Using image processing technology to study aeromagnetic and aeroanomalies in oil and gas fields

(Aerogeophysical Remote Sensing Center, Ministry of Geology and Mineral Resources, Beijing 100083)

Using image processing technology, the following problems related to oil and gas field anomalies are studied: rapid extraction of high-frequency weak magnetic anomalies, estimation of buried depth of high-frequency magnetic anomalies, normalization of radioactive anomalies, analysis of typical anomaly profiles and evaluation of oil and gas prospects, and four promising oil and gas anomalies are put forward. After verification, three of these four anomalies show oil and gas, of which 1 is industrial gas flow.

Key words: oil and gas field, superposition effect, image processing, aerial Th normalization processing, oil and gas prospect evaluation.

I. Introduction

In recent 10 years, people have paid more and more attention to the non-seismic geophysical exploration of oil and gas fields [1-8], one of the reasons is that the production cost of seismic exploration is increasingly expensive; On the other hand, in recent 10 years, great changes have taken place in aeromagnetism and aviation radiation technology. The sensitivity of pneumatic optical pump magnetometer is improved from 1nT to 0. 1 ~ 0.0 1NT. The introduction of large volume square cylindrical NaI(Tl) crystal and digital circuit into the airborne radioactivity measuring instrument not only improves its sensitivity 10. All these make geophysicists pay more attention to the application of large-area and low-consumption airborne geophysical prospecting technology in oil and gas exploration.

The distance of flight line measured by high-sensitivity aeromagnetic and airborne gamma-ray spectrometry in central Qaidam is 65438±0km, and the flying height is about 90 m; The aeroradio and aeromagnetic integrated station installed on the Twin Otto aircraft includes GAD-6 four-channel spectrometer, MAP-4 proton precession magnetometer, tape recorder, Doppler radar and aerial photography. The spectrometer uses three boxes of NaI(T 1) crystals with a total volume of 48L, and the sensitivity of the magnetometer is 0.5nT.

It has been more than 50 years since radioactive gamma measurement was used for oil and gas exploration. Practice shows that there are low radioactivity anomalies above the oil and gas fields, and there are abnormal peaks at the edge of oil and gas structures that are lower than the original ones. 1979, Donovan T.J and others in the United States discovered the high-frequency magnetic anomaly in cement oil fields [13], and people realized that the existence of this high-frequency magnetic anomaly related to oil and gas fields [9] has exceeded 10 years. Although there are many controversies about the mechanism of these anomalies, the most accepted hypothesis is stack effect. It is generally believed that almost all oil storage structures drain water from the deep, and stack effect brings hydrocarbons and other organic derivatives to the surface through three leakage ways: overflow, solution escape and gas diffusion. In the process of upward migration from oil and gas reservoirs, liquid and gas in the formation form magnetite and precipitate radioactive elements due to afteraction, thus forming aeromagnetic and aerial photography anomalies, which can be found by aerial survey and used as a sign to judge underground oil and gas accumulation.

Image processing technology is put forward to study this problem, based on two considerations: (1) large-area macro data of airborne geophysical exploration can quickly obtain intuitive results through image processing; The image system has the processing function suitable for potential field data (aeromagnetism) and statistical data (aeromagnetism). The basic image grid made in this paper is 3 17.5× 3 17.5m2 ... Besides aeromagnetism and aeroradio, MSS(7, 5 and 4 bands) data of Landsat images are also used.

Second, the study of abnormal characteristics.

1. Fast extraction of high-frequency weak magnetic anomalies

Qaidam basin is one of the important oil-bearing basins in China. There are many underground oil and gas displays in the exploration area. There are few people in the survey area and less human interference. It is an ideal test site to study the characteristics of high-frequency weak magnetic anomalies caused by magnetite after oil and gas fields.

The dynamic range of magnetic field in the survey area is 0 ~ 533 nt, and there are only 12 pixels at the low end of 0 ~ 127 nt, accounting for 5/ 100000 of the whole map, and it is concentrated in the local negative anomaly in the northeast, so the method of limiting stretching is adopted:

Zhang Yujun on new methods of geological exploration.

Where z(x, y) is magnetic field anomaly, and the unit nT, l(x, y) is image gray level or gray scale.

