracy of production capacity classification prediction, as shown in table 3.

Table 3 four kinds of sorted generalization algorithms obtain most high-accuracy on different characteristic selection algorithm

As shown in Table 3, for the data set of this low permeability oil field, the prediction effect of decision tree is better than support vector machine, artificial neural network and Bayesian network, and it obtains most high-accuracy and is substantially greater than 80%.

Secondly, utilize decision Tree algorithms in different parameters combination, set up the evaluating production capacity model of Pyatyi respectively, as shown in table 4.Pyatyi production capacity all payzones is divided into a class payzone (daily oil production >=20t), two class payzones (15t≤daily oil production < 20t), three class payzones (10t≤daily oil production < 15t), four class payzones (4t≤daily oil production < 10t) and five class payzones (daily oil production < 4t).

Table 4 decision tree in different parameters combination obtain the accuracy rate of evaluating production capacity model

The 2nd class parameter combinations (H in table 4 _e, Δ GR, DEN, CNL, AC, R _t, k,S _w, F _w) evaluate different production capacity rank result as shown in table 5.

Table 5 the 2nd class parameter combinations predicts predicting the outcome of different production capacity rank

As shown in Table 5, except two class payzones, the evaluation precision of other payzones is all greater than 80%.Be about 87.96% in the precision of this low permeability oil field production capacity quantitative evaluation, meet the demand that actual development is produced.

The present invention can come this low permeability oil field of quantitative evaluation A well, B well, C well, D well, E well capacity by instruments such as Geolog, GeoFrame, Forward, Lead, adopts Geolog instrument to draw this low-permeability oil Tanaka each well productivity quantitative evaluation result figure in the present embodiment.

As shown in Figure 2, be A well one class payzone (day produce oil 60.1t, daily output water 0m ³) logging data processing and interpretation results figure, in the 1st road, natural gamma and natural potential logging curve can indicate the rock signature on stratum, and section gauge logging curve can portray the situation of well, whether there is expanding or undergauge; In the 2nd road, be interval transit time, neutron porosity and volume density logging trace, the physical property characteristic on stratum can be reflected; In the 3rd road, be dark, in, shallow resistivity logging trace, the electrical property feature on stratum can be portrayed; In the 4th road, it is the net porosity calculated based on above-mentioned logging trace; In the 5th road, it is the effective permeability calculated based on above-mentioned logging trace; In the 6th road, it is the water saturation calculated based on above-mentioned logging trace.

As shown in Figure 3, be B well two class payzone (day produce oil 17.6t, daily output water 0m ³) logging data processing and interpretation results figure, in the 1st road, natural gamma and natural potential logging curve can indicate the rock signature on stratum, and section gauge logging curve can portray the situation of well, whether there is expanding or undergauge; In the 2nd road, be interval transit time, neutron porosity and volume density logging trace, the physical property characteristic on stratum can be reflected; In the 3rd road, be dark, in, shallow resistivity logging trace, the electrical property feature on stratum can be portrayed; In the 4th road, it is the net porosity calculated based on above-mentioned logging trace; In the 5th road, it is the effective permeability calculated based on above-mentioned logging trace; In the 6th road, it is the water saturation calculated based on above-mentioned logging trace; Due to net thickness H _erelatively large, this well is mistaken for a class payzone.

As shown in Figure 4, be C well three class payzone (day produce oil 10.5t, daily output water 1.6m ³) logging data processing and interpretation results figure, in the 1st road, natural gamma and natural potential logging curve can indicate the rock signature on stratum, and section gauge logging curve can portray the situation of well, whether there is expanding or undergauge; In the 2nd road, be interval transit time, neutron porosity and volume density logging trace, the physical property characteristic on stratum can be reflected; In the 3rd road, be dark, in, shallow resistivity logging trace, the electrical property feature on stratum can be portrayed; In the 4th road, it is the net porosity calculated based on above-mentioned logging trace; In the 5th road, it is the effective permeability calculated based on above-mentioned logging trace; In the 6th road, it is the water saturation calculated based on above-mentioned logging trace; Due to dark resistivity R _trelatively little, this well is mistaken for five class payzones.

As shown in Figure 5, be D well four class payzone (day produce oil 4.6t, daily output water 0m ³) logging data processing and interpretation results figure, in the 1st road, natural gamma and natural potential logging curve can indicate the rock signature on stratum, and section gauge logging curve can portray the situation of well, whether there is expanding or undergauge; In the 2nd road, be interval transit time, neutron porosity and volume density logging trace, the physical property characteristic on stratum can be reflected; In the 3rd road, be dark, in, shallow resistivity logging trace, the electrical property feature on stratum can be portrayed; In the 4th road, it is the net porosity calculated based on above-mentioned logging trace; In the 5th road, it is the effective permeability calculated based on above-mentioned logging trace; In the 6th road, it is the water saturation calculated based on above-mentioned logging trace.

As shown in Figure 6, be E well five class payzone (day produce oil 0t, daily output water 2.7m ³) logging data processing and interpretation results figure, in the 1st road, natural gamma and natural potential logging curve can indicate the rock signature on stratum, and section gauge logging curve can portray the situation of well, whether there is expanding or undergauge; In the 2nd road, be interval transit time, neutron porosity and volume density logging trace, the physical property characteristic on stratum can be reflected; In the 3rd road, be dark, in, shallow resistivity logging trace, the electrical property feature on stratum can be portrayed; In the 4th road, it is the net porosity calculated based on above-mentioned logging trace; In the 5th road, it is the effective permeability calculated based on above-mentioned logging trace; In the 6th road, it is the water saturation calculated based on above-mentioned logging trace.

