CN100437147C - Multi-parameter dimension-reducing oil-gas-water-layer identifying method - Google Patents

Abstract

The present invention relates to a multi-parameter dimension-reducing method for identifying oil, gas and water layers, which is mainly used for solving the problem of low coincidence rate of oil and water layers in the existing method for explaining oil and water layers. The present invention is characterized in that the method comprises the following steps: (1) the geological conditions of an oil reservoir are explored, the data of the oil test, production test or oil production of the oil reservoir is collected; (2), a target reservoir layer in a well is determined according to the comprehensive explanation of electrical well logging and the oil and gas display results of well logging; (3) the electric well logging and well logging reservoir layer parameters of the target reservoir layer are selected; 4, a computer program is compiled by a statistical mathematics-factor analysis method, the liquid producing data of the target reservoir layer is combined to linearly combine a plurality of electrical well logging and well logging parameters of the target reservoir layer in order to reduce dimension to two main factors; (5) an established figure board for oil, gas and water layers in the step (4) is used for distinguishing other oil, gas and water layers in the target reservoir layer of the oil reservoir. The method can be used for accurately explaining oil and gas layers, the coincidence rate of oil and water layers is high, the successful rate of obtaining industrial oil and gas flow is increased, and thus, the economic benefit of oil and gas exploration and development is increased.

Description

Multi-parameter dimension-reducing oil-gas-water-layer identifying method

Technical field:

The present invention relates to method, the especially multi-parameter dimension-reducing oil-gas-water-layer identifying method of oil-gas exploration and exploitation well oil gas and water layer interpretation.

Background technology:

For a long time, petroleum industry in the exploration in oil field and exploitation explanation hydrocarbon zone such as widespread use electric logging and log data this in petroleum industry, brought into play crucial effects.In long-term research and production practice, from exploring to the characteristics of developing different phase, formed many methods according to the oil field, disengage many hydrocarbon zone by hydrocarbon zone tool high resistivity, the characteristic solution of the praetersonic time difference, explored and developed the research and production practice and confirmed.But along with deepening continuously of oil-gas exploration, the target geologic condition of exploratory development becomes increasingly complex, and the method and the technology of original explanation oil-gas-water layer have inadaptability.Fuyu County, the Yang Dacheng seed oil layer in area to the east of the Daqing oil field placanticline in the past, general in conjunction with log datas such as dark side direction resistivity, spontaneous potential and induction just can discrimination of reservoir produce oil air water character.But along with exploring Fuyu County, the Yang Dacheng seed oil layer that direction goes to placanticline western part, with original method identification oil-gas-water layer, its coincidence rate obviously reduces.Especially when improving the oil prognostic reserves, because the oil reservoir reservoir properties is poor, mostly be the low porosity and low permeability reservoir, there are shale, calcareous content higher, the pore texture more complicated, making in the curve such as resistivity the information of reflection fluid is reflected, pore texture and lithology information partly covers, and is difficult to simply fluid properties in the reservoir correctly be identified from logging trace with original method in interpretation process.On the other hand, because low permeability reservoir water yield formation position no clear regularity on the zone, and formation testing is many to close the examination layer, this is for judging accurately that by the formation testing result individual layer fluid properties has certain difficulty, simultaneously also can't accomplish to change with the variation of lithology for the foundation of water saturation model, to making result of calculation have bigger error, the oil-water-layer coincidence rate of the chart interpretation of making of classic method is lower, especially can't formulate the electrical standard plate of oil reservoir under the individual layer formation testing condition lacking.Many-sided factor makes original method be difficult to satisfy oil-gas exploration and development and improves the reserves requirements of one's work.

Summary of the invention:

In order to overcome the lower deficiency of method oil-water-layer coincidence rate of existing explanation oil-gas-water layer, the invention provides a kind of multi-parameter dimension-reducing oil-gas-water-layer identifying method, this method makes hydrocarbon zone interpretation more accurate, the oil-water-layer coincidence rate is higher, improve the success ratio that obtains commercial hydrocarbon flow, thereby improved the economic benefit of oil-gas exploration and exploitation.Low hole osmotic pressure in oil-gas exploration and the exploitation splits transformation, about more than 50 ten thousand yuan of one deck reservoir formation testing expense.Therefore, use this method to improve the construction success ratio, have important economic implications.

Technical scheme of the present invention is: this multi-parameter dimension-reducing oil-gas-water-layer identifying method comprises the steps:

1. investigate the geologic condition of oil reservoir, collect formation testing, pilot production or the produce oil data of oil reservoir;

2. according to electric logging integrated interpretation and well logging show of oil and gas result, determine target reservoir in the well;

3. choose the electric logging of target reservoir and the reservoir parameter of well logging;

4. use the method for statistical mathematics-factorial analysis, technology uses a computer, be compiled into computer program, in conjunction with the 1. production fluid of resulting target reservoir (oil, gas, water) data,, carry out the linear combination dimensionality reduction to the binomial main factor to 3. multinomial electric logging of resulting target reservoir and logging parameters, it is main gene, be that the variance contribution of P=2 and binomial main gene accumulative total surpasses at 80% o'clock, foundation can be illustrated in the oil-gas-water layer classification plate of two dimensional surface, also claims oil-gas-water layer identification plate;

5. with oil, gas, the water layer plate of 4. middle foundation, other target reservoir of this oil reservoir is carried out oil-gas-water layer differentiate.

Above-mentioned multi-parameter dimension-reducing oil-gas-water-layer identifying method, for the oil reservoir that lacks the individual layer formation testing, before setting up oil-gas-water layer classification plate, also will be in conjunction with corresponding analysis method.

Be meant in uncased hole in the described well;

The electric logging parameter of described target reservoir is the data more than three in microballoon focused resistivity, dark side direction resistivity, spontaneous potential, interval transit time, compensated neutron, compensation density, the GR.

The logging parameters of described target reservoir is meant the geochemical logging data.

The logging parameters of described target reservoir is the gas detection logging data, total hydrocarbon, hydrocarbon component, nonhydrocarbon component.

