In hydrocarbon exploration, rock physics analysis plays a key role by connecting seismic data to rock properties. Analysis of rock physics data enables geophysicists to understand how fluid content affects the seismic response and what they should look for to improve the chance of finding hydrocarbons. In the Nong Yao oil field, the use of rock physics and AVO analysis was used to improve the hydrocarbon prediction process. The preferred method starts with rock physics analysis of key wells. Fluid Replacement Modelling (FRM) is then performed across many wells in order to generate a predicted seismic response for different pore fluids (gas, oil and brine). The predicted AVO response is then calibrated against the actual AVO response from the seismic data from key wells in order to build a database. In the Nong Yao Field, over four hundred and fifty drilled data points from sixty-nine wells were utilized in the analysis. This database is analyzed in order to find the optimal combination of parameters for hydrocarbon prediction, which is then used to improve hydrocarbon prediction for future near-field drilling candidates. Near-field appraisal programs in the Nong Yao oil field are driven strongly by amplitudes and AVO, as rock physics analysis has shown that sands and shale lithologies can be easily discriminated based on acoustic impedance. Fluid prediction is more difficult based on acoustic impedance alone, as other factors such as variable sand thickness and seismic data quality mean that there is significant overlap between hydrocarbon and wet sands. Rock physics analysis has shown that AVO behavior can be included to provide better separation between hydrocarbon sands and wet sands. AVO signatures from all the data points are then analyzed using intercept vs gradient cross-plots. A background wet trend is defined with the clear observation that increasing distance from the background wet trend correlates to increasing chance of hydrocarbon fill. Data are categorized into weak, moderate and strong AVO response based upon their distance from the background wet trend and then this is used to modify the chance of success of near-field appraisal drilling targets utilizing conditional probability. This results in an increased chance of success of up to 20% in a strong AVO supported target and around 10% in a moderate AVO supported target. Targets are then quantitatively high-graded in an appraisal portfolio.

The availability of high quality seismic data is of critical importance in trying to unravel the complexities of subsurface geology. This paper illustrates how proper selection of seismic acquisition parameters and data processing techniques can successfully overcome geological difficulties and minimize uncertainties when exploring for hydrocarbons in the northern part of Block G11/48, Gulf of Thailand, without compromising safety and cost efficiency. Fluvial and fluvio-deltaic sediments of early to mid Miocene age in low relief faulted structural traps constitute the most common hydrocarbon habitat in the area of investigation. Amplitude support in identifying potential targets is also proven by nearby discoveries. To fully evaluate the exploration potential of this area, a 3D seismic acquisition campaign was successfully carried out using a high-end seismic vessel and without HSSE incidents. The Nong Nuch dataset was acquired using a 10 deep-flat towed 5.1 km streamer configuration with triple sources to increase cross-line resolution and reduce operational time. This long streamer length relative to the target depth provides necessary information to Full Waveform Inversion and Q-tomography in order to correct push-down effects in broadband anisotropic Pre-stack Depth migration. The dataset also helps to obtain high accuracy velocity model, de-multiple and quantitative interpretation. This acquisition and processing approach significantly improved the ability to image thin reservoirs and correct push-down effects and energy absorption due to gas clouds or unconsolidated sea floor channels. The large streamer spread and deep tow did not create any major problems throughout the acquisition. The implementation of broadband acquisition and leading edge processing techniques resulted in good signal to noise ratio as well as high vertical and horizontal resolution with minimal acquisition footprint. In addition, long offset data acquisition contributes to successful attenuation of short and long period multiples. Channel-like features and fault plane reflections are very clearly imaged in the dataset, helping to better understand of the depositional environment and structural setting of the area. Severe push-down and abnormal amplitude absorption effects were significantly corrected and compensated by building a high resolution, Full Waveform Inversion (FWI) derived velocity model as well as application of reflection tomography and Q-tomography techniques. Thus, definition of potential traps beneath gas clouds has significantly improved.

No abstract

No abstract