In Lunnan area of Tarim Basin, using velocity spectrum data, the interval velocity is calculated by model iteration method, and its vertical and horizontal variation laws are studied. The research shows that the three-stage structure of interval velocity in Lunnan area can reflect the longitudinal change of large lithologic combination; Using the velocity of remaining layers, the physical properties of reservoirs and their favorable development areas can be further determined. This paper discusses the method of time-depth conversion by using seismic velocity spectrum, which lays the foundation for structural analysis and reservoir description in this area.

Velocity layer, velocity geological law, time-depth conversion between velocity and physical properties

I. Introduction

Seismic velocity is not only one of the most important parameters in seismic data, but also an important and commonly used information in structure, reservoir research and reservoir description technology, and its cognitive degree directly affects all aspects of oil and gas exploration. In Tarim basin, seismic velocity research has always been an important content in the exploration stage, and a lot of investment has been made in wave impedance inversion and PIVT profile making, but the cost is too high to be widely popularized in this area. Comparatively speaking, velocity spectrum data has the advantages of wide plane distribution and low investment. Under the condition of good data quality, the velocity spectrum data can also be used to qualitatively or semi-quantitatively explain the changes of underground lithology and physical properties.

Figure 1 layer velocity scatter map in Lunnan area

Lunnan area is rich in speed data and good in quality. Compared with VSP velocity data of seismic logging, it has good consistency (Figure 1) and high utilization value. This paper mainly introduces the geological law reflected by it and the application of velocity in lithology, physical property analysis and time-depth conversion.

Second, the calculation of interval velocity

In order to ensure the reliability of interval velocity calculation, mature model iteration method is adopted at present. The basic ideas are as follows: (1) Establish an initial seismogeological model by using the T0 result of seismic interpretation; (2) Assuming that the first layer is a homogeneous medium, calculating the layer velocity and the position of the first reflection interface; (3) firstly, the dip angle of the second layer is corrected, and the layer velocity is calculated by DIX formula as the initial layer velocity of this layer; ④ Calculate the time-distance curve of interface reflection wave on CMP gathers according to the actual observation system, and calculate the stacking velocity by theoretical curve fitting; ⑤ Compare the calculated stacking velocity with the actual stacking velocity, and if the error reaches a certain range, the interval velocity will be identified; Otherwise, adjust the layer velocity and return to the fourth step for recalculation; 6. And so on, step by step.

Three. Velocity characteristics and geological laws

The zonal velocity in Lunnan area not only has obvious segmentation characteristics (Figure 1), but also has the characteristics of vertical "three-segment structure", that is, there are three velocity segments and two main velocity interfaces, which correspond to the unconformity surface at the bottom of Triassic and the unconformity surface at the top of Mesozoic respectively.

Fig. 2 Vertical variation diagram of formation velocity in Lunnan 1 well.

As can be seen from Figure 2, the change of velocity does not increase with the increase of depth, which does not conform to the general geological laws. But in each "trend profile", the formation velocity follows its general law, that is, it increases with the increase of depth. This shows that the sedimentary environment is relatively stable (or constantly changing) in the geological age corresponding to each megatrend segment, but the velocity change caused by sedimentary rate in different stages is not the same, which shows that their lithologic combination and physical property change are different in different development stages.

The propagation speed of seismic wave in rock layer depends on the elastic modulus and density of rock, and the elastic modulus of rock depends on the mineral composition of rock. The propagation speed of seismic wave in rock is also related to external factors such as porosity, pore fluid properties, pressure and temperature. Although Cenozoic and Mesozoic in Lunnan area are dominated by sandstone and mudstone deposits, there are great differences in porosity, cement and pore fluid. Among them, a set of high-speed gypsum-bearing mudstone (widely distributed in Tarim basin) was deposited at the bottom of Cenozoic. Due to the intervention of gypsum components, the elastic modulus of rocks in Paleogene sandstone mudstone strata changes greatly, and the propagation speed of seismic waves increases greatly. The underlying Cretaceous sediments are mainly high porosity sandstone, with loose cementation and obviously reduced velocity. The velocity curve reverses near the bottom of Cenozoic, forming a velocity "reversal step"-the velocity interface at the top of Mesozoic. The velocity of Jurassic strata increased slightly, showing a gradual transition trend with Upper Triassic. Due to the fluctuation of ancient denudation surface and the influence of different provenances, the lithology and physical properties of the Lower Triassic have changed greatly, resulting in a great lateral change in velocity. However, the whole Mesozoic is characterized by stable changes, and the speed of overlying and underlying strata is relatively low. The Paleozoic lithology is quite different, and Silurian, Devonian and Permian are missing in Lunnan area. Carboniferous consists of calcareous sandstone and calcareous mudstone, high-speed limestone and constant-speed salt rock with higher velocity than normal clastic rock. The Ordovician is dominated by carbonate rocks, and the lithologic combination speed of this set is obviously accelerated, with a "positive step". Therefore, two velocity steps, one positive and one negative, determine the "three-stage structure" of interval velocity in Lunnan area.