Each gray level represents 1.598 nT, so it can be seen that after clipping and stretching, more weak abnormal information is retained in the original magnetic image, and then a three-dimensional shadow map is made according to formula (2) [10] (color layout 9(A)).

Zhang Yujun on new methods of geological exploration.

Where θ is the azimuth of the light source; φ is the height angle of the light source; λ is the included angle between the normal vector of the magnetic plane and the light source vector.

The color stereoscopic shadow map of the magnetic field in the survey area shows that there are a group of high-frequency weak magnetic anomalies along the diagonal from the northwest corner to the southeast corner, with a width of 1 ~ 2 km and an amplitude of 2 ~ 15 nt, and the black-and-white image has 3 ~ 20 gray changes.

Literature [2] holds that drilling casing and manhole cover may cause anomalies with amplitude of 2 ~ 5 nt, but their wavelength is short (< 300 m), that is to say, the anomalies caused by casing and manhole cover are only isolated points on the 3 17.5×3 17.5m grid image. In order to highlight the useful weak magnetic anomalies, the template in figure 1 is used as the horizontal second derivative image, and the isolated anomalies are eliminated by neighborhood filtering.

Attached Figure 9(B) of the color swatch is a superimposed image of aeromagnetic second derivative, normalized U and abnormal distribution of oil and gas. The cyan point group is the second derivative anomaly of aeromagnetic level; Red is normalized u; Green and blue ellipses are known abnormal positions; The red ellipse is a newly discovered promising anomaly. All the known 10 oil and gas displays have high-frequency weak magnetic anomalies, which strongly shows that the second-order horizontal derivative diagram of magnetic field can be completed in just a few minutes, which is of great significance for extracting high-frequency anomalies related to oil and gas fields.

Figure 1 Deriving Convolution Template

2. Buried depth of high frequency weak magnetic anomaly magnetic source

Perform fast Fourier transform (mirror symmetric edge expansion) on the magnetic field, then calculate the power spectrum of the transformation result, take the radial cross section value from the power spectrum image, and finally make a radial power spectrum diagram (Figure 2). For the convenience of drawing, the number of points on the image is directly used as the abscissa.

The spectrum shows that the whole map consists of low-frequency power spectrum Pl and high-frequency power spectrum Ph, the low-frequency part is mainly caused by the magnetism of pre-Devonian metamorphic rocks in this area, and the high-frequency part is mainly caused by local mineralization of shallow post-strata. Calculate the buried depth of magnetic source by parameter method [14].

Zhang Yujun on new methods of geological exploration.

take the logarithm

Zhang Yujun on new methods of geological exploration.

Figure 2 Radial power spectrum

If plotted in semi-logarithmic coordinates, Equation (4) is a linear equation, -2h and -2h are the straight line slopes of low frequency band (1 ~ 8 points) and high frequency band (12 ~ 50 points) respectively, lna and lnb are the intercepts of two straight lines respectively, and△ r is the radial frequency interval.

R=2π/ (number of points × spacing between points) (5)

The number of points in this survey area is 5 12, the distance between points is 3 17.5m, and R=38.65× 10-6m. The intercept and slope of two straight lines are obtained by least square method.

a=254. 1429，Kt=- 13.5595，b= 1 13.72 12，Kh=-0.9 182，

Zhang Yujun on new methods of geological exploration.

S is introduced into the formula because the logarithmic power spectrum value displayed by image processing has only relative significance, and the unit of S is m, and its value can be obtained from the buried depth of known abnormal sources. It is known that low-frequency magnetic anomalies are caused by pre-Devonian metamorphic rocks with a buried depth of -5000 to-15000 meters, then

Zhang Yujun on new methods of geological exploration.

Substituting h, s is 0.0285~0.0855m m.

Zhang Yujun on new methods of geological exploration.

That is, the buried depth of high frequency weak magnetic source is between 0.3 ~ 1 km.