The present invention has following beneficial effect: owing to take such scheme, considered the influence of target reservoir fluid and rock skeleton to multinomial well logging, logging parameters, used the mathematical algorithm of multiparameter mathematical statistics, set up the plane plate by dimensionality reduction, the method for only setting up two-dimentional plate with two parameters than the past has bigger practicality.Compare with method of discrimination such as past multiple parameter method, neural network method and filter methods, the advantage of self arranged, it be with actual formation testing result for just drilling oil, gas, the water layer method of discrimination that the basis is set up, be applicable to the area that degree of prospecting is higher.The actual measurement sample is many more, and plate precision and differentiation coincidence rate are high more.Especially close the more area of examination data for multilayer,, filter out the way of typical oil, gas, water layer, set up oil-water-layer and differentiate plate by rejecting abnormalities point repeatedly.In addition,,, parameter is carried out various combined transformations, form new parameter, can adapt to the reservoir characteristic of target area, improve the coincidence rate of differentiating in conjunction with artificial experience for the complicated reservoirs condition.

Description of drawings:

Accompanying drawing 1 is that XZ area grape flower oil-water-layer is differentiated plate;

Accompanying drawing 2 is that Fuyu County, bycg area, Yang Dacheng seed oil layer oil-water-layer are differentiated plate.

Embodiment:

The invention will be further described below in conjunction with specific embodiment:

The present invention comes from the inventor oil gas and water layer interpretation, the submission reserves to prospect pit in the oil field prospecting is made the new knowledge that oil-gas-water layer is differentiated plate and on-the-spot actual formation testing effect analysis.

On the basis of above-mentioned principle, we adopt the delphi language to work out multi-parameter dimension-reducing oil-gas-water-layer Classification and Identification computer program, use several data forms such as menu Chinese character, * .dbf, * .xls, * .vts in the program, flexible and convenient operation.

This multi-parameter dimension-reducing oil-gas-water-layer identifying method comprises the steps:

1. investigate the geologic condition of oil reservoir, collect formation testing, pilot production or the produce oil data of oil reservoir;

2. according to electric logging integrated interpretation and well logging show of oil and gas result, determine target reservoir in the well;

3. choose the electric logging of target reservoir and the reservoir parameter of well logging.

More than three the step be prior art, in " petroleum and natural gas calculation of reserves methods " (1990) the 18th to 49 pages of writing such as Yang Tongyou detailed argumentation is arranged, be not described at this.

4. use the method for statistical mathematics-factorial analysis, technology uses a computer, be compiled into computer program, in conjunction with the 1. production fluid of resulting target reservoir (oil, gas, water) data,, carry out the linear combination dimensionality reduction to the binomial main factor to 3. multinomial electric logging of resulting target reservoir and logging parameters, it is main gene, be that the variance contribution of P=2 and binomial main gene accumulative total surpasses at 80% o'clock, foundation can be illustrated in the oil-gas-water layer classification plate of two dimensional surface, also claims oil-gas-water layer identification plate.

The resulting data of different curves of well logging have reflected the multiple information of reservoir, and acoustic logging and density logging can reflect the porosity of reservoir preferably; The gamma well logging can reflect the shale index in the reservoir preferably; Resistivity logging can reflect the oiliness of reservoir preferably; Inductolog and natural potential logging can reflect the property of water-bearing of reservoir preferably.Geochemical logging S ₁And S ₂Value has reflected reservoir hydrocarbon-containifirst fluid situation.Electric logging that these are various and log data have reflected the feature of reservoir from many aspects, for the moisture character of discerning reservoir provides multiple parameters.We can regard oil gas water reservoir as by a plurality of parameter characterizations space geometry figure, in order to reach the purpose of differentiation, must carry out linear combination to multiple parameters, dimension-reduction treatment, form new combination parameter-factor, from the engineering mathematics angle, the space geometry graphic feature that multiparameter characterizes is by amount and shape two aspect factor affecting, i.e. amount-reflection is big or small; Shape-reflection space geometry form.Complicated nondescript its space geometry figure of oil gas water reservoir should be different, through the tiltedly very big quadrature rotation of variance, can promptly be shown to two dimensional surface to the space geometry graphic feature that characterizes by multiparameter to the multiparameter dimensionality reduction, find their figure of difference in the plane.Thereby reach purpose with multiparameter identification oil, gas, water layer.Utilize the multiparameter of target reservoir,, set up oil-gas-water layer plane plate, thereby realize the oil-gas-water layer character of target reservoir in the well is differentiated by dimension-reduction treatment.Starting point of the present invention that Here it is.

In the present invention, the parameter amount (n) that can obtain according to electric logging and well logging, a plurality of samples (m) of target reservoir therefrom extract 2 principal elements, represent with the linear combination of raw parameter, and the purpose of the information that reflects mass data basically and provided is provided.

Be provided with n sample, each sample comprises m variable, and the value of a n sample m variable is write as the data matrix form:

$Z=\left[\begin{array}{cccc}{z}_{11}& z_{12}& \·\·\·\·\·\·& z_{1n}\\ z_{21}& z_{22}& \·\·\·\·\·\·& 2_{2n}\\ z_{m1}& z_{m2}& \·\·\·\·\·\·& z_{\mathrm{mn}}\end{array}\right]$

In matrix Z, z _j=(z _1jz _2j... z _Mj) ' (j=1,2 ..., n) observed reading of a j sample m variable of expression, i.e. j the sample of Z; z _i=(z _I1z _I2... z _In) (i=1,2 ..., n) observed reading of i variable of expression in n sample, i.e. i the variable of Z.

In matrix Z, z _j=(z _1jz _2j... z _Mj) ' (j=1,2 ..., n) observed reading of a j sample m variable of expression, i.e. j the sample of Z; Zi=(z _I1z _I2... z _In) (i=1,2 ..., n) observed reading of i variable of expression in n sample, i.e. i the variable of Z.

If variable z _iStandard deviation standardization (abbreviation standardization) variable be x _i, and a corresponding observed reading is designated as matrix

$X=\left[\begin{array}{cccc}{x}_{11}& x_{12}& \·\·\·\·\·\·& x_{1n}\\ x_{21}& x_{22}& \·\·\·\·\·\·& x_{2n}\\ x_{m1}& x_{m2}& \·\·\·\·\·\·& x_{\mathrm{mn}}\end{array}\right]$

During follow-up reader in factorial analysis discussed, if no specified otherwise, all arranging variable was standard.

