Gulf of Mexico case study
Introduction
This deepwater Gulf of Mexico case study illustrates how seismic data obtained with Q-Technology improve reservoir definition, ultimately improving production rates and reducing drilling costs. The Q-Marine data resolved stratigraphic details of likely production barriers and reservoir sand units as thin as 5 m (15 ft). An integrated approach was used that combined the Q seismic data, structural and stratigraphic interpretation, rock-physics modeling, and advanced seismic inversion methods (Figure 1). The approach predicted reservoir parameters such as water saturation (Sw), porosity (f), and the probability of commercial gas sand.
Geologic setting and interpretation
The study area lies in an intraslope salt withdrawal basin in the southern East Breaks protraction area, western Gulf of Mexico. Water depths range from 1,400 to 1,520 m (4,600 to 5,000 ft). Gas was discovered in 1990 at about 3,140 m (10,300 ft) subsea. The narrow, downdip oil rim was developed first and gas development has begun. Production is not meeting expectations, most likely due to compartmentalization caused by thin lenticular sand lobes that are evident on the Q data (Figure 2).
The trap is primarily stratigraphic, as sands pinch out on the flank of a salt dome to the northwest. Turbiditic sands were deposited in lobes that downlap and fill in lows to the southeast. The Lower Pleistocene reservoir interval, commonly known as A-50, consists of fine- to very fine-grained, unconsolidated, turbidite sands that were deposited in a deepwater environment. Below the A-50 sands is a shaley interval about 100 ft thick, followed by another sandy layer, known as the A-70. Reservoir quality for the A-50 is good, with average permeability about 800 md and porosity about 30%. Gross reservoir thickness ranges from around 30 m (100 ft) in the northern, more proximal, portions of the field to roughly 15 m (50 ft) to the south. Q data, with bed resolution as thin as 5 to 6 m (15 to 20 ft), reveal that sand beds are often thin and compartmentalized by layers of low permeability silts and shales (Figure 3).
Sand sections are low-impedance events relative to surrounding shales, so the seismic signature of the field exhibits a "bright spot". Flat events, diagnostic of fluid contacts, are visible along much of the downdip edge of the field.
Seismic inversion and classification
A multidisciplinary approach involving seismic inversion, geological interpretation, petrophysics, and rock physics was used to estimate seismic-based reservoir properties. Each step of this approach benefits from the enhanced data quality provided by Q-Technology.
Hybrid seismic inversion technology, which combines full waveform prestack inversion and AVO inversion, produced relevant 3D reservoir attributes, e.g., acoustic impedance, shear impedance, and Poisson's ratio. Quality of inversion, defined in terms of resolution and accuracy, greatly benefits from the high signal-to-noise ratio of the Q data. The benefits of the high s/n ratio on the resolution and accuracy became clear when remarkably good agreement was found between observed and predicted elastic properties at two "blind" well locations, where log data were not used for calibration.
Log data from three wells were used to define rock-physics relationships between reservoir parameters and seismic elastic attributes. These transforms were applied to the inversion outputs using a Bayesian estimation approach. The goal is to generate probability estimates of lithology and fluid combinations for a given attribute (Figure 4). These estimates can be improved by providing additional independent information: in our case, seismic interpretation at the reservoir level. Here again, increased resolution from Q seismic data helped in fine-tuning the reservoir geological features. This integrated workflow produced 3D maps of reservoir lithofacies (lithology and fluid), with associated probabilities and uncertainties (Figure 5).
Another methodology, mixing full waveform prestack inversion and geostatistical cokriging, was used to produce geocellular 3D models of water saturation (Figure 6) and porosity at the reservoir level. These models could be used directly by the reservoir engineer as an input to reservoir simulators. The quality of Q data enables the use of the small grid cells that are required to consider both vertical and lateral geological heterogeneities.
This hydrocarbon field has been in production since the early 1990's. The water saturation analysis indicates production-related effects of water coning in the vicinity of producing wells (Figure 7). Furthermore, the accuracy of the definition of the fluid contacts is significantly enhanced compared to simply interpreting seismic amplitudes or direct hydrocarbon indicators.
Results and potential benefits
Q seismic data resolved thin sand beds (5 m) and revealed vertical heterogeneities within the pay interval. Reservoir characterization results-porosity, water saturation, and probability of commercial gas sand-indicate that the field is also laterally heterogeneous. Using Q-Technology, strategic well planning will help produce from the sweet spots and improve production rates.