PetroChina has began focusing on such domains as high-steep structure, carbonate rocks, low permeability, and volcanic rocks, and carried out geophysical research based on the thought of "highlighting key basins, relying on key projects, popularizing mature techniques, tackling bottlenecks, and leading technical development" since 2006. Since the "11th Five-Year Plan" period, PetroChina has developed four key techniques, remarkably improving the quality of seismic data and the accuracy of reservoir prediction and hydrocarbon detection. In addition, PetroChina has obtained a series of important geologic achievements and explored a set of effective management methods by applying the key geophysical techniques mentioned above, which vigorously enables it to get a growth spurt for oil and gas reserves.

Prestack depth migration has become an effective approach to solving seismic imaging problems in complex area. However, it is more sensitive to velocity as compared with prestack time migration, which constrains its application. Dabei area is located in the west of Kelasu structure belt of Kuqa Depression, Tarim Basin, where seismic imaging is very difficult because of complicated topography and subsurface structure. This paper introduces the integration of procession and interpretation, and the prestack depth migration velocity modeling method under multi-constraints. Based on the integration and application of surface investigation, tomographyic inversion, well constraints, non-seismic data and other information, velocity field is built, and then anisotropic prestack depth migration is used to image complicated structure. It has successfully solved the imaging problem in the area.

The earth's surface of Kedong tectonic zone is deep loess tableland, the seismic data has an extremely low signal to noise ratio and serious static correction problem. Its complex structure superposition makes the velocity change so violent that it's hard to get accurate imaging. We take full advantage of multiple seismic data processing system and take the study thought of joint processing and interpretation in imaging research.We effectively improve the quality of seismic imaging by taking advantage of wide-line static correction technology,multi-domain prestack noise attenuation technology, common reflection stack technology and fluctuant surface prestack deep migration technology. We further indicate the structural characteristics of the tectonic zone and confirm several favorable traps by fine interpretation research

The southern edge is the most complex area within Junggar Basin. Seismic imaging in high and steep structures has always been a bottleneck of exploration for oil and gas in southern edge. A set of techniques for processing and interpretation suitable for high and steep structures have been developed, based on new and old 3D multiline stack depth migration processing and interpretation of Gumudi area on southern edge of Junggar Basin. The static correction technique, the prestack amplitude preservation and denoising technique, the consistency processing technique, the prestack depth migration processing technique and other technical means have been adopted to improve imaging accuracy in processing, and then all kinds of information have been applied to rational structure modeling and fine velocity modeling in terms of interpretation. Through integrated research, the image of faulted zone has been improved a lot, and the structure of formations has been more clarified. The geological structure of "north-south zonation and vertical stratification" has got more rationalized; main orientation and targets of exploration has been clear.

Lithology is complicated in Permian Fencheng Formation of Fencheng Field in the northwestern margin of Junggar Basin. It is dominated by mud dolomite and has pore-fracture dual media reservoir. To predict fracture strike and density is one of the key parts of oil and gas exploration. The fractured reservoir prediction technique based on P-wave azimuthal anisotropy using prestack seismic data is selected by comparing all kinds of fracture prediction techniques. First, forward modeling based on azimuthal anisotropy is conducted, determining main factors of anisotropy. Secondly, different azimuth seismic data processing is made, and azimuthally anisotropic attributes using prestack seismic data are calculated. Combined with data of the well, fracture strike and density are predicted, and the direction of further oil and gas exploration is pointed out. Drilling results prove that the predicted results are coincident with drilling results, and the technique is feasible in the area

