Reservoir Characterization by Investigating the Reservoir Fluid Properties and their Effect on Seismic Response of Fenchuganj Gas Field , Bangladesh

However, for the interpretation and evaluation of structural or stratigraphic features in the subsurface, the seismic data are commonly used. The physical properties of pore fluids have a vital effect on the seismic response of a porous rock containing it. It is essential to have an understanding of the changes in p-wave (compressional) velocity, s-wave (shear) velocity, and density as fluid or rock properties change to know or predict the effect of changes in seismic amplitudes and travel times.


Introduction
Reservoir characterization incorporates all the characteristics of the reservoir that are relevant to its ability to store hydrocarbons and also to produce them. Models for reservoir characterization are used to suggest the behavior of the fluids within the reservoir under different sets of situation and to find the best possible production techniques that will maximize the production.
However, for the interpretation and evaluation of structural or stratigraphic features in the subsurface, the seismic data are commonly used. The physical properties of pore fluids have a vital effect on the seismic response of a porous rock containing it. It is essential to have an understanding of the changes in p-wave (compressional) velocity, s-wave (shear) velocity, and density as fluid or rock properties change to know or predict the effect of changes in seismic amplitudes and travel times.
The amplitude versus offset (AVO) is a general term in reflection seismology for referring to the dependency of the seismic attribute, amplitude, with the distance between the source and receiver (the offset). AVO analysis is a method that geophysicists can accomplish on seismic data to determine a rock's fluid content, porosity, density or seismic velocity, shear wave information, fluid indicators (hydrocarbon indications [12]). The P-wave and S-wave velocity, bulk density, acoustic impedance, Poisson's ratio (PR), and bulk modulus are determined from Batzle and Wang [8] without considering the rock matrix and from Gassmann-Biot models as a function of the saturating rock fluids.
Fenchuganj Gas Field is one of major gas producing fields in the Surma basin with estimated reserves of 553 Bcf . These authors already have worked on first two zones (New gas Zone [13] III and New gas Zone II) of Fenchuganj Gas Field. However, it is required to work with all four layers for better reservoir characterization [31]. Here we aimed i) to predict the velocity, density and modulus of fluid/fluid saturated rock matrix samples for both varying saturation with constant/varying pressure using Batzle and Wang model and Gassmann-Biot model, and ii) to predict seismic response from the layered rock properties using Abstract Fenchuganj Gas Field is located in the Surma Basin of Bangladesh and characterized by water-drive gas field. In the reservoir condition, water saturation increases as gas production rise. The fluid properties of the four individual gas zones of this reservoir at the present condition and at the gas depleted condition should be addressed with proper prediction. In this paper, we characterize the total reservoir with special emphasis on Upper Gas Zone and New Gas Zone I which are compared with other two gas zones (New Gas Zone III and New Gas Zone II) representing some modeling results (has done before by these authors) which evidences that the pore fluids have a significant effect on the acoustic impedance and the Poisson's ratio of the reservoir rock which is directly correlated with seismic amplitudes at constant pressure with Batzle-Wang model and Gassman-Boit models. These models with varying pressure and water saturation conditions show the reasonable predicted fluid modulus against pressure for all four gas sands. The reservoir modeling from irreducible water saturation condition (90% gas saturation) to residual gas condition (10% gas saturation) provides a way to estimate values at reservoir conditions from logging conditions. Fluid bulk density increases when water saturation increases with constant pressure and stay around constant when water saturation increases with pressure drop. But overall it increases through the production path that we assumed. Amplitude versus Offset (AVO) analysis is also compared with other study models which show that seismic reflection of p-wave changes due to change of pressure and water saturation of the reservoir rock layers. This study is also showing that all four gas zones of the Fenchuganj are under gas sand category 3. We propose the modeling of fluid property in determining the convenience of time lapse seismic, predicting AVO and amplitude response, and forecasting in the study field and making production and reservoir engineering decisions.

