低频地震探测方式

低频地震探测方式(LFS)是基于自然背景地震光谱在低频(1至10赫兹)下的分析。

LFS方式的原理是基于流体饱和、断裂以及多孔的介质操作而成的。在一个均质的地质环境内,微震的特性显示了地震能量的单调分布。地质的不均质物体,例如结晶基底、断层构造、碳氢化合物矿床和盐层的形成,引发在光谱里地震能量的重新分布(发生在能够清晰分辨的光谱峰值)。

由各种地质不均质产生的地震能量光谱峰值具能够辨别的特征。因此,石油饱和的储层发出强烈反常,以便我们预计在油藏的流体饱和度。

针对勘探区进行的地震能量峰值分布的分析能够让我们识别不同地质构造的区域。

Low-Frequency Seismic Sounding investigations include field surveys, data processing and interpretation.

Modifications of field surveys:

  • line (250 m, 500 m and more)
  • grid (250x250 m, 500x500 m and more)

Processing is carried out using our original software involving calculation procedures that enable to effectively extract the signal against anthropogenic and natural noises. The procedures include both classic signal processing algorithms based on Fourier analysis and correlation techniques, as well as methods based on wavelet analysis.



Stages of Low-Frequency Seismic Sounding Method

Stages of Low-Frequency Seismic Sounding Method


Numerical simulations enable us to forecast behaviour of microseismic field for certain geological conditions and to compare simulated and real spectral curves, increasing the reliability of data interpretation.

LFS survey allows to identify and stratify hydrocarbon deposits in geological section. The LFS signal is calibrated using measurement on nearest wells with known fluid saturation.

The main result of LFS survey is a map of oil & gas potential zones for the surveyed area.


Physical Theory

Low-Frequency Seismic Sounding method is based on the effect of increased low-frequency energy (1-10Hz) in the spectrum of microseisms over the hydrocarbon pools, which was discovered in 1989. During 2002-2009, the effect was studied on over 100 areas in course of the research studies and exploration operations onshore. Testing of the technology based on such a vast amount of data in different geological conditions has confirmed the theoretical understanding of the nature of anomaly associated with multiple reflections (resonance) of seismic waves between the surface and the oil and gas deposit (Fig. 1).


Impedance frequency dependence for oil and water

Fig. 2. Impedance frequency dependence for oil and water


An abnormally high rate of wave reflection from the oil-saturated reservoir can be explained based on the ratio of the basic physical properties of water and oil, as well as taking into account that exactly the low-frequency component of the waves is being significantly reflected. The compressibility of oil is 2-5 times higher than the compressibility of the water and surrounding rocks. The compressibility of water differs for only 10-20% as compared with the compressibility of surrounding rocks. The compressibility is affected by the gas factor which considerably increases it. The viscosity of oil is two orders of magnitude greater than that of the water and surrounding rocks. Under the low-frequency influence, the forces of viscosity within the deposit weaken and, as a result, the deposit is becoming easily deformable by the infrasonic wave. This has an impact on decreasing of its acoustic impedance and, as a result, promotes significant reflection coefficient (Fig. 2). 

Thus, the reflection coefficient of the oil and gas deposits depends on the frequency of seismic waves and increases with its decrease. The intensity of anomaly depends on the thickness of the oil reservoir and its physical properties - viscosity, compressibility and oil saturation.


Data acquisition

We have more than 60 kits of advanced field equipment adapted to different operating conditions onshore.


Experience of surveys at temperatures from +40 to -60 deg C. Field crews are equipped with satellite communications for operational data transmission at inaccessible terrain.


Processing

Block diagram of the LFS processing sequence consists of the following stages:

  • Preparatory stage;
  • Narrowband noise filtering;
  • Transient noise filtering;
  • Generation of the final Fourier spectra;
  • Spectral anomaly parameters estimation;
  • Correction for spectral anomaly parameter variations.


The preparatory stage includes assessment of the filed data quality by means of visual analysis of the spectra and dynamic spectrograms of seismic signals. In case of poor data quality, field crew shall carry out reacquisition of the rejected receiver point. Distinguishing features of the desired signal and noise enable to identify the frequency ranges where the desired signal events are unaffected by significant noises. This stage also includes adjustment of automatic filtering techniques for individual features of the microseismic signals in the current survey area.



Narrowband noise filtering is done using the optimization method of quasiharmonic noise filtering with preservation of the background noise level in order to exclude the narrowband components caused by the monochromatic vibrations of the near-surface industrial sources;


Transient noise filtering
is carried out in order to exclude the snapshot waveforms of the increased energy caused by the influence of the additive surface noise, which significantly distorts the spectrum of the desired background microseismic noise.


Generation of the final Fourier spectra.




Spectral anomaly parameters estimation(calculation of spectral anomaly’s frequency, anomaly bandwidth, local signal-to-noise ratio) is carried out in an automatic mode using the algorithm of parameterization of the spectral peaks, which was developed based on the optimization of local peaks of the wavelet image of microseismic signal spectrum. In case of the complex spectrum (overlapping of the spectral peaks of several anomalous zones of geologic section), estimation of anomalies’ intensity is carried out “manually”.



