Combined with supercomputing technology, FWI provides our geoscientists with unparalleled insights into subsurface geologic structures and the physical characteristics of rocks.
In the past, the industry could only utilize a portion of the data recorded in 3D seismic surveys due to the high complexity of the seismic waves and the sheer quantity of data. Now, FWI has provided us the ability to better leverage all of the data from a seismic survey to produce high-definition images of the subsurface. Specifically, FWI enables us to more accurately map geologic structures and determine with more certainty whether or not the rock may contain hydrocarbons. In doing so, we can help make oil and gas resources easier to identify and target in exploration, development and production programs. For fields where ExxonMobil is currently producing oil and gas, FWI is applied to time-lapse, or 4D seismic surveys to help map out remaining hydrocarbons. These capabilities can make resource development and production more efficient, less costly and more environmentally friendly through optimised drilling.
The full wavefield inversion process begins with a geologic model of the subsurface rock properties. Computer-generated synthetic seismic data derived from the model is then compared to the real-world field seismic data. The differences between the two are used to update the model to more accurately reflect reality. This process is repeated until there is a match between the synthetic and real data. The simulation needs to incorporate a high degree of realism, correctly representing the seismic acquisition geometry and the physics governing seismic wave propagation in the subsurface.
The FWI approach is a significant departure from the way seismic data are currently used to image the subsurface. Standard seismic analysis is time- and labor-intensive, consisting of a series of operations designed to remove noise and unwanted signals from the data prior to imaging. After imaging, subsurface properties are estimated, but there is no explicit feedback loop to compare these estimates with the original field data. In contrast, the FWI process is largely automated using computer simulation to bypass the time-consuming steps of the traditional approach.
ExxonMobil’s proprietary algorithms and investments in supercomputers allow us to increase the range of frequencies used by full wavefield inversion so we can create remarkably precise maps of subsurface structures. In most published FWI studies, only the lowest frequency portion of the data (less than 10 Hz) is inverted, resulting in low-resolution models. At ExxonMobil, we have run FWI on 3D seismic surveys using much higher frequencies, generating high resolution models of the subsurface.
The improved imaging and reservoir characterisation provided by FWI has quickly translated to improved results in the field. For example, high-resolution rock property volumes derived from our FWI technology were used to optimally target drilling and form an integral part of our ongoing Black Sea exploration and development efforts.
Full wavefield inversion remains an active area of research at ExxonMobil. Advancements in computing and algorithms remain necessary as we use more complex simulation physics and push the resolution limits of seismic data. Our investments in high-performance computing and in full wavefield inversion technology have allowed us to be an industry leader in seismic imaging.