For any two variable x _i=(x _I1x _I2... x _In) and x _j=(z _J1z _J2... z _Im) their related coefficient is $r_{\mathrm{ij}}=\frac{1}{n-1}\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{x}_{\mathrm{ik}}\·x_{\mathrm{jk}}(i,j=\mathrm{1,2},\·\·\·,m)$

The related coefficient of m variable constitutes the matrix of a m * m

$R=\left[\begin{array}{cccc}{r}_{11}& r_{12}& \·\·\·\·\·\·& r_{1n}\\ r_{21}& r_{22}& \·\·\·\·\·\·& r_{2n}\\ r_{m1}& r_{m2}& \·\·\·\·\·\·& r_{\mathrm{mn}}\end{array}\right]$

And r _Ij=r _Ji, r ₁₁=r ₂₂=...=r _Mn=1, and claim that R is the correlation matrix of variable.

Any two sample x _i(=x _1ix _2i... x _Mi) ' and x _j(=x _1jx _2j... x _Mj) ' similarity coefficient be

$q_{\mathrm{ij}}=\frac{\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{x}_{\mathrm{ki}}\·x_{\mathrm{kj}}}{\sqrt{\underset{k=1}{\overset{m}{\mathrm{\Σ}}}{x}_{\mathrm{ki}}^2\underset{k=1}{\overset{m}{\mathrm{\Σ}}}{x}_{\mathrm{kj}}^2}}(i,j=\mathrm{1,2},\·\·\·,n)$

The similarity coefficient of n sample is write as matrix form

$Q=\left[\begin{array}{cccc}{q}_{11}& q_{12}& \·\·\·\·\·\·& q_{1n}\\ q_{21}& q_{22}& \·\·\·\·\·\·& q_{2n}\\ q_{m1}& q_{m2}& \·\·\·\·\·\·\·& q_{\mathrm{mn}}\end{array}\right]$

And q ₁₁=q ₂₂=...=q _Mn=1, q _Ij=q _Ji, claim that Q is the similar matrix of sample.

R mode factor analysis is the inner structure of research correlation matrix R, therefrom finds out p generalized variable f that all variablees is started to control making usefulness _k=(k=1,2 ..., p, p＜m), and variable x _iBe expressed as f _kLinear combination, promptly

x _i＝a _i1f ₁+a _i2f ₂+...+a _ipf _p+a _ie _i (i＝1，2，…，m)…… (1)

When p＜＜during m,, and further explore the controlling factor of Changing Pattern in the origin cause of formation contact of variable and the space by formula (1) Study on Simplifying system.

The factorial analysis of Q type is by the research to similar matrix Q inner structure, seeks p generalized variable f of restriction sample similarity _k=(k=1,2 ..., p, p＜n), and sample x _iBe expressed as f _kLinear combination, promptly

x _j＝a _j1f ₁+a _j2f ₂+...+a _jpf _p+a _je _j (j＝1，2，…，n)…… (2)

When p＜＜during n, by formula (2) abbreviation sample research system, and further study sample produces the main cause of similarity.

It is to be noted at this: formula (1) and formula (2) also are the basic assumption conditions of R type and the factorial analysis of Q type.E wherein _iAnd e _jBe respectively to obey that average is 0, variance is σ _i ²And σ _j ²Normal distribution.

Above-mentioned by R type and the analysis of Q type factor-analysis approach, integration objective reservoir data, disclose between the variable and between the sample in the contact on the origin cause of formation or on the space, adopt this algorithm to play and do not lose or lose less oil geology origin cause of formation information, reach the effect of Dimension Reduction Analysis problem.Promptly when p=2, set up the discriminant classification plate of oil-gas-water layer.

Above-mentioned multi-parameter dimension-reducing oil-gas-water-layer identifying method, for the oil reservoir that lacks the individual layer formation testing, before setting up oil-gas-water layer classification plate, also will be in conjunction with corresponding analysis method.Concrete combination sees " correspondence analysis oil-gas-water layer method of discrimination and application " literary composition that the author is published in 2002 the 6th phases " grand celebration oil geology and exploitation ".Be meant in uncased hole in the described well; The electric logging parameter of described target reservoir is the data more than three in microballoon focused resistivity, dark side direction resistivity, spontaneous potential, interval transit time, compensated neutron, compensation density, the GR; The logging parameters of described target reservoir is meant the geochemical logging data; The logging parameters of described target reservoir is the gas detection logging data, total hydrocarbon, hydrocarbon component, nonhydrocarbon component.

5. with oil, gas, the water layer plate set up in 4., to other target reservoir of this oil reservoir (not formation testing) carry out oil-gas-water layer and differentiate.To target reservoir (not formation testing) carry out oil-gas-water layer when differentiating, employed parameter order and number with 4. in identical when discerning plate of the oil gas water of foundation.

Specifically carry out in two steps:

The first step: set up oil-gas-water layer identification plate

At first, in concrete oil reservoir or zone, collect the formation testing pilot production or the produce oil data of individual well, and then read the log data and the logging data of target reservoir in these wells, as seven parameters such as dark side direction resistivity, compensated neutron, compensation density, interval transit time, spontaneous potential, GR, hole diameters.The computer program that utilization weaves utilizes R type and the dual factor-analysis approach of Q type to train, classify.The two-dimentional plate table of classification results after with dimensionality reduction coated different colours to oil district, pool, oil-water common-layer district respectively then, and general oil district is for red, and the pool is blue look, and the oil-water common-layer district is a sky-blue, as shown in Figure 1.So far, differentiation plate in multiparameter dimensionality reduction oil gas pool is built up.

Second step: differentiate unknown layers runoff yield volume property

After above-mentioned plate builds up, just can differentiate unknown fluid properties of reservoir, both can be that batch data is handled in the differentiation process, can be again that single sample is directly imported, after differentiating the result, waiting to report the sign of Lock-in oil, gas, water layer in the fixedly row of sample data, whether these differentiate the result accurate, can verify by formation testing technology.This method is differentiated oil-gas-water layer, realizes software implementation, and it is objective to differentiate the result.Simultaneously, it is that form with data exists in the computing machine, in using, actual research and production verify the result by formation testing, the oil-gas-water layer plate that continuous correction has been built, because this method has been considered the influence of multiple parameter to reservoir produce oil air water, improved the precision of the oil-gas-water layer differentiation of complex reservoir reservoir.