The various embodiments described above are only for illustration of the present invention; the structure of each parts, size, setting position and shape all can change to some extent; on the basis of technical solution of the present invention; all improvement of carrying out individual part according to the principle of the invention and equivalents, all should not get rid of outside protection scope of the present invention.

Claims (7)

1. a low permeability reservoir production capacity method for quantitatively evaluating, is characterized in that, the method step is as follows:

1) data prediction, comprises data transformation and data cleansing:

(1) based on the construction parameter of low permeability reservoir, log parameter and reservoir parameter, carry out data transformation to these parameters, obtaining can the derivative parameter of more accurate characterization low permeability reservoir production capacity;

(2) carry out data cleansing based on the construction parameter of low permeability reservoir, log parameter and derivative parameter, optimize and low permeability reservoir relationship between productivity series of parameters closely, the series of parameters that removing is little with low permeability reservoir relationship between productivity;

(3), after carrying out data cleansing and data transformation to low permeability reservoir data set, select the parameter after cleaning as the parameter sets setting up low permeability reservoir evaluating production capacity model;

2) extract character subset: the parameter after low permeability reservoir cleaning is analyzed, adopt cluster analysis, correlation rule and feature selecting algorithm carrying out pretreated data centralization, analyze the correlativity between all parameters and low permeability reservoir production capacity; Based on the analysis result of each algorithm, optimize the parameter sets higher with low permeability reservoir production capacity correlativity, and then extract the character subset intended for setting up low permeability reservoir evaluating production capacity model;

3) set up evaluation model: for extracted character subset, adopt sorted generalization algorithm, in conjunction with parameter each in character subset physical meaning and and low permeability reservoir production capacity between relation, set up multiple low permeability reservoir evaluating production capacity model;

4) assess evaluation model: from multi-solution, completeness, practicality and accuracy rate four aspects, each low permeability reservoir evaluating production capacity model is assessed, select best low permeability reservoir evaluating production capacity model;

5) evaluation model is revised: based on the low permeability reservoir evaluating production capacity model of low permeability reservoir the best, the actual information of low permeability reservoir each side is incorporated best low permeability reservoir evaluating production capacity model, low permeability reservoir evaluating production capacity model is revised, improve the applicability of low permeability reservoir evaluating production capacity model, realize carrying out quantitative evaluation to low permeability reservoir production capacity.

2. a kind of low permeability reservoir production capacity method for quantitatively evaluating as claimed in claim 1, is characterized in that: described step 1) in, construction parameter comprises sand amount, sand ratio, broken pressure and discharge capacity; Log parameter comprises: hole diameter CAL, natural gamma relative value Δ GR, spontaneous potential SP, volume density log value DEN, neutron porosity log value CNL, compressional wave time difference log value AC and deep resistivity log value R _t; Derivative parameter comprises: sandstone gross thickness H, sandstone net thickness H _e, net porosity effective permeability K, oil phase effective permeability K _o, water saturation S _w, producing water ratio F _w, three porosity factor TPI, reservoir quality factor RQI, the volume of voids factor thickness factor H _e/ H, resistance enhancement factor the lithology factor saturation degree factor Δ R _t× Δ AC, two differential divisors and lg|dR _t/ dS _w|.

3. a kind of low permeability reservoir production capacity method for quantitatively evaluating as claimed in claim 1, it is characterized in that: described step 2) in, the extracting method of character subset is: to think poorly of permeable reservoir strata production capacity for target, the analysis result of incorporating parametric susceptibility and the physical meaning of each parameter, permutation and combination is carried out to parameter, forms the character subset thinking poorly of permeable reservoir strata production capacity.

4. a kind of low permeability reservoir production capacity method for quantitatively evaluating as claimed in claim 2, it is characterized in that: described step 2) in, the extracting method of character subset is: to think poorly of permeable reservoir strata production capacity for target, the analysis result of incorporating parametric susceptibility and the physical meaning of each parameter, permutation and combination is carried out to parameter, forms the character subset thinking poorly of permeable reservoir strata production capacity.

5. a kind of low permeability reservoir production capacity method for quantitatively evaluating as described in any one of Claims 1 to 4, is characterized in that: described step 3) in, sorted generalization algorithm comprises decision tree, support vector machine, Bayesian network and artificial neural network.

6. a kind of low permeability reservoir production capacity method for quantitatively evaluating as described in any one of Claims 1 to 4, is characterized in that: described step 4) in, multi-solution refers to whether low permeability reservoir evaluating production capacity model only provides unique solution to low permeability reservoir production capacity; Completeness refers to whether low permeability reservoir evaluating production capacity model can cover the different production capacity ranks of low permeability reservoir completely, and all can provide optimum production capacity rank to each class low permeability reservoir; Practicality refers to whether formed low permeability reservoir evaluating production capacity model can predict the production capacity rank of low permeability reservoir effectively; Accuracy rate refers to whether the precision of built low permeability reservoir evaluating production capacity model reaches the precision of Accurate Prediction low permeability reservoir production capacity.

7. a kind of low permeability reservoir production capacity method for quantitatively evaluating as claimed in claim 5, is characterized in that: described step 4) in, multi-solution refers to whether low permeability reservoir evaluating production capacity model only provides unique solution to low permeability reservoir production capacity; Completeness refers to whether low permeability reservoir evaluating production capacity model can cover the different production capacity ranks of low permeability reservoir completely, and all can provide optimum production capacity rank to each class low permeability reservoir; Practicality refers to whether formed low permeability reservoir evaluating production capacity model can predict the production capacity rank of low permeability reservoir effectively; Accuracy rate refers to whether the precision of built low permeability reservoir evaluating production capacity model reaches the precision of Accurate Prediction low permeability reservoir production capacity.