3. The reflection of aerial images on oil and gas reservoirs

The aerial three-element synthetic image [1 1- 12] is as shown in the color board figure 9(C), on which the aeromagnetic secondary horizontal derivative anomaly is superimposed. Red is K, and its content varies from 0 to 7.8%. The green color is Th, and the content varies from 0.0 to 39×10-6. Blue is U, and its content varies from 0 ~ 14.9× 10-6, and the color code is shown in the upper picture of the color swatch 9(D). The yellow dot group is the second derivative anomaly of aeromagnetic level. Compared with the second derivative of aeromagnetic level, the correlation between aerial images and oil and gas display is much more complicated. Although not all known oil and gas displays have clear aviation anomalies, some oil-bearing structures are clearly reflected. For example, the No.5 anomaly in the color version of Figure 9(C) is a known oil-bearing dome in Huya Lake. On the binary map (the color code in the figure below the attached drawing of color swatch 10(A) and attached drawing of color swatch 9(D)), the Yahu structure is also obvious; The upper part of the figure is a composite image of R-Th, G-K and B-K, and the arrows refer to the known oil-bearing structures in Jinaier, Dongtai. The lower part is a composite image of R-Th, G-U and B-U, and the arrows refer to known oil-bearing structures Sebei 1 No.2, all of which show radioactive anomalies with high periphery and low middle.

How to enhance or highlight the information related to oil and gas in aviation data is the key to reduce or suppress the influence of lithology. So we use principal component analysis (called KL transform in image processing) and Th normalization [15].

The attached drawing of the color board 10(B) is an aerial three-factor principal component analysis (KL transform) image, and the upper color standard of the attached drawing of the color board (D) is also its legend; In the figure, A, B and C are the 1, the second and the third principal components respectively, which are represented by KL (1), KL(2) and KL(3) respectively, and D is the synthetic image. The composition of each main component is as follows:

Zhang Yujun on new methods of geological exploration.

Where Xk, XTh and x represent The content values of pixels k, th and u, respectively.

In 1 principal component ξ 1, K is dominant, which reduces the role of Th; Because there is a positive correlation between K and Th in most sandstone and conglomerate, KL( 1) processed by this method greatly suppresses the influence of lithology and highlights the reflection of geological processes unrelated to K and Th, such as the deposition and superposition effect of potassium salt.

ξ2 synthesizes the changes of K, Th and U, reflects the geological environment of the whole region, and has no obvious effect on the study of local oil and gas anomalies.

U is dominant in ξ3, which weakens the role of Th and K and greatly inhibits the lithologic influence, which is very important for studying local anomalies of oil and gas.

In order to reduce the influence of lithology, according to the literature [15], when sorting out the ground or airborne radioactive measurement data, it is often divided by the Th content value, which is called Th normalization processing. When doing This treatment, it is considered That th can represent the change of lithology in gravel sediments, and K, U and Th have certain positive correlation in these sediments, so the influence of lithology can be reduced or suppressed by dividing by th value. Our improvement is to carry out correlation analysis before normalization, get the correlation coefficient, and then normalize according to the correlation coefficient. The attached drawing 10(C) of the color swatch is the comparison diagram of K and U before and after normalization, the original drawing is on the left, and the color code is shown in the attached drawing 9(D) of the color swatch; On the right is the normalized result.

Table 1 is a statistical table of known and newly discovered anomalies, which has * * 8 characteristics: K original image, K transformed by KL, K normalized by th, U original image, U transformed by KL, U normalized by Th, U halo and halo reflection on MSS (bands 7, 5 and 4) images. Eight (80%) of the known 10 oil and gas displays have obvious or obvious aviation anomalies; Its typical characteristics are: The peripheral halo of low k, low u and high u in the center of the th normalized map.

Th normalization has obvious abnormal effects on oil-bearing structures as follows: changing the K height of V 1, V2 and V5 to K low, and significantly reducing the U height of V 1 and V5. Therefore, standardization has obvious effect on unifying abnormal attributes.

4. Typical abnormal profile images

Hu Ya (V2) and Hongsanhan No.4 (V 1) were studied as typical known anomalies.

Hu Ya anomaly has the following typical characteristics (color attached figure 10(D)):

Before and after normalization (1)U (blue curve on the right of the color swatch 10(D)), the abnormal center is low;

(2)K was originally a high value in the center of the anomaly, but it was also a low value after normalization (the red curve on the right of the swatch chart 10(D));

(3) There are obvious U peaks and halos around the anomaly;

(4) There are obvious high-frequency weak magnetic anomalies, and the center is strong (color chart 10(D) green curve on the right).

Table 1 Statistics of known and newly discovered anomalies

The abnormal characteristics of Hongsanhan 4 are as follows:

(1)U is strong and high in the abnormal center before normalization, and significantly decreased after normalization.