Embodiment one:

1. investigate and study the geologic condition of oil reservoir, collect formation testing, pilot production or the produce oil data of oil reservoir.

The XZ construction location is in the XZ nosing structure in the neat family in central down warping region district, loose distant basin-Gu Long depression south, and the fundamental purpose layer is one section Putaohua reservoir of system Yao Jia group under the Cretaceous System.Putaohua reservoir end face main body is to be the wide slow nosing structure of pitching in east northeast-southwest, the saddle of the XZ nosing structure transition of pitching for east northeast in southwest.The total structure form is high in the east and low in the west, and high south, north is low, two nosing structures and two tectonic frameworks that distribute between monoclinic phase.Be subjected to tomography cutting, have the structural attitude that thing divides band, grown some level Four structures based on fault block, disconnected nose.East-west direction structure drop is bigger, can see on the Putaohua reservoir end face structural map that drop reaches 600m in East and West direction 15km scope.Secondary structure is zonal arrangement, can be divided into western to oblique sloped region, western graben, middle part horst gentle slope belt, four parts of east graben zone of fracture from West to East.XZ-441.5 survey line west side is ancient imperial to the Xie Dong slope, this district's tomography agensis, and structure is mild, buries darker; Northwest (NW) is relatively growth of degree structure a little in graben, and type based on synsedimentary anticline, brokeback tiltedly; Central authorities' horst gentle slope belt tomography agensis, slope steepening gradually from west to east; North-south, east graben zone of fracture is made up of three groups of tomographies.T _1-1Layer explains that tomography is trap-down, fault strike based on northwest (NW) to, nearly north-south, development length 2.0～10.0km, general 5.0km, turn-off is generally 30.0m, maximum can reach 70.0m.Fault plane is " Y " font and distributes, and under the tomography cutting, the XZ nosing structure is further complicated, helps forming the oil enrichment region.

Putaohua reservoir belongs to one section stratum of Yao, with the green hill mouth group that underlies tangible difference is arranged.On the lithology, the black mud stone of the grey of Putaohua reservoir bottom, celadon, aubergine mud stone and green hill mouth group has tangible interface; On the electrical measurement curve, the Putaohua reservoir resistivity curve is a zigzag, and green hill mouth group resistivity curve is straight, and interval transit time curve is compared with country rock up and down and is tangible low value, generally its half range point is pushed up the layering sign on circle, the end as Putaohua reservoir.In order further to study the single sand body feature of Putaohua reservoir, we are divided into 11 substratums of 3 sandstone groups, the wherein I of Portugal with Putaohua reservoir ₁The I of～Portugal ₃Layer is last sandstone group, the I of Portugal ₄The I of～Portugal ₆Layer is middle sandstone group, the I of Portugal ₇The I of～Portugal ₁₁Layer is following sandstone group, and the skeleton sectional view of the establishment detail correlation of reservoir bed.

The Putaohua reservoir sandstone thickness in Xz area is extremely southern by north, by to eastern attenuate, the g634 graben block of northern sedimentary system control belongs to the lateral margin of placanticline delta body, based on distributary channel under water, the agensis of sand sheet body, sandstone thickness changes (8.0-14.0m) greatly, sandstone thickness is thinner on the whole, the g62 of western sedimentary system control, g652 well graben block is grown nearly north-south sandstone development band, g649-g648-g62 wellblock sandstone thickness is greater than 18.0m, northern, western sedimentary system is after the XZ oil field is converged, at g652, g628 and g639, the g648 wellblock forms East and West direction lensing sandstone development district, more than the sandstone thickness 18.0-20.0m.From Putaohua reservoir individual layer net thickness distribution histogram as can be seen, XZ oil field net thickness accounts for 88.7% of total number of plies less than the number of plies of 1.5m, and thickness accounts for 67.7% of gross thickness.Analysis-by-synthesis, the Putaohua reservoir reservoir is based on thin interbed; Net thickness distributes relevant with the oil reservoir origin cause of formation and type on the plane, being positioned at east, western graben g634, g648-g62 wellblock oil reservoir relatively grows, net thickness is greater than 6.0m, middle part g652,628, x160-76 wellblock lensing lithologic oil pool district form local oil reservoir and grow the district, net thickness is greater than 5.0m, influenced by deposition, g605-g69 wellblock, southeast net thickness is less, is kept to 2.5m gradually by 4.0m.

Putaohua reservoir core analysis statistics, net porosity generally are distributed between 10～23%, average out to 15.5%; Air permeability generally is distributed in 0.3～99.2 * 10 ^-3μ m ², average out to 11.3 * 10 ^-3μ m ²The high value of net porosity district mainly concentrates on XZ nosing structure axial region on the plane, and analyze reason: on the one hand, the reservoir properties distribution characteristics mainly is subjected to the influence of primary deposit system, it is better that distributary channel sand is grown district's rerum natura under water, and it is relatively poor that the leading edge sand sheet is grown district's rerum natura; On the other hand, because XZ structure of oil field drop is bigger, maximum can reach 600m, and compaction also will produce certain influence to its rerum natura.

Nenjiang group deposition latter stage, loose distant basin is owing to be subjected to northwest (NW)--east southeast to the pressure property turned round stress, be accompanied by the protuberance of grand celebration placanticline, form six nosing structures in the placanticline west side.The XZ nosing structure is exactly one of them, and its typing phase is open fire group deposition latter stage, and the xz nosing structure forms period and ancient dragon depression green hill mouth group oil source rock and arranges the hydrocarbon phase in a large number and be complementary, and helps hydrocarbon accumulation and becomes to hide.The xz nosing structure has been controlled favourable oil domain generally.And after the tectonic movement of many phases at end of the Ming Dynasty and Paleogene period end, Gu Long depression center obviously moves westwards, the hydro carbons that impels Gu Long depression hydrocarbon source rock to generate is migrated to grand celebration placanticline direction along xz nosing structure slope, XZ textural " Y " the font activity of many phases of rupturing under tectogenetic influence simultaneously, form complex structures such as local disconnected nose, graben, help assembling in the local trap in the XZ zone of fracture in the hydro carbons migration process.Analyze from mode of deposition in addition, the xz oil reservoir is in northern sedimentary system and west and south sedimentary system alternate zone, and detrital grain is thin, the shale index height, and the east-west direction drop is big in addition, and the Diagn obvious difference helps forming lithologic trap.XZ fault block oiliness difference from different places is big, the profit complex distribution, and rich accumulation of oil and gas is subjected to multifactor controls such as fault block, lithology.