Result and Discussion
Four gas bearing zones are present and identified in the Fenchuganj gas field. They are New Gas Zone III, New Gas Zone II, Upper Gas Zone and New Gas Zone I. Analysis of all four gas zones is presented in this paper while analysis of the first two gas zones has already published [17]. We emphasized on the mainly Upper Gas Zone and new gas Zone I, and compared the results with other two zones. The results of all zones are tabled of course.
New Gas Zone III was found at a depth of 1656-1680m,with pressure16.3888 MPa (2377psi), temperature 46.67°C (116°F), porosity 27.3%, gas saturation 54%, water saturation 46% and salinity 8500 ppm. The New Gas Zone II with depth 1992-2017 m, pressure 19.7328 MPa (2862psi), temperature 51.11°C (124°F), porosity 14.5%, gas saturation 36%, water saturation 64% and salinity 9500 ppm. The Upper Gas Zone with depth 2030-2086 m, pressure 20.1121 MPa (2917 psi), temperature 51.67°C(125°F), porosity 25%, gas saturation 60%, water saturation 40% and salinity 10000 ppm. And the New Gas Zone I with depth 2148-2154 m, pressure 21.2841 MPa (3087psi), temperature 55.56°C (132°F), porosity 24.8%, gas saturation 57%, water saturation 43% and salinity 10500 ppm. For all gas zones, density of airs 0.00122g/cc, the API gravity of condensate is 31.86°, gas-condensate ratio is 142260, gas constant(R) is 8.3145, and specific gravity of gas is 0.5624. These are the initial condition of gas layers Table 2. Fluid properties for gas zones were analyzed using the aforementioned methods of the previous paper by these authors [17]. Our main aim was to investigate and forecast the behavior of reservoir fluid properties during production which are discussed below.

Fluid models for Varying Saturation under Constant Pressure
Batzle and Wang Model: Using Batzle and Wang model for varying saturation with constant pressure for initial conditions, i.e. parameters, we have calculated density (ρ), acoustic velocity (V P ) and modulus (k) of gas, brine and mixture phase ( Table 3).
The calculated result shows that the density(ρ), acoustic velocity (V p ) and bulk modulus (k) are 0.1367 g/cm 3,549.46 m/s, and 41.25 MPa for Upper Gas Zone, whereas New gas Zone I shows these values are 0.1412 g/cm3, 559.70 m/s, and 44.24 MPa, respectively for gas phase (Table 3). For brine and mixture phase, all the parameters (density, acoustic velocity and bulk modulus) have changed significantly (Table 3).
The cross-plot between bulk modulus and density for changing the saturation of gas zones have been formulated which are given in Figure  2.The values of modulus and density for changing saturation are listed in Table 4. AVO (Zoeppritz equation) for all four layers with special emphasize on Upper Gas Zone (most targeted zone) and New Gas Zone I of Fenchuganj Gas Field.

Geological Setting and Stratigraphy of the Study Area
Fenchuganj structure of Surma basin lies under Fenchuganj Upazila of Sylhet district. It is about forty kilometers southeast of Sylhet town. Geographically, it is bounded by longitude 90°53'-92° east and latitude 24°30´-24°37´ north and it is tectonically located in the transition zone between the central Surma basin and the Folded Belt in the east (Figure 1).
The Surma Basin of Bangladesh experienced a variety of sediment facies, indicating a range of depositional environments during the Neogene time [15]. Furthermore, during the Miocene time, the Sylhet Basin has a noticeable subsidence and marine transgression. The transgression of the Miocene certainly affected the coastline. It is believed that the Surma Basin has undergone two successive phases of evolution; the marine transgressive phase, followed by a regressive phase resulting in a series of continental fluvio-deltaic to marginal marine sedimentation during the Neogene. The thickness of the late Mesozoic and Cenozoic strata in the Surma Basin range is from about 13 to 17 km [15,16], and much of this group is Neogene in age. The Great Himalayan Orogeny and related tectonics subject the Surma Basin during the Miocene-Pliocene times. However, major changes in sea level for Neogene are transgressive-regressive phenomena suggested by [15].
Fenchuganj structure is an elongated structure and about 30 km long and 8 km wide. It is a reversibly faulted asymmetrical anticline with NNE-SSW trending axis [16]. The eastern flank has sharp dip than the western flank. The amount of dip in the eastern flank varies from 30°-35°, whereas in the western flank dip varies from 20° to 25°. Basin is same as Sylhet Trough (after [14]); the purple asterisk locates the reservoir area.