Correction for spectral anomaly parameter variations. Repeated studies of the microseismic wavefield spectra have shown that there are variations of the anomaly parameters which are evident while comparing cumulative spectra recorded, for example, during the day and night. The technology enabling to take into account the changes of anomaly parameters given the limited number of simultaneously observed points is used under the conditions of temporal non-stability and spatial heterogeneity of the microseismic wavefield. To speed-up this stage, the algorithm for automatic calculation of correction coefficients was developed, which performs calculations based on the preliminarily prepared spread-sheet of measurements metadata.


Interpretation


Interpretation of low-frequency seismic sensing includes:

  • Analysis of anomaly’s stability during the day;
  • Typification of spectra;
  • Cluster analysis;
  • Mapping.
  • Calculation of predicted frequencies;
  • Comparative Analysis;
  • Numerical modeling;


Analysis of anomaly’s stability during the day. Stability of the anomaly during the day is studied here using the method of recording the temporal variations of the anomaly’s “signal-to-noise” parameter.


Typification of spectra.

Cluster analysis of the parameters of spectral anomalies (frequency, width and "quality" of an anomaly), which enables to:


  1. visually assess the distribution of parameter values that form the cluster;
  2. separate stable spectral anomalies over the survey area from random ones;
  3. automate the generation of the physical fields of the values of anomalies’ parameters;
  4. numerically compare the parameters of anomalies recorded on different territories.


Calculation of predicted frequencies is carried out in accordance with the results of cluster analysis based on the well-velocity survey and VSP data on the distribution of longitudinal velocities in the section. As a result of comparison of the actual anomaly frequency with the set of predicted frequencies attributed to stratigraphic horizons, possible interval of oil pool occurrence in the section is determined.

Comparative analysis of the acquired spectra with the spectra from the territories with proven oil bearing capacity as per deep drilling data;

Numerical modeling of the seismic waves’ distribution in the geological environment is carried out in order to calculate the synthetic spectra of the microseismic vibrations assuming both absence of the oil saturation in the section and the presence of oil saturation in one or more target horizons. Acquired synthetic seismograms are compared with the field data which allows determining the most probable type of hydrocarbon accumulations in the section.


Mapping and comparative analysis of the spatial distribution of anomalies’ parameters for evaluation of oil potential of the objects in question.


Software Package

Processing of low-frequency passive seismic data is characterized by the large volumes of the raw data as well as by the need to subject such data to multi-variant resource-intensive calculations. For this purpose we apply an in-house software package based on the most advanced software and hardware solutions.



The software package is built on three-level architecture. All resource-intensive calculations are installed in the application server implemented on the basis of J2EE technology. The application server actively uses distributed calculations ensuring high scalability of the software package.


In order to ensure flexibility, the PC system manages the data processing according to a user-defined processing sequence. The processing sequence is compiled by the user from a great number of standard processing functions and may be extended by the user-defined Java and Matlab modules. Processing sequence may include branchpoints and savepoints for intermediate results which makes it possible to minimize repeated calculations.

Raw, intermediate and resulting data are stored in the relational database. The relational database has a flexible structure which may be expanded by the user. Along with the low-frequency seismic data, the system may also use seismic data, well logging data and other factual and cartographic geological-geophysical information.


All available information may be graphically visualized at any stage of processing. Visualization includes plots, maps and 3D models. The contents and the form of data presentation may be dynamically changed by the user in course of data processing.



Numerical simulations

In the process of waves propagation in geological environment they are influenced not only by the hydrocarbon deposits, but many other contrasting geological objects as well, which resulting impact is similar to that of the oil and gas pools. We use numerical modeling to study the behavior of microseisms in the complex geological environments.





Given the available seismic survey data on the area, seismic-mechanical model of the geologic section is compiled which is then subjected to the numerical experiments on the microseismic waves distribution. Synthetic seismic gathers are calculated assuming both absence of the oil saturation in the section and the presence of oil saturation in one or more target horizons.


Acquired synthetic gathers are compared to field data using different statistical techniques which allows determining the most likely option of availability and location of hydrocarbon accumulations.


Numerical modeling is performed on a high-performance computer based on the massively parallel TESLA processor coupled with CUDA 2.1 programming environment from nVidia. Application of these solutions allows calculating multi-variant models of the geological environment in the resolutions of up to 1m in 2D version and up to 10m in 3D. Calculations are carried out 30-200 times faster as compared to the single-core 3GHZ Intel processor.


Efficiency of technology

If you are not satisfied with standard geophysical methods, then you need something more!

Up to date over 126 projects are completed with total researched area of 1133.5 sq. km.

The efficiency rate of the LFS technology amounted to 85.3%

95 wells were drilled by the Low-Frequency Seismic Sounding interpretation results. 81 wells have confirmed the forecasting. 28 wells produced hydrocarbons from the carbonate reservoirs, and 35 wells produced hydrocarbons from the terrigenous reservoirs.


The efficiency of Low-Frequency Seismic Sounding (2005-2010)