The compound oil reservoir of fault block-lithology is the main oil reservoir type in xz area, basic characteristics are that the profit distribution is controlled by fault block, and concrete oil-containing enrichment positions is restricted by lithological change, in the plane oil domain as a whole with fault block in local structure match, but oil bondary and structure contour are inconsistent.In general, oiliness is good near extending long major fault in east, xz oil field, two zone of fracture in west, and the saddle between west, Portugal, new website nosing structure, especially area, the obsolete slope of tomography oiliness is relatively poor, individual well is mainly seen oil-water common-layer, water layer, as g602, g633 well, g639 wellblock, be the profit complex area; From concrete trap condition, XZ oil field local structure is mainly relevant with the faulting that " Y " font distributes.Form a series of drag structures and rupture band at tomography two dishes, these local structures form the compound oil reservoir of fault block-lithology under lithological change cooperates.

Lithologic oil pool mainly is positioned at the g605 block that the XZ nosing structure is inclined and there is not the position, forms the stratiform lithologic oil pool by lensing, strip sand sheet body, and profit distributes simpler, and individual well is vertically gone up to oil reservoir-dried interlayer mutual, and oil-containing in flakes on the plane.

Xz oil reservoir Putaohua reservoir profit distribution relation is complicated, mainly contains following 5 kinds of forms: full section net pay zone, up and down between oil reservoir therebetween water layer, oil reservoir and oil-water-layer mutually, full well oil-water common-layer and full well water layer, the whole district does not have unified water-oil interface; Vertical upward drop is big, nearly 600m (1280-1900m).The different buried depth reservoir, its lithology, rerum natura and rock compaction degree difference are bigger, and electrical property feature is the concentrated expression of reservoir lithology, rerum natura and fluid properties, in conjunction with this district's geologic feature and well-log information, for reducing the influence of reservoir lithology and rerum natura, divide axial region and alar part (is the boundary with 1600m) to formulate oil-water-layer identification plate respectively to logging trace.

Finish 159 layers in 58 mouthfuls of wells of formation testing by Dec calendar year 2001, obtain the oil reservoir formation testing data (as table 1) of industry oil stream Jing47Kou.

Table 1 XZ area Putaohua reservoir formation testing tables of data

Pound sign	Sequence number	Well section 1	Well section 2	Bed thickness	Level number	The formation testing mode	Day produce oil	Daily output water	Daily gas	The formation testing result
											ao40 1	1	1260	1261.8	2	(33，)	MFEII	2.609	Trace	1.79	Commercial oil reservoir
ao40 2	1	1281	1294.4	8.4	(mend 3, mend 2)	MFEII	0.026		0.03	Low pay sand
											ao40 3	S2-1	1220	1225.8	3.8	(43，44，)	Press the back pumping	0.171		1.64	Low pay sand
ao40 3	C2-2	1220	1225.8	3.8	(43，44)	Press back MFEII	4.060			Commercial oil reservoir
											ao7	4	1245	1250	3.4	(30，31，)	MFEII	2.610	Trace	2.66	Commercial oil reservoir
ao9	2	1298	1322.8	5.4	(outside 10,11,1,2,3)	MFEII	14.750		2.51	Commercial oil reservoir
											ao90 3	1	1506	1507.2	1.2	(45，)	MFEII			2.35	Water layer
ao90 4	C1-1	1569	1570.2	1.4	(15，)	MFEII	0.202		0.19	Low pay sand
											ao90 4	S1-2	1569	1570.2	1.4	(15，)	Press the back pumping	3.889		35.33	Commercial oil reservoir
da11 1	1	1616	1660.2	15.4	(64-72,74, mend)	MFEI	Oil bloom		0.20	Water layer
											da40 1	1	1599	1632.4	15.7	(40, mend 1,41-45)	MFEI	14.306	554	4.77	Commercial oil reservoir
da40 2	1	1656	1683.8	9.4	(37-41，43，44，)	MFEI	2.493	175	2.31	Commercial oil reservoir
											da40 6	3	1689	1715.2	10	(13-17 mends)	MFEI			12.23	Water layer
da40 7	1	1687	1886.2	0.8	(52，)	MFEI	0.005		0.01	Do layer
											da41 4	C3-1	1620	1622	2.4	(43，)	MFEII			8.24	Water layer
da41 9	C1-1	1490	1511.6	8	(43-49，)	MFEI	1.974		1.47	Commercial oil reservoir
											da42 0	S3-1	1681	1683.2	2.6	(45，)	Bailing	0.127		1.03	Low pay sand
....

2. according to electric logging integrated interpretation and well logging show of oil and gas result, determine target reservoir in the well.

According to electric logging integrated interpretation and well logging show of oil and gas result, determine target reservoir in the well, set up all target reservoir tables of data of every mouthful of well of the whole district, as table 2, this step is exactly general said infiltration sandstone integrated interpretation.

Table 2 XZ area Putaohua reservoir electrical measurement parameter list

Pound sign	Classification	Dark side direction resistivity	Compensated neutron	Compensation density	Interval transit time	Spontaneous potential	The nature coffee	Hole diameter
									pu40-32	Oil reservoir	10	19	2.43	82	3	90	8.7
pu401-13	Oil reservoir	12	17.5	2.42	82	4	59	9.5
									pu401-14	Oil reservoir	9	18.5	2.35	90	3	81	8.5
pu401-15	Oil reservoir	11	23.5	2.4	88	8	81	8.6
									pu401-16	Oil reservoir	10	24	2.34	92	4	75	9
ao158-38	Oil reservoir	10.9	22.6	2.33	88	10	59	8.7
									ao158-41	Oil reservoir	9.3	23.6	2.32	88	5	69	8.6
ao904-15	Oil reservoir	23	15	2.4	78	8	52	8.7
									mao17-21	Oil reservoir	23	22	2.33	82	3	60	9.3
mao203-3 9	Oil reservoir	7	27	2.36	94	2	85	9.4
									mao203-4 1	Oil reservoir	6.5	25.5	2.45	88	3	85	9.7
mao203-4 2	Oil reservoir	5	24	2.47	90	4	92	9.4
									pu381-52	Oil reservoir	10.8	22.3	2.42	86	2	77	8.8
pu381-53	Oil reservoir	9.1	25.4	2.36	90	0	93	8.7
									pu381-55	Oil reservoir	16.8	19.1	2.43	76	3	80	8.5
pu381-56	Oil reservoir	23.7	23.9	2.31	88	8	62	8.4
									...