Four. Velocity, lithology and physical properties

When the interval velocity reaches a certain accuracy requirement, it can be directly used to study geological problems [1]. The Mesozoic sedimentary environment in Lunnan area is relatively stable, and the vertical and horizontal changes of formation velocity are relatively small, and the velocity change range is not large. The velocity difference between sandstone and mudstone is between 200-400 m/s, which provides a certain basis for studying the lithology and reservoir properties of important target intervals.

1. Velocity-lithology relationship

Through the analysis and comparison of the velocity of sandstone and mudstone in Jing Shangshun, the velocity law of sandstone and mudstone in Lunnan area is different from general lithology. The velocity of shallow mudstone is higher than that of sandstone, and the intersection of them is located at the corresponding depth in Jurassic (Figure 3), which means that it is impossible to make a good lithologic analysis of Jurassic and above strata by using velocity data. Although Triassic can be distinguished, according to the general law, the basis for calculating the percentage of sandstone and mudstone in the target layer by velocity is that the velocity difference between pure sandstone and mudstone must reach 20% ~ 30% [2], while the velocity difference between Triassic sandstone and mudstone in Lunnan area is not large. Therefore, we should be careful when using velocity to analyze lithology. Although predecessors have done a lot of work in this field, the author does not approve of using velocity to predict lithology.

2. Relationship between velocity and reservoir physical properties

Fig. 3 Velocity curve of sandstone and mudstone in Well Lunnan 26

According to logging data and seismic interpretation results, the Triassic standard layer of uranium-bearing mudstone in the main target layer is relatively flat, the buried depth of the stratum increases by more than 500 m from south to north, and the dip angle of the stratum is only 2; Even in the structural position, the dip angle of strata increases slightly, only 5, so the influence of buried depth on velocity may not be great. Triassic sediments in Lunnan area are relatively stable, mainly composed of sand and mudstone, with sandstone content of 20% ~ 60% and sandstone porosity of about 20%. Therefore, the influence of rock composition, porosity and buried depth on velocity will not be too great. For a certain sand group (or oil group), the factor affecting the velocity is the fluid composition. Take Well Lunnan 58 as an example, there is a set of sandstone reservoir in the lower part of the Triassic Gongyou Formation (the velocity is 3200m/s), which is lower than that of argillaceous surrounding rock reservoir (the velocity is 3380m/s) [3], because the reservoir contains gas. Therefore, it can be considered to remove the regional velocity background and keep its anomalies to detect the physical differences in this area.

Taking Triassic Ⅰ sand group in Lunnan area as an example, the trend surface is obtained by using the method of residual layer velocity trend analysis, and then the residual layer velocity is obtained by using the following formula:

△Vi=Vi-fi(V)

Where: vi-velocity of residual layer, m/s;

Vi—— known interval velocity value, which is V(x, y) and m/s on the plane;

Fi(v)- trend value, that is, f(x, y) and m/s on the plane.

As can be seen from fig. 4a, Lunnan fault zone is a low-speed anomaly zone as a whole. Using the velocity plan of the remaining layers, combined with the sandstone percentage map and oil and gas detection results, the reservoir evaluation map of Lunnan area can be drawn, and the first and second favorable oil and gas-bearing areas can be obtained (Figure 4b).

This method is effective in Lunnan area. Due to the improvement of reservoir performance and the enrichment of oil and gas in Lunnan fault barrier zone, this zone presents obvious negative velocity anomaly, which is confirmed by the exploration and development of Lunnan 2, Lunnan 3, Lunnan 5 and Lunnan 10 fault block oilfields. In the sub-favorable evaluation area, several exploration wells have also achieved certain exploration results.

Fig. 4 Evaluation diagram of remaining layer velocity and reservoir physical properties of Triassic Ⅰ sand group in Lunnan area.