(2)K is high before normalization and low after normalization;

(3) After normalization, obvious U peak appears around the anomaly, reflecting the halo phenomenon;

Fig. 3 Surface energy spectrum measurement profile of Hongsanhan No.4.

(4) There are obvious high-frequency weak magnetic anomalies.

Based on the comprehensive analysis of typical anomalies, the identification marks of aeromagnetic and aeromagnetic anomalies related to oil and gas reservoirs are: local high frequency and weak variation of aeromagnetic; After Th normalization, local potassium and uranium are low (some of them can't reach low values); And the high value halo phenomenon of uranium around the anomaly.

We conducted a ground inspection of the Hongsanhan No.4 anomaly, and measured the profile with four energy spectrometers. the total length 10km and the point distance 100 ~ 200m ... The profile (Figure 3) obtained is consistent with the aerial survey results, and the height of K, U and Th is at the center of the anomaly, which first decreases and then increases from the center to both sides.

Three. Discussion on oil and gas prospect evaluation

The known oil-bearing structures or points in this survey area are *** 10, as shown in table 1: Hongsanhan No.4 (V 1, gas), Hu Ya (V2, gas), Ma Xin Gaodian (V4, oil), Dongtai Jinaier (V5, see industrial oil flow) and Nannan. These 10 known oil and gas displays all have obvious aeromagnetic second derivative anomalies without exception. The first eight have obvious or obvious aviation anomalies, which are manifested as the central low value and peripheral halo phenomenon of U on the K-element image and U-element image normalized by Th. The aviation anomalies in the latter two known areas are not obvious enough, mainly because the two known oil and gas display points are located in areas rich in potash deposits, and the radioactive changes caused by potassium are stronger than the weak radioactive anomalies related to hydrocarbons, which are not obvious. Therefore, the coincidence rate of known oil and gas display structures or points in this survey area with aeromagnetic horizontal second derivative anomalies is 100%, and the coincidence rate with aeromagnetic K and U anomalies is 80%.

Li luling and others reported the results of detailed aeromagnetic survey in the central Qaidam basin. Aerogeophysical Remote Sensing Center, Ministry of Geology and Mineral Resources, 1985.

The oil and gas prospect of this survey area is evaluated by traditional methods, and eight most promising oil and gas areas are circled, and the evaluation results are completely confirmed by image processing. After image processing, we think that the following four promising oil and gas anomaly areas should be added:

(1) yikeraoru anticline structure (V3) is located 30km east of Hu Ya oil and gas anticline structure and15km north of Taj Nair oil and gas anticline. The anomalies of aeromagnetic, aeroradiometric and satellite images are very typical, very similar to the Hu Ya anomaly, and the basis of anomaly characteristics is sufficient, which can be regarded as the most promising oil and gas anomaly newly discovered.

(2) Nanlingqiu west anomaly (V 10) is located about 5km west of known oil-bearing structures in Nanlingqiu. There are local high frequency and low amplitude magnetic anomalies, low potassium and uranium, and obvious high value U halo with closed periphery. It can be used as a new and promising oil and gas anomaly.

(3) The flood anomaly (V 1 1) is located in the upper part of the taijinaier diluvial fan. In addition to local high-frequency weak magnetic anomalies, there are low U and faintly visible peripheral U high halo, which can be used as promising oil and gas anomalies newly discovered.

(4) The Nanbaxian anomaly (V 14) is located in the northern margin of the southwest high point in xian county. There are local high frequency weak magnetism, low potassium and low uranium anomalies. It can also be used as a newly discovered promising oil and gas anomaly.

As mentioned above, Figure 9(B) of the color swatch shows the known (blue, green and yellow ellipses) and four newly discovered oil and gas anomalies (red ellipses) at l0. It is superimposed with the normalized U (red) and the second derivative anomaly (green dot) of aeromagnetic level.

The above prospect evaluation was completed in 1990. According to the exploration results of the Ministry of Energy in the central Qaidam Basin in 199 1 year, the comparison is as follows: in the above four oil-bearing prospect areas, oil and gas shows have been seen in three areas: that is, oil and gas shows have been seen in the Yike Jaú soft oblique structure (V3) of the "Zhong Yi 1" well, and flood anomalies have occurred in Taichung.