3. choose the electric logging of target reservoir and the reservoir parameter of well logging.

4. use the method for the factorial analysis in the statistical mathematics, the technology that uses a computer is compiled into computer program, in conjunction with the 1. production fluid of resulting target reservoir (oil, gas, water) data, to 3. multinomial electric logging of resulting target reservoir and logging parameters, set up oil-gas-water layer classification plate.

Set up oil, gas, water layer identification plate with factor analysis, requirement is clear to the attribute of each sample of raw data, just will there be the individual layer oil test data at the scene, perhaps multilayer is closed the pure oil of test manufacture, perhaps multilayer is closed the test manufacture pure water, this multilayer is closed should the possibility of doing layer in the examination, set up in the plate process and should reject.The data informations such as formation testing of XZ area Putaohua reservoir are just meeting the requirements, and differentiate plate so this method of utilization has been set up oil reservoir, water layer and the oil-water common-layer of XZ area Putaohua reservoir, as shown in Figure 1.Raw data adopts 34 oil reservoirs of 16 mouthfuls of wells, 9 water layers, 8 oil-water common-layers, the logging trace of selecting for use has: seven parameters such as dark side direction resistivity, compensated neutron, compensation density, interval transit time, spontaneous potential, GR, hole diameter, wherein there are two oil-water common-layers to be strayed into the oil reservoir district, the plate precision is 94.1%, and effect is quite desirable.

Embodiment two:

1. investigate and study the geologic condition of oil reservoir, obtain formation testing, pilot production or the produce oil data of oil reservoir.

Bycg oil reservoir structure is positioned at western part, northern central down warping region district, loose distant basin, strides dragon and tiger and steep-pacifies terrace and Qi Jia-Gu Long two second-order structure zones that cave in greatly.Exploration fundamental purpose layer is the FY oil reservoir on four sections in Cretaceous System Lower Cretaceous Series spring and three sections tops of spring.

From F oil reservoir end face structural map, this area is positioned at western part, northern central down warping region district, loose distant basin.Stride dragon and tiger and steep-pacify terrace middle part and neat family-Gu Long two second-order structure zones that cave in greatly.Overall be the ancient monocline of Ta Laha to oblique western lifting, and fracture is grown, active be three groups of north north east to zone of fracture, from be followed successively by the extra large zone of fracture of Harrar to the east of the west, Bayan is looked into dried zone of fracture, Taihe county zone of fracture.

From the F of this area oil reservoir end face structural evolution trend, local area deposits period to Nenjiang group deposition at the F oil reservoir, is that construct on a long-term slope that tilts from the north-westward southeast of growing on the main body, and presents the structural attitude of local buckling along Harrar sea zone of fracture.Its Zhongquan head group is subjected to tectogenetic the influence to green hill mouth group deposition, form a large amount of T ₂Layer north north east is to tomography, comprise that it all is to form this period that typical Harrar sea, Bayan are looked into dried, Taihe county zone of fracture, eastern generally thick west, stratum is thin, form strong deflection along Harrar sea zone of fracture, Ta Laha is to tiltedly having begun to take shape, this tectonic framework develops with regard to inheritance before Yao Jia group deposition, but attitude of stratum tends towards stability gradually, and Ta Laha is to oblique axially being offset eastwards gradually by north and south, Nenjiang group and twice tectonic movement of open fire group, Ta Laha makes it typing to tiltedly beginning to take shape up to Paleogene period tectonic movement in latter stage.Because this district's tectonic framework formation time is early, adjoin Gu Long depression oil generation center again, therefore arrange the hydrocarbon phase in a large number at blue or green one section oil source rock, this district has certain sensing and inducing action to oil-gas migration.

Main oil bearing reservoir are the delta distributary channel deposit.Sand body is nearly East and West direction and stretches, from west to east attenuate gradually.Fuyu County's oil reservoir list layer of sand number is generally at the 4-8 floor in the district, and single sand layer thickness is generally at 0.8-4.0m, and boring and meeting the sandstone total thickness is 6-40m, and sandy ground is than changing at 6.5-48.6%.

Fuyu County, Bycg area oil reservoir is controlled by mainly in western thing source, is delta front and delta plain deposition, and sand body is based on waterborne, distributary channel sand and crevasse-splay deposit under water.Fuyu County's reservoir rock type belongs to feldspar landwaste packsand, detrital grain is based on 0.05-0.15 μ m, shale index is 2-10%, clay mineral is mainly based on illite, secondly content be chlorite and illite/smectite mixed layer at 3-95%, and content is respectively at 1-50% and 1-15%, contain a small amount of smalite, content is 2-4%.Cementation type has forms such as regeneration-hole, regeneration-film, hole-film, regeneration-contact, regeneration-filling, filling-film like.Analyze per sample, Bayan is looked into and is done regional Fuyu County oil reservoir net porosity at 7-15%, and permeability is in 0.07-2 * 10 ^-3μ m ²Find out that from factor of porosity, permeability histogram factor of porosity mainly is distributed in 8-13%, permeability mainly is distributed in 0.2-2 * 10 ^-3μ m ²Average pore is 12%, and mean permeability is 0.5 * 10 ^-3μ m ²Sample distribution is more diffusing, shows that reservoir heterogeneity is strong, thereby it is bigger to make this distinguish each well production rate difference.

By the end of calendar year 2001 by the end of September, finished 44 mouthfuls of prospect pits in the district, get 20 mouthfuls in core, 21 mouthfuls of formation testings.The formation testing data see Table 3.