Page 3 of 25
density (ρ w ), bulk modulus (K w ), hydrocarbon density(ρ hyd ) and bulk modulus (K hyd ).The important output values are the bulk density (ρ d ), P-wave velocity (V p ), S-wave velocity (V s ), acoustic impedance (A I ), and Poisson's ratio(σ) as they vary due to changes in saturation. The dry frame modulus is held constant.
Using the Gassmann-Biot model at the reservoir condition, we found that the values of dry frame rigidity (G) are 6.58438 and 3.83328 GPA, bulk density (ρ) are 2.19998 and 2.1947707 g/cm 3 (Table 6). Some other values of output parameters for different saturation are listed in Table 6.
Saturations of gas zones change from the initial condition with production. In Figure 2, the red marked position is the initial condition, whereas the green line indicates the production direction. These Figures show that for saturation changes with constant pressure, both the bulk modulus and density increases as the water saturation increases. However, the density increases very rapidly and bulk modulus increases slowly at the initial stage of production, whereas reverse situation (density increases slowly and bulk modulus increases very rapidly) exists at the later stage.

Gassmann-Biot model:
In this section, Table 5 and Table 6 demonstrate the use of the Gassmann-Biot equations with fluid and rock properties to determine the overall reservoir rock seismic properties such as velocity and density. The input values include porosity (ρ), solid material bulk modulus (K s ) and density (ρ s ), brine        In this paper, we have assumed full water (wet) and gas saturation is 95% water and 90% gas, respectively. We also considered the irreducible gas saturation or depleted reservoir condition at 20% gas or 80% water ( Figure 6 and Figure 7).
Based on the theory on pore fluid distribution in pore space, it is resolved that pore fluid should rise compressional velocity and decline slightly shear velocity of the rocks [2,3]. Moreover, experimental results also demonstrate that compressional and shear velocity are related to saturation [18,19]. According to Gregory [21], the saturation has larger effect on rock velocities in low porosity rocks than that of rock with high porosity. VP in fully water-saturated rocks is apparently larger than those in partially water-saturated rocks. V S do not always fall with the rise of saturation. Instead V S is related with pressure, porosity and the chemical interactions between pore fluid and rock skeleton. Our modeling result also displays ( Figure 3) the compressional velocity (V p , blue curve) increases with the increasing water saturation. During the production for upper Gas Zone and New Gas Zone I the water saturation varies from water saturation 40% to 80% and 43% to 80% respectively at reservoir conditions. Within this saturation range a significant variations has been found in the compressional velocity (20.67% and 24.68% for Upper Gas Zone and New Gas Zone I, respectively ( Table 6). The

Fluid models for varying saturation and pressure
Batzle and Wang model: In this section, we considered gas saturation would change from 90% to 10% at reservoir pressure of 2000 psi, 1500 psi and1000 psi pressure for New Gas Zone III and at reservoir pressure of 2500 psi, 2000 psi, 1500 psi and1000 psi for New Gas Zone II, Upper Gas Zone and New Gas Zone I. For different saturation and pressure conditions the moduli and densities were calculated from this model are listed (Table 7).The cross plot between fluid modulus and density versus pressure shows that the fluids have a wide range of fluid moduli and densities of different saturation conditions ( Figure 6). The different fluid moduli and densities for the initial reservoir pressure conditions are shown by the yellow diamonds.   decreases through the assumed production path. So, the effect of pressure fall is dominating here for fluid modulus changes.
However, a reverse condition has been seen from the Figures 6e-h, which indicates that the fluid density of the reservoir is not affected as strongly as the modulus by pressure changes and variations of saturation. The density is increasing with the increase of water saturation and decreases with pressure fall for both gas zones.

Gassmann-Biot Model:
In this section, the initial conditions of the Figs. are same as the Figure 6. All the outputs from Gassmann-Biot model are listed in Table 8. As we know that the Biot-Gassmann theory precisely forecasts velocity ratios with respect to differential pressure for given porosity. However, because the velocity ratio is weakly associated to porosity, it is not suitable to investigate the velocity ratio with respect to porosity (φ). The velocity ratio has been used for many purposes, such as a lithology indicator, determining degree of consolidation, identifying pore fluid, and predicting velocities [18][19][20]. The velocity ratio usually depends on porosity, degree of consolidation, clay content, differential pressure, pore geometry, and other factors. The velocity ratio for dry rock or gas-saturated rock is almost a constant regardless of porosity and differential pressure, whereas the velocity ratio of wet rock depends significantly on porosity and differential pressure. P-wave to S-wave velocity ratio (V p /V s ) for the all gas zones Fenchuganj Gas Field show value of more than 2.0 which show the    presence of gas in the unconsolidated rock [21] with higher porosity [22]. Figure 7 displays the predicted P-wave velocity, PR and acoustic impedances under differential pressure due to gas production rises. The black connecting line with arrow head displays the downward curve for the decreasing value of all parameters respects to pressure fall during production. These parameters rise with water saturation at a constant pressure, but falls with pressure drop. So, it can forecast that the New Gas Zone II has greater effect on parameters change than the other three gas zones.