It is worth noting that the velocity converted from velocity spectrum data is limited by its accuracy, which qualitatively but not quantitatively reflects the changing trend of parameters within a certain accuracy range.

Verb (abbreviation of verb) speed and time-depth conversion

One of the main purposes of velocity research is time-depth conversion, drawing more accurate structural maps and providing necessary data for final well location establishment and drilling design. The conventional velocity-time-depth conversion method is mainly scale method, and its calculation process is stacking velocity-root mean square velocity-interval velocity-average velocity-fitting scale-converting time-domain data into depth domain.

According to the analysis results of average velocity, the average velocity in Lunnan area shows a variation law of low in the southeast and high in the northwest, with a difference of about 300 m/s. If the time-depth conversion is directly carried out by using the fitting velocity scale, the error will be great, and even false structures will be caused, which can not meet the needs of high accuracy in compiling structural maps in this area. Therefore, the author slightly modified the gauge method, and directly converted the time to depth with the average speed instead of fitting the average speed, which is called the conversion from variable speed time to depth (Figure 5).

Fig. 5 Diagram of Time-depth Conversion Mode of Gauge and Variable Speed

The time-depth conversion method is to cut the velocity field along the seismic interpretation horizon to obtain the average velocity profile, and directly convert the time-depth from the velocity and horizon time, that is:

Essay on exploration and development in Shengli oil region

Where: h (x, y)- the depth of coordinate (x, y), m;

T0(x, y)—— When the seismic horizon travels in two directions at coordinate (x, y), s;

V (x, y, t0)/2 —— the average speed at t0 at (x, y), m/s.

As can be seen from the above formula, the accuracy of depth map strictly depends on the accuracy of velocity field. Through the comparison between VSP velocity and seismic velocity in well 14, it is found that the error distribution between VSP velocity and velocity obtained by velocity spectrum is uneven, which is not a systematic error. Therefore, after the velocity database is established by using seismic velocity, it must be constrained by more accurate VSP data. The constraint of velocity field is mainly completed by the following steps: ① Using the velocity in velocity data volume as a gridding method, the estimated value of well point velocity is obtained, and the difference between the estimated value and the measured velocity in the well is obtained; (2) Obtaining the difference values of all well points; (3) adjusting the speed of the speed data volume by using the difference; (4) Go back to the first step and perform iterative calculation until the difference between all well points is small enough.

The constrained velocity field is used for time-depth conversion, and the converted buried depth map is close to the actual drilling depth (table 1), which can meet the requirements of reservoir description and achieve the purpose of variable speed mapping.

Table 1 Error Statistics of Isobathymetric Chart of Time-depth Conversion

The faults in Lunnan slope zone are not developed, so it is difficult to form large structural traps. However, according to previous exploration experience, tectonic platform (or slope break zone) is often the focus of exploration. Therefore, the main target strata in Lunnan area are explained in detail. Firstly, the density of interpretation network is increased to 50m×50m, which ensures the accuracy of seismic interpretation. Then the variable-speed time-depth conversion is unified, and the "platform" with structure display in the structural map is selected from the interpretation data, and large-scale accurate mapping is carried out independently, which overcomes the possibility of erasing local structure due to smooth grid during regional mapping.

Through comprehensive analysis, three large traps and more than ten local small structures were found in the bottom structural layer of uranium-bearing mudstone on the south slope of Lunnan Oilfield Development Zone, which provided a basis for further exploration in Lunnan Oilfield.

Six, some understanding

Under certain seismic geological conditions, seismic wave velocity can reflect a set of large lithologic combination changes, especially in Lunnan area, the sudden change of velocity trend corresponds to a large stratigraphic unconformity interface.

Under certain geological conditions, favorable exploration areas can be determined by using velocity data.

Time-depth conversion is not suitable for areas with large changes in plane velocity, and variable-speed time-depth conversion can accurately reflect the true face of the structure.

I would like to thank the chief geologist Song and the deputy chief geologist for their guidance in the research process, and express my heartfelt thanks here.

Main references

Liu. Using seismic information to predict oil and gas. Beijing: Petroleum Industry Press, 1994.

Zhu Guangsheng. Reservoir prediction method based on seismic data. Beijing: Petroleum Industry Press, 1995.

Chen Yongwu. Reservoir and oil and gas distribution (oil and gas exploration series in Tarim Basin). Beijing: Petroleum Industry Press, 1995.