Table 3 Bycg area formation testing tables of data

2. according to electric logging integrated interpretation and well logging show of oil and gas result, determine target reservoir in the well

According to electric logging integrated interpretation and well logging show of oil and gas result, determine target reservoir in the well, set up all target reservoir tables of data of every mouthful of well of the whole district, form such as table 4, this step is exactly general said infiltration sandstone integrated interpretation.

3. get the reservoir parameter of well logging of the electric logging of target reservoir earlier

This distinguishes most of formation testing all is that profit goes out together, and individual layer formation testing data are few, can't set up oil-gas-water layer identification plate with factorial analysis like that to embodiment two.Set up oil-gas-water layer identification plate with factorial analysis, requirement is clear to the attribute of each sample of raw data, just will there be the individual layer oil test data at the scene, perhaps multilayer is closed the pure oil of test manufacture, perhaps multilayer is closed the test manufacture pure water, this multilayer is closed should the possibility of doing layer in the examination, set up in the plate process and should delete.Yet, in many areas often is that full well closes formation testing water with going out, these close the examination layer might all be oil-water common-layer, also have net pay zone and pure water layer, oil-water common-layer perhaps out of the ordinary, owing to then can't determine the production fluid attribute of each destination layer like this, just can't set up oil-water-layer identification plate with factorial analysis.Past runs into this situation at work, all can not set up oil-water-layer with any traditional method and differentiate plate.Adopt the method for correspondence analysis just can solve this difficult problem well.Correspondence analysis is a kind of multivariate statistical analysis method that grows up on the factorial analysis basis, it is that R type and the factorial analysis of Q type are combined, the method that variable and sample unification are analyzed and researched, advantage is that sample and variable can be reflected on the same factor planar graph simultaneously, are convenient to variable and sample unification are carried out geologic interpretation and deduction intuitively.

Briefly say, on the factor planimetric map of correspondence analysis, can provide following information: the relation between the i variable, some contiguous variable points represent that these variablees are closely related, promptly they have the contact on the origin cause of formation, indicate the geologic function of a certain characteristics; The relation of ii sample room, contiguous sample spot has similar character, belongs to same type, is the product of same geologic function; The relation of iii variable and sample, the sample spot of same type will be characterized by contiguous variable point, that is to say, and similar sample spot is close to product under the indicated geologic function of variable point for it.

Fuyu County, bycg area, Yang Dacheng seed oil layer, 53 layer data of 13 mouthfuls of wells of formation testing that can be used for plate are collected, many wells all are the examination data of closing that profit goes out together, and the logging trace of selecting for use has: seven parameters (table 4) such as dark side direction resistivity, microballoon focused resistivity, spontaneous potential, interval transit time, compensated neutron, compensation density, GR.

Fuyu County, table 4 bycg area, Yang Dacheng seed oil layer electrical measurement parameter list

Pound sign	The microballoon focused resistivity	Dark side direction resistivity	Interval transit time	Compensated neutron	Compensation density	Spontaneous potential	GR
								L19-36	110	43	70	12.5	2.51	0	55
L19-38	38	30	72	13	2.5	2	53
								L23-27	35	20	74	14	2.4	0	78
L23-34	70	42	73	12	2.47	0	70
								L23-35	40	31	73	13	2.5	0	72
L23-38	40	29	73	14	2.46	0	70
								L23-39	60	33	72	11	2.51	0	75
L25-32	70	43	71	12	2.46	0	82
								L25-33	41	31	70	12.5	2.46	0	80
L25-38	40	27	68	9	2.48	0	78
								T231-21	52	40	68	12.5	2.5	2	60
T232-34	18	38	70	10	2.47	4	60
								T282-24	60	60	72	16	2.42	0	60
T282-25	39	35	72	15	2.44	0	67
								T282-26	35	39	73	19	2.35	2	60
....

4. use the method for the correspondence analysis in the statistical mathematics, the technology that uses a computer is compiled into computer program, in conjunction with the 1. production fluid of resulting target reservoir (oil, gas, water) data, to 3. multinomial electric logging of resulting target reservoir and logging parameters, set up oil-gas-water layer classification plate.

The utilization factor analysis is carried out classification based training, and the correspondence analysis difference is the result show, the first main gene (F ₁) variance contribution account for 56.9%, the second main gene (F ₂) variance contribution account for 26.3%, the two the accumulation variance contribution be 83.2%, can represent whole data matrix fully.7 log parameters are at F ₁And F ₂Score on the main gene load axle sees Table 5.

Table 5 correspondence analysis factor loading score

As can be seen from Table 5, the higher log parameter of score absolute value is respectively dark side direction resistivity, spontaneous potential and GR.Differentiate plate from Fuyu County, bycg area, Yang Dacheng seed oil layer oil-water-layer, as shown in Figure 2, promptly can obviously find out on the factor loading figure: dark side direction resistivity, spontaneous potential and GR are distributed in 3 end points on the factor planimetric map respectively.According to the ultimate principle of electric logging and exploration practices as can be known, the dark side direction resistivity value of oil reservoir is higher, water layer spontaneous potential amplitude is big, the big reflection reservoir of GR contains the shale height and causes moisture height, in conjunction with indivedual single examination layer data, divide 3 districts according to the intensity of factor planar sample, be respectively oil reservoir district (44 points), pelagic division (8 points), oil-water common-layer district (7 points).Thereby set up oil-water-layer identification plate after the pressure of Fuyu County, bycg area, Yang Dacheng seed oil layer.

In addition, oil-water-layer that correspondence analysis is set up is differentiated plate, be profit from many wells with oil reservoir, water layer and the oil-water common-layer of determining by classification based training the examination data that close that goes out, be not strayed into sample, also just do not have the plate precision problem, this point is different with previous methods.

In order to verify the practicality of this method, to this distinguish all not the reservoir of formation testing differentiate again, totally 8 mouthfuls of wells are 29 layers, confirm that through actual formation testing its result meets substantially, have only No. 15 floor of T23 well, No. 26, No. 28 3 floor erroneous judgement of T34 well to be oil reservoir, illustrate that the explanation precision that oil-water-layer is differentiated has reached 89.6% (table 6), effect is fairly good.