There are rugged surfaces and complex subsurface structures which lead to raw seismic data with poor quality in Baka area of Northern Turpan-Hami Basin. Especially for the main parts of subsurface structures, there are seriously steep formations and abundant faults. So it is difficult for seismic migration and to determine the subsurface structures. And the reservoirs feature in low permeability, low porosity and inconspicuous geophysical response. So there are few reservoir prediction and identification methods. The above factors restrict the progress of oil and gas exploration of Baka area. Technical series have been developed based on research, such as seismic imaging, structural interpretation and reservoir prediction techniques. And the key innovation techniques include model constrained integral static correction method for rugged surface, "six division" prestack noise attenuation method and floating datum pre-stack reverse time migration method. On the basis of combining research and production, the new research results could more accurately describe the subsurface faults and structures, and the overall imaging effect has been improved effectively. Application of new research results helps us determine the structural types again, further realize the faults characteristics and formation distribution rules, find and determine some new traps. On this basis, research reservoir prediction is conducted and several beneficial drilling targets are chosen

After outlining the basic situation of regional geology and exploration of Hongqikan (HQK) structure, this paper clearly points out existing problems in seismic processing and interpretation, which could be described as low S/N ratio and incorrect imaging, directly related to very complicated faults and structure near the mountain area. In other words, the huge errors at main target layers between the seismic and well-logs tell us that how to get correct seismic image has been a key problem in the area for further seismic processing and interpretation. The main reason for the imaging problem is that traditional imaging technology only for simple cases cannot be applied to the complicated situation like HQK structure. RTM(reverse-time migration), the best imaging technology in the world, has not been utilized and reported in China so far. Hence it is believed that RTM is going to be the most key technology for 3D seismic depth processing and interpretation in HQK. Based on those considerations, this paper focuses on three key technologies, i.e. amplitude-preserved noise reduction, non-linear inversion of interval velocity, and pseudo-spectral double-wavefield reverse-time prestack depth migration (PSDW). They have been applied to HQK processing and interpretation project and achieved the results that are positively far beyond initial prediction. In conclusion, RTM could be used to get the greatest seismic images and successfully solve the problems designed at the very beginning. Besides, RTM experience in HQK will definitely benefit depth processing and interpretation in other complicated areas

Ordovician carbonate fracture-vuggy reservoirs are well developed in Halahatang area. The crack-cave distribution is mainly affected by faults and paleo-water systems. The oil layer is thin, and water bodies are active. The fracture-vuggy reservoirs feature in different scale, irregular spatial distribution, and various filling properties. It is difficult to accurately describe the spatial location of carbonate fracture-vuggy reservoir, the volume of fracture-vuggy reservoir, and oil-gas-water distribution in carbonate reservoir exploration and development. With the support of high-density and wideazimuth seismic acquisition, this paper puts forward such predicting methods as the prestack depth migration, the quantitative characterization combining forward and inversion, the prestack anisotropy analysis, and the prestack elastic inversion. Field application of practical seismic data shows that prestack depth migration could accurately characterize the spatial location of fractured body; prestack anisotropy analysis could effectively predict fracture density and direction; the prestack elastic inversion combined with well data could predict the spatial form of fractured reservoir and the oil and gas condition of the reservoir. The application of various means effectively solves the problems in study on carbonate fracture-vuggy reservoirs, and provides powerful supports for well deployment in the carbonate reservoir exploration and development in Halahatang area

3D spatial carving and description are conducted on Lower-Middle Ordovician fracture-cavity carbonate system in Tarim Basin, based on integrating seismic data, underground logging and drilling data. Seismic geometric attributes are extracted by utilizing seismic data to describe such geological properties as karst watercourse, karst cave, plane round or elliptic fracture features and fracture channel related to karst reservoirs. Seismic facies analysis could be used to confirm the spatial distribution of major karst features like karst cave and fracture channel and the threshold value of seismic geometric attributes. The geocellular model is constructed to keep the same resolution with seismic data, which could be helpful to simulate geological properties of karst reservoir accurately. The spatial connectivity of major karst features is analyzed to build the spatial structural model of fracture-cavity connectors. Log facies analysis and calibration with core data and borehole imaging logs lead to the definition of reservoir rock types at wells.Modeling techniques help to characterize reservoir heterogeneity such as reservoir types and porosities derived quantitatively from log information and conditioned with seismic impedance. Fracture-cavity connectors could be incorporated with spatial structure to obtain the reservoir geological model of them. This paper introduces the application of quantitative seismic characterization approaches to the fracture-cavity carbonate reservoirs of the Ordovician. It is shown that quantitative seismic characterization of karst reservoirs provides new understandings of reservoir origins and underlying geological determination.