AVO analysis
At this final stage, we determined and analyzed the AVO response for Upper Gas Zone and New Gas Zone I and compared with New Gas Zone III and II (AVO response for New Gas Zone III and New Gas Zone II has already published as Islam et al. 2014 [16] of Fenchuganj Gas Field. In geophysics and reflection seismology, amplitude versus offset (AVO) is the general term for referring to the dependency of the seismic attribute, amplitude, with the distance between the source and receiver (the offset). According to Schlumberger Oilfield Glossary, AVO analysis is a technique that geophysicists can execute on seismic data to determine a rock's fluid content, porosity, density or seismic velocity, shear wave information, fluid indicators. If matches with the standard figures AVO gives the validity of the reservoir properties which are used to interpret it.
Zoeppritz [1] equations provide a complete solution for amplitudes of transmitted and reflected P-and S-waves for both incident P-and Swaves. The equations are very complex and subject to troublesome sign, convention, or typographic errors. Hilterman [18], Aki and Richards [22], and Shuey [23], developed simplifications and approximations for Zoeppritz equations.
So, at first, we took assumptions given by Shuey [23] for simplification of Zoeppritz [1] equation and interpreted AVO that empirical equation for 0-30° incident angle. The phenomenon is based on the relationship between the reflection coefficient and the angle of incidence. Zoeppritz Equation: Where, R (θ) = reflection coefficient (function of θ), θ= angle of incidence, A = zero-offset reflection coefficient (AVO intercept) and B = slope of the amplitude (AVO Gradient).
The input values (V p , ρ b and σ) for AVO interpretation are taken from the output values (Table 9)  Secondly, we have interpreted AVO without any assumptions. We used full Zoeppritz [1] equation with the help of an AVO calculator by Timothy et al. which starts with negative values and decreases with the offset indicating of low impedance gas sand class 3 AVO [11]. This characteristic of AVO indicates bright zone that is potential for hydrocarbon zone [11].
The input values used to determine AVO, were taken from the results of Gassmann-Biot model which is also dependent on outputs of Batzle and Wang model. Both approaches applied in this section, the AVO reflection indicates class 3 type and it is a good sign to be sure that Fenchuganj is a gas reservoir and ironically we know it. So, it can be inferred surely that the fluid properties we determined in previous sections by Batzle and Wang model and Gassmann-both models are fairly correct.

Conclusion
The fluid properties for both of the constant/varying saturation with constant/varying pressure condition are analyzed by the Batzle and Wang model that predicted near precise forecasting. Increase  of water saturation effects on fluid properties by increasing the fluid density, modulus and acoustic velocity among the all gas sand layers of the Fenchuganj Gas Field. The compressibility of fluid declines as the water in fluid rises. Similarly, due to temperature rises, the velocity and density of the fluid fall. The cross plots using the Batzle and Wang model on densities and moduli allows to predict the fluid properties as the reservoir is produced and shows the effect on the reservoir as water saturation rises and gas saturation drops. The modify in P-wave and S-wave velocity, bulk density, acoustic impedance, Poisson's ratio, and bulk modulus were predicted using the Batzle and Wang and Gassmann-Biot model which show that the reservoir changes from irreducible water saturation conditions in residual gas conditions which provide an avenue to calculate values at reservoir conditions from logging conditions. Coupling with the Batzle and Wang, Gassmann-Biot, the AVO models can be used to determine expected seismic responses throughout the production path of the reservoir and coincide with previous results [17]. In case of the Fenchuganj Gas Field, it is shown that an AVO response is presented as a result of the fluid and rock properties and also show that the reservoir is pressure decreases due to increasing the gas production. The AVO modeling for fluid property investigation will help in determining the usefulness of time lapse seismic, predicting AVO and amplitude response, and making decision on forecasting and production for the reservoir.