No abstract

The super-giant ACG field lies in the Azerbaijani sector of the south Caspian Sea. Since the signing of the "Contract of the Century" in September 1994, the significant complexity of the geohazards setting over this field has required near continual acquisition of geophysical imagery to better understand the various geohazard issues faced.Upon signing of the PSA it was known that there was very little geophysical imagery of the shallow section in existence. As such, the first geophysical operation over the PSA contract area in 1995 was a blanket regional 2D survey of the entire 450km 2 PSA contract area to allow regional geohazards mapping to be undertaken. This entailed acquiring a regional grid of HR2D seismic data and a seabed survey that included collection of swath bathymetry. The challenges of importing acquisition systems into Azerbaijan, and mobilizing the equipment onto vessels of opportunity, limited the systems that could be used at the time. Therefore, the bathymetric model that was produced, for example, was useful, but limited by a swath bathymetry system that could only image to~210m of water depth while PSA contract area water depths vary between 96 and 425m.Over the following decade, geophysical site investigations of appraisal wells, platform sites and pipeline routes followed industry norms of site specific 2D surveys.However, in 2004, one pass HR3D and seafloor surveys were acquired making use of a 3D seismic vessel. The use of the larger vessel had various advantages. Firstly, the capability to safely tow four streamers and two sources allowed acquisition of eight lines of subsurface coverage per sail-line. Secondly, the more stable vessel platform, allied with a more robust design of over-side mountings, provided a far superior platform for acquisition of swath bathymetry data. The resulting seabed and sub-seabed imagery were far superior to preceding imagery.In 2007, another step change in data quality was achieved with the acquisition of the first ever deep-water AUV survey acquired in the Caspian Sea. Using a Hugin 3000 vehicle, the resulting swath bathymetry, sonar and sub-bottom profiler imagery saw another step change in quality.HR3D and AUV data of the entire PSA have now been acquired in a phased approach. Included in the HR3D acquisition were undershoots of the six producing platform complexes to verify the integrity of overburden conditions below the platforms. This paper will show the improvements in data quality that have been achieved over life of the PSA, and demonstrate the impact these improvements have had on better understanding of geohazards for future development and ongoing operations across the field.

Shallow localized gas pockets cause challenging problems in seismic imaging because of sags and wipe-out zones they produce on imaged reflectors deep in the section. In addition, the presence of shallow gas generates strong surface-related and interbed multiples, making velocity updating very difficult. When localized gas pockets are very shallow, we have limited information to build a near-surface velocity model using ray-based reflection tomography alone. Diving-wave refraction tomography successfully builds a starting model for the very shallow part. Usual ray-based reflection tomography using single-parameter hyperbolic moveout might need many iterations to update the deeper part of the velocity model. In addition, the method generates a low-velocity anomaly in the deeper part of the model. We present an innovative method for building the final velocity model by combining refraction, reflection, and wave-equationbased tomography. Wave-equation-based tomography effectively generates a detailed update of a shallow velocity field, resolving the gas-sag problem. When applied as the last step, following the refraction and reflection tomography, it resolves the gas-sag problem but fails to remove the low-velocity anomaly generated by the reflection tomography in the deeper part of the model. To improve the methodology, we update the shallow velocity field using refraction tomography followed by wave-equation tomography before updating the deeper part of the model. This step avoids generating the low-velocity anomaly. Refraction and wave-equation-based tomography followed by reflection tomography generates a simpler velocity model, giving better focusing in the deeper part of the image. We illustrate how the methodology successfully improves resolution of gas anomalies and significantly reduces gas sag from an imaged section in the Greater Cassia area, Trinidad.

No abstract