Reef reservoir is important for high-yield carbonate oil and gas field, and it has large development efficiency and economic value, but there are many uncertain factors influencing reef reservoir study. Firstly, it's difficult to process seismic data of mountain area, and the problems include static correction, prestack noise removing, and migration imaging, etc. Secondly, it is difficult to identify and interpret potential reef reservoir. In the light of these problems, this paper studies reef reservoir of Changxing Formation in the west of Longgang area, Sichuan Basin. Equivalent near-surface model tomographic static correction and global optimization of residual static correction are respectively employed to resolve the long-wavelength and shortwavelength static correction problems. Then frequency-dependent noise attenuation method and conical shaped filter method in the F-Kx-Ky domain are selected to be used to improve the S/N ratio and solve the problem of prestack noise removing. High-resolution wide-band constrained deconvolution, surface consistent deconvolution and single-channel predictive deconvolution techniques are used to improve vertical resolution. Finally, anisotropic ray-bending prestack time migration is used to improve the imaging accuracy. Besides, many techniques for reef reservoir prediction and fluid detection are tested, and techniques suitable for this area are selected, achieving better results. The techniques suitable for reef reservoir prediction and fluid detection of the area are summarized and refined, and they could be taken as guidance for reef reservoir study and oil/gas detection in similar area.

New breakthrough has been made in Changxing Formation of Longgang 62 well and the third member of Xujiahe Formation of Jianmen 102 well recently, which indicates that reef reservoir of Changxing Formation and clastic reservoir of Xujiahe Formation have great exploration potential. Based on comparing and analyzing the dada in work area, PP-PS joint inversion, curvature attributes and instantaneous amplitude are used in the study. The platform-edge is described accurately; clastic rock fracture development and favorable hydrocarbon accumulation zones are forecasted. The study has brought about accurate data for further exploration, and a set of technical countermeasures are developed.

To further expand the exploration results in shallow layers of Jiulong Mountain, a series of technology research about structure interpretation, reservoir prediction, fluid detection and fracture forecast are conducted by using 3-D seismic data. Some achievements have been made. Firstly, favorable sedimentary facies distribution is forecasted through paleogeomorphology restoration and seismic attribute analysis. Secondly, thickness, porosity of reservoir is forecasted through pseudo-acoustic inversion, wave impedance inversion, and gamma inversion. Zhenzhuchong reservoirs were mainly developed in the middle and southeast of the study area; the thick area of the upper of the second member of Xujiahe Formation is mainly distributed in the northwest of the study area, of which the distribution law coincides with the geological reality. Thirdly, gas potential and fractures are forecasted with poststack and prestack seismic data. Field application shows that major gas zones and fracture development areas as well as reservoir distribution areas of all target layers own better overlapping features.

Drilling and exploration show that favorable reef-pore and karst-cave reservoirs developed in Ordovician of the platform margin belt in northern Tianhuan area, Ordos Basin. However, it is faced with such difficulties as complicated reservoir-forming conditions, poor seismic inside imaging, and strong reservoir heterogeneity. In order to open up new areas of exploration, this paper focuses on static correction, multi-domain denoising and prestack migration and other processing techniques to improve the Ordovician inside imaging accuracy. Seismic reflection pattern classification and attribute analysis techniques are enhanced, and the direct gas detection method about prestack carbonate reservoirs prediction is explored and applied, based on the study on favorable sedimentary facies belt. 1218 square kilometers of favorable reservoir development areas are evaluated comprehensively. New discovery is made in Kelimoli Formation of one exploration well confirms broad exploration prospects in the area once again

