M. Garden, T. Bunting, S. Ng, E. Frugier Dorrington - WesternGeco
B. Flack - Schlumberger
Gao Wei Yi, Zhang Hao - Shanghai Petroleum Corp.
Shanghai Petroleum Corp. (SPC), deployed theGeco Searcher to image the Pinghu formations, offshore Shanghai. This was the first use of Q-Marine technology in Chinese waters and for a Chinese oil company. Acquisition started in August 2007, but the initial project discussions started early in 2006, with a survey design project.
The Pinghu field has two proven reservoirs: the Huagong formation with three distinct hydrocarbon charged zones between 2,200 and 2,700 m (7,218 and 8,858 ft) depth and the Pinghu formation with four distinct hydrocarbon charged zones between 2,850 and 3,850 m (9,350 and 12,631 ft) depth. Vertical thickness of the individual sand beds varies between 1 and 60 m (3.3 and 197 ft).
SPC contracted WesternGeco to survey the Pinghu field. Prior to acquisition, a full survey design project was completed.
Survey design, acquisition
The survey design analyzed existing well log data and surface seismic data from Pinghu field and was primarily a 1D modeling study. The survey design suggested that a correctly designed survey in concert with a fully integrated Q-Marine study could provide an additional 0.6 octaves of bandwidth, and subsequent improvements in vertical resolution, of up to 50%, if the noise floor could be controlled. Q-Marine technology focuses primarily on lowering the noise level of the measured seismic response and subsequently increases signal bandwidth and resolution.
Earlier rock physics models suggest the hydrocarbon sands have a distinct acoustic impedance/Poissons ratio response. As such, the survey design additionally included some focus on enabling AVO analysis and inversion. An existing rock physics model, developed using logs of the Pinghu B8 well, separated shales and sands using gamma ray measurements. Cross-plotting this model suggested that sands and shales in Pinghu have distinct acoustic impedance and Poisons ratio response, indicating that seismically derived rock properties may help distinguish high quality sands from shales on a survey-wide scale.
Additional survey design recommendations included a 5,000-m (16,404-ft) streamer length to increase the angle range available for AVO inversion, and an 18.75-m (61.5-ft) crossline bin to meet structural and fault imaging requirements.
Signal and noise extraction from migrated sections including target zone. Legacy (right), new acquisition (left).
Acquisition went very smoothly despite two typhoons and two surface obstructions within the area of seismic operations. Typhoons are relatively common offshore China and shutdown seismic operations due to the increased noise levels and risk to surface equipment. Use of the single sensor measurements to measure and subsequently separate the low velocity noise from the signal extended the operational weather window and allowed operational loss as a result of the Typhoons to be minimized.
The survey area was populated with two surface obstructions separated by 2 km (1.25 mi) in the crossline direction and 5.6 km (3.5 mi) in the inline direction. The use of the streamer steering technology and good operational practices allowed the vessel to traverse between (2,000 m [6,562 ft] gap) the two obstructions and either side of the obstructions, with the full 5,000-m (3.1-mi) streamer length deployed. The availability of the streamer steering technology and a strong operational process resulted in the oilfield operators reducing the surface exclusion zone of the obstructions from 500 to 100 m (1,640 to 328 ft), significantly reducing the coverage hole to be filled in with existing data.
Data processing
In order to meet drilling plans, the project timeline was very aggressive. This was further complicated by the high levels of noise as a consequence of extending the weather window, external rig noise, high levels of complex surface multiples, and matching the existing seismic to fill in the holes as a result of the surface obstructions.
Processing was completed at WesternGeco in Kuala Lumpur with strong interaction from Shanghai reviewing and feeding back on all processing tests.
Comparison of zoomed images from legacy (right) and new acquisition (left). It is now possible to interpret subtle faults which were smeared with the existing image.
Surface multiples were addressed using a deterministic water layer de-multiple followed by velocity discrimination.
Results, conclusions
The primary driver for the survey was to improve resolution of the existing targets and also to open new deeper plays.
Comparisons of signal and noise extractions of the legacy and new seismic are shown in the accompanying chart. The spectra were computed from migrated sections over a time window including shallow and deep target. The new seismic provides an additional 50% (0.7 octaves) improvement in bandwidth. This matches well with the predictions from the survey design study. The analysis also suggests an additional 10 dB of separation between the signal and noise floor.
This improvement in bandwidth results in a subsequent improvement in resolution in the seismic image. This resolution improvement is detailed in the images below both in terms of ability to resolve thinner beds and better distinguish subtle faults.
The ultimate goal of any surface seismic reservoir study is to provide seismically derived reservoir properties to aid in building and refining the reservoir model. Currently, the new seismic dataset is being used in an AVO inversion project run by SPC and Schlumberger. Using the velocity field, generated during the data processing effort, it is possible to compute the incident angle as a function of depth and offset.
Based on the previous rock physics modeling and the improved seismic dataset (signal-noise, bandwidth, and useable offset), the results of this AVO inversion study are eagerly anticipated.
Acknowledgements
The authors greatly thank Shanghai Petroleum Corp. for permission to publish this work.
