Application of Forward Modeling in Putaohua Oilfield

: Putaohua Oilfield is bounded by the large fault on the south side of Pu A1 Well. It is divided into Pubei and Punan development zones on the plane. It is an oilfield that has been developed by water injection for more than 30 years. With the increase of development time, the water content of oil wells increases rapidly, and the oil production of the oilfield decreases rapidly,In order to realize the study of petroleum geological potential in this area, find out the favorable target area for well layout, and effectively improve the drilling success rate,first of all, we need to do forward modeling for the study area to lay a solid foundation for subsequent research. This paper analyzes the change characteristics of seismic wave frequency attenuation attribute in reservoir oil and gas, and applies it to the study of reservoir oil and gas prediction in Putaohua Oilfield.


Introduction
In geophysical exploration research, according to the shape, occurrence and physical property data of the geological body, the theoretical value (mathematical simulation) is calculated through the structural mathematical model, or the numerical value (physical simulation) of the geophysical effect generated by the observation model through the structural physical model is called forward simulation [1] [2] . In the process of geophysical data interpretation, the forward simulation results are often compared with the actual geophysical exploration data, and the model is constantly revised to make the simulation results as close as possible to the actual data, so as to make the interpretation results closer to the objective reality. This process of comparison is also called selection method [3] [4] .
For example, before the processing and analysis of Pubei 3D seismic attributes, it is necessary to use forward modeling experiments to simulate the seismic response results of Puyi group of sand and mud thin interbedding and channel sand body deposition, and to study the reservoir identification ability of a certain seismic frequency.
Forward simulation includes physical simulation and computational simulation. The former has the advantage of being closer to the actual situation, while the latter has high efficiency, convenient calculation and modification of parameters, and is more widely used [5] [6] . This paper analyzes the change characteristics of seismic wave frequency attenuation attribute in reservoir oil and gas, and applies it to the study of reservoir oil and gas prediction in Putaohua Oilfield. Based on the data analysis of the seismic response results and forward and inverse results of the two-dimensional geological model, it is considered that the seismic profile can not fully reflect the characteristics of the thin sand body of the reservoir, and the effect is better for the relatively thick sand body in the middle of the Putaohua Formation; For the thin sand layer near the top and bottom of the grape flower layer, only the amplitude energy is reflected.

One-dimensional Computational Simulation
Basic Assumptions of One-dimensional Computational Simulation [7] .
One-dimensional simulation includes the following basic assumptions (commonly referred to as Gupitt model): (1) The stratum is horizontally uniform and vertically composed of a series of parallel thin layers with different elastic properties; (2) The thin layer is assumed to be isochronous, and the time thickness of each layer is half  (sampling interval); (3) The seismic wave is plane wave normal incidence; (4) The reflection wavelet shape of each interface is the same, only the amplitude and symbol change according to the reflection coefficient of each interface; (5) Ignore transmission effects and multiple reflections.

Basic Formula of One-Dimensional Numerical Simulation
Under the above assumptions, reflection seismic records ) (t x can be regarded as the convolution of seismic wavelet and reflection function ) (t R (reflection coefficient sequence in discrete time) [8] . * (1) This is the basic formula for one-dimensional calculation and simulation. If the theoretical seismic record can be calculated according to the above formula, which is called synthetic seismic record. Therefore, one-dimensional simulation is also called the production of synthetic seismic records.