Of the four oil-bearing prospects predicted in this work, three have been shown, which shows that it is effective to study the abnormal characteristics of oil and gas fields by using aeromagnetic and aeroradiometric parameters for image processing. This method is used to predict the anomalies not found by conventional methods, and it is verified, which shows the value of this method.

Other personnel involved in the field verification include Shui Enhai, Shi and Guo Yi. The screen images were all taken by Yang and supported by Ye Hefei from the Institute of Remote Sensing Geology of the Petroleum Exploration and Development Institute of the Ministry of Energy. I would like to express my sincere thanks here.

refer to

Guo Wulin Wald. Using aeromagnetic and radioactive survey to directly search for oil and gas reservoirs. Geological science and technology trends, 1984, (24): 2 1 ~ 24.

Guan zhining. Exploration of oil and gas reservoirs related to chemical remanence by aeromagnetism. Foreign geological exploration technology, 1985, (7): 18 ~ 23.

, ten thousand, gold. Further test of magnetic oil prospecting and research on related problems. Petroleum geophysical exploration, 1985, (20): 397 ~ 404.

Guo yukun, ed. Summary of geophysical direct oil exploration methods in Soviet Union. Translation series of petroleum geophysical exploration, 1985, (1): 90 ~ 95.

[5] Wang Xiuwen's translation. New prospect of oil and gas radioactive exploration. Geological science and technology trends, 1984, (18): 17 ~ 20.

Zhao improved the compilation. Radiological methods for direct exploration of oil and gas fields. Translation Series of Petroleum Geophysical Exploration, 1987, (6): 95 ~ 104.

Li Shuyi. Application of nuclear technology in petroleum, natural gas and geothermal exploration. Foreign geological exploration technology, 1986, (4): 22 ~ 27.

Edited by Wang Xitian. Remote sensing detection of hydrocarbon microleakage in oil and gas exploration. Geological science and technology trends, 1987, (11):13 ~16.

Cai Zhenjing. Application of high-precision large-scale (1∶50000) aerial geophysical prospecting in oil and gas reservoir exploration. Geophysical and geochemical exploration,1989 (13): 401~ 410.

Dr. Bloom Tesky. Computer programs for making shaded landforms and three-dimensional shaded landforms. GSC paper. 84— 1 B，375—389， 1984

Zhang Yujun. Application of image processing technology in airborne radiation data. Minutes of the 2nd Symposium on Exploration Geophysics, Xi 'an, 525-526, 1986.

Zhang Yujun. Digital image processing of airborne radioactive and magnetic data in central Qaidam basin. Overview of exploration geophysics in China, 517-535, 1989.

Donovan, T.J., Fauci, R.L., Roberts ... Aeromagnetic exploration of diagenetic magnetite above the oilfield. Bulletin of American Association of Petroleum Geologists, 63,245—248, 1979.

Donovan, T.J., Bryan, D.P. Bryan, J.G., Cunningham, K.I. Near-surface magnetic display of buried hydrocarbons. Separation of aeromagnetic detection and false signals, IGARSS' 87 remote sensing: understanding the earth as a system, 219-232, 1987.

[15]Alton V.Gallagher. Radiation measurement-a practical detection technology. Weotern oil reporter in March, 173, 1982.

Research on aeromagnetism and aerodynamics. Study on radioactive anomalies in some oil and gas fields by image processing technology

Zhang

(Aviation Geophysics and Remote Sensing Center, Ministry of Geology and Mineral Resources, Beijing 100083)

abstract

This work represents the results of studying the following problems with image processing technology: rapid extraction of high-frequency weak magnetic anomalies, estimation of the depth of high-frequency magnetic anomalies, normalization of radiation anomalies, analysis of typical anomaly profiles and prediction of finding oil and gas potential anomalies. Four new anomalies of oil and gas prospect have been discovered. This study was completed in early 1990. Compared with the exploration results of 199 1, we are encouraged that not only oil and gas were found in three of the four abnormal areas, but also natural gas was found in one of them.

Keywords oil and gas field, stack effect, image processing, second-order horizontal gradient of magnetic field, power spectrum, Th normalization of airborne radiometry, oil and gas prospect evaluation.

Journal of Geophysics, 1994, Volume 37,No. 1.