The eighth member of Shihezi Formation is a major reservoir of Sulige gas field in Ordos Basin. It is a thin reservoir of low porosity and permeability, so it is hard to be predicted based on regular seismic prediction. Multi-wave exploration technique could be used to improve accuracy of rock character and fluid property by adding with S-wave data. This paper focuses on multi-wave application in Sulige gas field, mainly discusses converted wave static correction, converted wave imaging, fine PS wave calibration, P and S joint interpretation, joint inversion, and thus better solves difficulties in converted imaging. Meanwhile, the success ratio of drilling is improved greatly by using the fluid detection methods based on multi-wave AVO and multi-wave joint inversion. This paper presents a set of reservoir prediction methods of multi-wave fluid detection focusing on western area of Sulige, which are proved to be feasible in the area

Sulige gas field, a very rare low permeability, low pressure, and low abundance gas field, is composed of highly heterogeneous tight gas reservoirs. It is an exploration and development model of low permeability reservoirs in China. Because of the small acoustic impedance difference between mudstones and gas-bearing sandstones, and the influence of seismic resolution, the distribution of effective reservoirs in Sulige area cannot be clearly characterized using conventional post-stack inversion method. Targeting at the problems encountered in the exploration and development of low permeability sandstone reservoirs in Sulige gas field, a technical process and procedure relating to rock physics, multi-component seismic data processing, inversion, interpretation, and geological analysis is set up systematically, which contains well-controlled seismic data processing, fine event match of PP and PS, simultaneous inversion of PP and PS, and some other corresponding techniques. Different from single P-wave, multi-component exploration uses the introduction of converted wave seismic data as a constraint condition, to improve the stability of inversion, and signal to noise ratio of its results, thus improving accuracy of reservoir prediction. Its application in Sulige area has proved the effectiveness and feasibility of the techniques, and accumulated valuable experiences for exploration and development of similar oil/gas fields in future.

There is huge gas resource potential in the continental facies clastic reservoir of Xujiahe Formation in Upper Triassic of Sichuan Basin. In recent years, commercial gas flow has been got in central Sichuan, such as Hechuan and Guang'an etc. In the study area, Xujiahe gas reservoir is mainly lithologic gas reservoir and compound traps of lithologic-structural gas reservoir. The reservoirs are very heterogeneous horizontally and vertically, and most of them are thin interbedding with low porosity and low permeability. Moreover, the reservoir fluid contents are very complex, so it is very difficult to perform reservoir prediction and fluid detection. Aimed at this type of lithologic gas reservoir, prestack high-fidelity processing technique has been applied to obtain prestack gathers with high quality, then, prestack quantitative reservoir prediction technique, mainly including elastic parameter inversion, as well as hydrocarbon detection technique which integrates elastic multi-parameter method and expanded elastic impedance inversion have been applied to improve the accuracy of reservoir prediction and fluid detection, and good application effects have been achieved

2D seismic project in Langzhong-Guang'an area has been operated over years and crossed vast terrain with strong ground roll. The target is thin layers deposited on flow phase with low porosity and low permeability, changing a lot laterally. Dependent on reliable enclosure checking procedures to timing and energy distribution in the seismic data volume, high-quality gathering and final results are provided for structure interpretation and reservoir prediction by laying emphasis on preserving preservation of amplitude and fidelity. Those elastic parameters relatedwith lithology and hydrocarbon potential, i.e. ratio of P/S wave and Poisson's ratio, are obtained after prestack inversion. Prediction of reservoir and hydrocarbon potential is carried out through statistics and cross-plot analysis of reservoir sensitive parameters, and remarkable geological results are obtained.