Calculation In More Complex Situations
By extending the meaning of seismic wavelet or reflection coefficient sequence in the above basic formula, it can be used for calculation in more complex situations to obtain more practical synthetic seismic records [9] .
(1) Consider transmission effect According to the Gupitt model, the upper incident reflection coefficient of a reflection interface is i R , while the lower incident reflection coefficient is 0, and the upper and lower incident transmission coefficients are 1, which is obviously unreasonable. If the transmission coefficient is not considered to be 1, the transmission effect is considered. At this time, it is only necessary to expand the reflection coefficient sequence to be * . * ∏ 1 The synthetic record calculation formula considering transmission effect is * * * (2) Consider absorption effect Considering the influence of absorption, convolution formula can still be used to calculate. Or include the absorption into the reflection coefficient sequence, calculate the pseudo-reflection coefficient sequence including the absorption from the known absorption coefficient and reflection coefficient sequence, and calculate the synthetic seismic record including the absorption effect with wavelet convolution. Or the effect of absorption can be included in the wavelet term to form the convolution of time-varying wavelet and reflection coefficient sequence to obtain synthetic seismic records.
As absorption is usually related to frequency, a convenient method is to calculate in the frequency domain. Especially when the effects of transmission, multiple reflection and absorption are considered at the same time, the calculation in frequency domain is more simple. The calculation of time domain is much more complicated, and some approximation is often taken.
Although theoretically one-dimensional simulation is only applicable to horizontal layered media, it is sometimes used in more complex situations. In order to carry out forward simulation calculation, the seismic model must be established first. In the work area, the wavelet of seismic data in this area is finally estimated by making synthetic seismic records of multiple evaluation wells. The seismic forward response of the Puyi sandstone section is simply explained by one-dimensional synthetic records and two-dimensional forward results; In Fig.1 is the synthetic seismic record of Pu C well. According to the acoustic curve and density curve, the wave impedance curve is obtained, and the synthetic record of the well is calculated to test the underground formation information reflected by the reflection event of the actual seismic record, or to determine the seismic response characteristics of the sand body. From the perspective of Pu C well, due to the thin sand body thickness and the small difference between the upper and lower lithological impedance in the Putaohua oil layer, its seismic reflection is not obvious, but there is good reflection at the strong impedance interface at the top and bottom of the Putaohua oil layer. Through its analysis, it can be seen that the top surface of T1-1 sandstone corresponds to the wave crest, the bottom surface of T1-1 sandstone corresponds to the wave trough, and the sandstone in T1-1 has weak seismic reflection. Fig. 1 The synthetic seismic record of Pu C well

2D Geological Model and Seismic Response
According to the understanding of the grape flower sediment sand body, it is necessary to design a two-dimensional model [10] . The results of two-dimensional model forward modeling should achieve the following purposes: (1) Analyze the ability of seismic profile to identify the thickness of sandstone vertically, and determine the geological meaning of the weak reflection layer between T1-1 reflector and T1-1b reflector by analyzing the reflection characteristics of thin sandstone layers with different thickness on the forward profile; (2) The forward profile is similar to the actual seismic profile, and verifies the actual geological conditions reflected by the designed geological model; (3) Combining forward modeling and inversion, comparative analysis is conducted to demonstrate the feasibility of seismic attribute research methods in identifying reservoir sand bodies, and recognize the recognition ability of wave impedance data body and reflection coefficient data body to Putaohua Formation sedimentary sandstone, as well as the parameter selection of relevant research time window in other seismic attribute extraction practical work. In this study, a two-dimensional geological model with 200 CDP points is designed at the location of the main survey line of Well Pu C. The geological model is mainly composed of the stratigraphic framework and rock physical parameters. The stratigraphic framework is composed of two main horizontal reflection interfaces. It is composed of a 60-meter-thick sandstone layer with the same thickness as the Putaohua oil layer, and has obvious velocity change interface with its top and bottom mudstone layer, There are some small lenticular channel sand bodies in the sandstone layer; The velocity and density of sandstone and mudstone layers in each layer are mainly obtained from logging curves (In Fig. 2).

Fig. 2 Seismic resolution of sedimentary channel sand body in Putaohua oil layer
In the model, the horizontal velocity of Putaohua oil layer is 3000m/s, and the channel sandstone velocity is 3400m/s. The average interval velocity of upper and lower mudstone of Putaohua oil layer is 2700m/s. Based on the data analysis of seismic response results and forward and inverse results of the two-dimensional geological model (In Fig. 3), it is considered that the seismic profile can not fully reflect the characteristics of the thin sand body of the reservoir, and the effect is better for the relatively thick sand body in the middle of the Putaohua Formation;For the thin sand layer nera the top and bottom of the grape flower layer,the intensity of amplitude energy is only reflected.

Conclusion
(1) Through the combination of forward modeling and inversion and the method of comparative analysis, the feasibility of seismic attribute research method in identifying reservoir sand body is successfully demonstrated, and the recognition ability of wave impedance data volume and reflection coefficient data volume to the sedimentary sandstone of Putaohua Formation is recognized, (2) The reliability is provided for the subsequent application of seismic attribute analysis technology to verify seismic inversion and reservoir prediction, At the same time, it also provides help for the optimization of favorable blocks and the proposed well location deployment.