Table 6 bycg holds up in the area poplar layer oil reservoir oil-water-layer decryption table

Claims (1)

1, a kind of multi-parameter dimension-reducing oil-gas-water-layer identifying method is characterized in that: this method comprises the steps:

1. investigate the geologic condition of oil reservoir, collect formation testing, pilot production or the produce oil data of oil reservoir;

2. according to electric logging integrated interpretation and well logging show of oil and gas result, determine target reservoir in the well;

3. choose the electric logging of target reservoir and the reservoir parameter of well logging;

4. use the method for statistical mathematics-factorial analysis, technology uses a computer, be compiled into computer program, integrating step is the production fluid oil of resulting target reservoir 1., production fluid gas, production fluid water number certificate, to step multinomial electric logging of 3. resulting target reservoir and logging parameters, carry out the linear combination dimensionality reduction to the binomial main factor, it is main gene, be that the variance contribution of P=2 and binomial main gene accumulative total was above 80% o'clock, foundation can be illustrated in the oil-gas-water layer classification plate of two dimensional surface, also claim oil-gas-water layer identification plate, the i.e. parameter amount (n) that can obtain according to electric logging and well logging, a plurality of samples (m) of target reservoir therefrom extract 2 principal elements, and concrete grammar is as follows:

Be provided with n sample, each sample comprises m variable, and the value of a n sample m variable is write as the data matrix form:

$Z=\left[\begin{array}{cccc}{z}_{11}& z_{12}& ......& z_{1n}\\ z_{21}& z_{22}& ......& z_{2n}\\ z_{m1}& z_{m2}& ......& z_{\mathrm{mn}}\end{array}\right]$

In matrix Z, z _j=(z _1jz _2j... z _Mj) ' (j=1,2 ..., n) observed reading of a j sample m variable of expression, i.e. j the sample of Z; z _i=(z _I1z _I2... z _In) (i=1,2 ..., n) observed reading of i variable of expression in n sample, i.e. i the variable of Z;

If variable z _iStandard deviation standardization (abbreviation standardization) variable be x _i, and a corresponding observed reading is designated as matrix

$X=\left[\begin{array}{cccc}{x}_{11}& x_{12}& ......& x_{1n}\\ x_{21}& x_{22}& ......& x_{2n}\\ x_{m1}& x_{m2}& ......& x_{\mathrm{mn}}\end{array}\right]$

For any two variable x _i=(x _I1x _I2... x _In) and x _j=(x _J1z _J2... z _Jm) their phase $r_{\mathrm{ij}}=\frac{1}{n-1}\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{x}_{\mathrm{ik}}\·x_{\mathrm{jk}}$ (i，j＝1，2，…，m)

The pass coefficient is

The related coefficient of m variable constitutes the matrix of a m * m

$R=\left[\begin{array}{cccc}{r}_{11}& r_{12}& ......& r_{1n}\\ r_{21}& r_{22}& ......& r_{2n}\\ r_{m1}& r_{m2}& ......& r_{\mathrm{mn}}\end{array}\right]$

And r _Ij=r _Ji, r ₁₁=r ₂₂=...=r _Mn=1, and claim that R is the correlation matrix of variable

Any two sample x _i(=x _1ix _2i... x _Mi) ' and x _j(=x _1jx _2j... x _Mj) ' similarity coefficient be

$q_{\mathrm{ij}}=\frac{\underset{k=1}{\overset{m}{\mathrm{\Σ}}}{x}_{\mathrm{ki}}\·x_{\mathrm{kj}}}{\sqrt{\underset{k=1}{\overset{m}{\mathrm{\Σ}}}{x}_{\mathrm{ki}}^2\underset{k=1}{\overset{m}{\mathrm{\Σ}}}{x}_{\mathrm{kj}}^2}}$ (i，j＝1，2，...，n)

The similarity coefficient of n sample is write as matrix form

$Q=\left[\begin{array}{cccc}{q}_{11}& q_{12}& ......& q_{1n}\\ q_{21}& q_{22}& ......& q_{2n}\\ q_{m1}& q_{m2}& ......& q_{\mathrm{mn}}\end{array}\right]$

And q ₁₁=q ₂₂=...=q _Mn=1, q _Ij=q _Ji, claim that Q is the similar matrix of sample

R mode factor analysis is the inner structure of research correlation matrix R, therefrom finds out p generalized variable f that all variablees is started to control making usefulness _k=(k=1,2 ..., p, p＜m), and variable x _iBe expressed as f _kLinear combination, promptly

x _i＝a _i1f ₁+a _i2f ₂+...+a _ipf _p+a _ie _i (i＝1，2，...，m)……(1)

When p＜＜during m,, and further explore the controlling factor of Changing Pattern in the origin cause of formation contact of variable and the space by formula (1) Study on Simplifying system

The factorial analysis of Q type is by the research to similar matrix Q inner structure, seeks p generalized variable f of restriction sample similarity _k=(k=1,2 ..., p, p＜n), and sample x _iBe expressed as f _kLinear combination, promptly

x _j＝a _j1 f ₁+a _j2 f ₂+...+a _jpf _p+a _je _j (j＝1，2，...，n)……(2)

When p＜＜during n, by formula (2) abbreviation sample research system;

It is to be noted at this: formula (1) and formula (2) also are the basic assumption conditions of R type and the factorial analysis of Q type.E wherein _iAnd e _jBe respectively to obey that average is 0, variance is σ _i ²And σ _j ²Normal distribution;

5. with 4. middle oil, gas, the water layer plate of setting up of step, other target reservoir of this oil reservoir is carried out oil-gas-water layer differentiate;

Above-mentioned for the oil reservoir that lacks the individual layer formation testing, before setting up oil-gas-water layer classification plate, also will be in conjunction with corresponding analysis method; Be meant in uncased hole in the described well; The electric logging parameter of described target reservoir is the data more than three in microballoon focused resistivity, dark side direction resistivity, spontaneous potential, interval transit time, compensated neutron, compensation density, the GR; The logging parameters of described target reservoir is meant the geochemical logging data; The logging parameters of described target reservoir is the gas detection logging data, total hydrocarbon, hydrocarbon component, nonhydrocarbon component.

Images