The natural gas reservoir of Xujiahe Formation and Leikoupo Formation in Jianyang-Dazu field of middle Sichuan Basin belong to tight sandstone gas reservoir and carbonate gas reservoir respectively. The variation of geophysical characteristics of tight sandstone gas reservoir is relatively small, due to low porosity and change of hydrocarbon saturation. So the main challenge for seismic prediction is to predict reservoir and gas strata. Diagenesis of carbonate rocks is so complex that the connection among velocity, density, and porosity is bad, thus it is difficult to predict physical properties of the target reservoir quantitatively. Through prestack seismic data processing and reservoir prediction in Jianyang-Dazu field, following results have been obtained. Firstly, primary factors affecting amplitude-preserved processing of seismic data are analyzed, key techniques for prestack processing are developed, and an effective flow is finally formed. Secondly, petrophysical analysis of typical tight sandstone indicates that elastic impedance (EI) and P-wave-to-S-wave velocity ratio could differentiate lithology effectively; the correlation between P-wave and S-wave as well as porosity is so good that it could be used to predict physical properties. Therefore physical properties are differentiated later than lithology. Thirdly, elastic-impedance coefficient (EC) is developed to predict gas strata in tight sandstone, and new technique to quantitatively predict carbonate reservoir based on seismic techniques constrained by typical reservoir characterization parameters is also presented.

Gaotaizi area has been a major target of Changyuan tectonic zone to submit reserves in recent years. Fuyu oil layer was dominated by terrigenous fluvial faces and delta at sedimentary stage. It has thin sandstone (less than 5m in general) and small sand bodies (300m~600m), which is far lower than the seismic limiting resolution. So it is difficult to identify river sand bodies. To solve the problem, taking accurate seismic identification of river sand bodies in the oil layer as the objective, a technique and procedure for seismic reservoir predicting is developed, which is based on high resolution of relative preservation of amplitude and waveform features and employs such means as the optimal stack identification of sand bodies, the prestack AVO prediction of effective reservoirs, and the inspection of oil layer by prestack elastic parameter inversion. It has been applied well in work area, and meandering river point bar sand bodies are described clearly, laying solid basis for the discovery of thin-bed lithologic oil and gas reservoirs in the area. There are 18 recommendation wells in total, all of which have obtained good oil and gas shows. 10 wells have been drilled and completed, 9 have been tested, 7 have got commercial oil, and 2 have produced less oil. Wildcat success rate reaches 77.8%, 1.78 times higher than the 43.9% before the research is conducted.

The effectiveness of thin interbed prediction is the main bottleneck restricting the growth of oil production in Songliao Basin. It is confronted with such problems as thin layer, poor property, obscure reservoir disciplines and low predication accuracy of effective reservoir. Study has been conducted on seismic amplitude-preserved processing, resolution enhancement and effective prediction with pre-stack seismic data, and a series of technical achievements have been obtained. A suit of techniques are used to preserve the amplitude in data processing, such as the two-step deconvolution in shot and receiver domain, the anisotropic pre-stack time migration, and the gather regularization, etc. In terms of resolution enhancement, the techniques including the reduction of T2 effect, the enhancement of resolution in pre-stack gathers, the alignment of wavelet stretching in far-middle offset, and the optimization of effective offset distance are used to expand frequency band and reduce interference of thin interbed. In the aspect of prediction with prestack seismic data, the techniques including AVO qualitative reservoir description, quantitative porosity prediction, and pre-stack elastic impedance inversion are used to delineate the thickness of effective reservoir. Through the application of the techniques above, the matching probability of reservoir prediction is higher than 78.5% in Saertu survey

There exits obvious limitations by using P-wave data for reservoir characterization alone in Sanhu area, Qaidam Basin, due to energy absorbing and structure distortion caused by underground gas. Affected by many factors, using P-wave data alone may lead to multi-solution problem. This problem could be solved by using multi-wave seismic exploration techniques and converted-wave data, and the accuracy of reservoir forecast and oil-gas detection could be improved a lot. Taking seismic exploration in Sanhu area based on multi-wave data as an example, the converted-wave data processing technique and the interpretation technique combining multiple waves have been improved. Based on making breakthroughs in converted-wave static correction and prestack time migration, and using high-quality multi-wave data, the interpretation technique combining multiple waves has been improved; joint analysis of multi-wave seismic attributes and oil-gas detection have been conducted to confirm lithologic gas reservoirs. Good geological results have been achieved in practical drilling and exploration.