Introduction to InSAR Data
Interferometric Synthetic Aperture Radar (InSAR) is a powerful remote sensing technique that produces a very detailed picture of the shape of the ground surface. The data returned are so precise that they can reveal changes in the elevation of the earth down to the level of millimeters per day, enabling very small changes in surface elevations to be quickly discovered, which can be critical to land management. Additionally, the method offers speed and ease of data acquisition: The data are typically collected by orbiting satellites and are immune to cloud cover and other access issues. As they are space borne, they have an inherent repeatability, often on the timescale of weeks between duplicate data collection over the same location. In summary, InSAR data are a powerful and ubiquitous tool for detecting ground elevation changes.
Use Case: InSAR Analysis of the Belridge Field
The potential for high density / high accuracy surface deformation (uplift and/or subsidence) information from oil field operations has significant financial, operational, and, critically, safety implications. This is especially acute in the San Joaquin Valley of California where heavy oil production, and water and steam injection, are often from very shallow reservoirs (<1000 feet below surface). Production from these shallow reservoirs is rapidly and strongly reflected in surface conditions, and if not correctly mitigated can lead to infrastructure challenges up to, and including, well failure. InSAR data are the ideal tool for monitoring these entire fields. However, the true value of InSAR data is revealed when the data are fully integrated in a diverse, contextual 4D environment. This must necessarily include temporal records of production and injection data, and can include surface infrastructure, subsurface geologic models, well trajectories and, potentially, microseismic and tilt meter data. The temporal component is paramount in this integration.
Figure 1. Location of study area: Belridge field in the central San Joaquin Valley, California.
Using the CoViz 4D software from Dynamic Graphics for this temporal data integration, we present a case history spanning nearly twenty years for the Belridge field in the central San Joaquin Valley, California (see Figure 1). Multiple years of InSAR data (© 2016 SkyGeo) were combined with digital elevation data and publicly available well locations and production and injection records.
We observed long-term subsidence patterns that are clearly related to fluid production, plus pockets of local uplift interpreted as relating to over-injection. Furthermore, assessing forward modeling with simple geomechanical models showed these can be used to quantify and predict injection performance. Thus, careful integration of InSAR data can yield benefits for operators, including:
- Planning injection interventions
- Fewer well integrity issues
- Savings on drilling costs
- Better targeting and monitoring of injection campaigns
The movie (Figure 2) offers an integrated, albeit qualitative, view of the well and InSAR data. The correlation between drilling activity and long-term subsidence of the field is easily observed.
Figure 2. Animation of the well and InSAR data showing correlation between drilling activity and field subsidence.
To quantify this relationship, we subjected the InSAR data to a rigorous spatiotemporal workflow which subdivided the field into ~6-acre segments and studied the net production / injection history versus the uplift / subsidence history in each segment, as illustrated in Figure 3.
Figure 3. Workflow for comparing net production / injection history with the uplift / subsidence history in roughly 6-acre segments of the study area.
This workflow was facilitated by the flexible and scalable tools available in the CoViz 4D Developers’ toolkit. These allow for a wide variety of data science-type workflows to be quickly and easily constructed for spatial and temporal data. In this case we captured and archived the workflow in a Jupyter Notebook environment. The conclusion from this analysis is captured in the movie in Figure 4 which shows the clear correlation between well production / injection activity and subsidence.
Figure 4. Animation showing the correlation between well production / injection activity and field subsidence.
To further study the nature of subsurface response to the oilfield operations we built a first-order geomechanical model. Using simple homogeneous relationships and approximate assumptions regarding subsidence modeling, we observe that the InSAR-derived subsidence can be modeled using reservoir compaction linearly scaled by production. Figure 5 captures the essence of this correlation, showing that a first-order reservoir thickness change related to fluid production can explain the observed subsidence.
Figure 5. A first-order geomechanical model demonstrated that fluid production related reservoir thickness changes accounted for the observed field subsidence.
As illustrated in Figure 5, overwhelmingly the InSAR signal indicates a broad subsidence related to production from the reservoir which is not mass-balanced with an equivalent volume of injection. This broad long-term subsidence is expected from these reservoirs, and with prior planning and attention the most severe impacts can be mitigated. However, imposed on top of this broad subsidence are very short wavelength areas of uplift, as illustrated in Figure 6. These much smaller, but much more rapidly developing features are typically related to an over-injection of water or steam, and possibly indicate compromised subsurface communication. The size and speed of the uplift was a significant concern to operators: If the uplift is not identified and arrested quickly it can soon lead to a “surface expression” in which fluids can migrate from the subsurface to the subsurface along newly created fractures. Again, InSAR proved invaluable in rapidly identifying these issues and allowing operators to take immediate corrective action.
Figure 6. Localized areas of rapid surface uplift related to an over-injection of water or steam were observed and mitigated.
In summary we have seen how InSAR data are extremely valuable to onshore operators for revealing both long-term subsidence and short-term uplift events. When integrated in a quantitative platform like CoViz 4D from Dynamics Graphics, the full value of these data can be harnessed and utilized. These workflows can help lower risk and reduce lost-time events in shallow onshore oil provinces such as the San Joaquin Valley of California.
See the CoViz 4D page for more information.
