RRC Technical Meeting on Seismicity
Summary and Recommendations
Introduction
On Friday, June 5, 2015, Commissioner Ryan Sitton and RRC technical staff met with authors of
the study entitled, “Causal factors for seismicity near Azle, Texas.” Also in attendance were Dr.
Dan Hill, Chair of the Petroleum Engineering Department at Texas A&M University, and
scientists from EnerVest Corporation. The meeting was open to the media and live streamed for
the public to view.
The goals for the meeting were: (1) to understand the nature of this and similar studies; (2) to
explore what additional data could contribute to better scientific understanding of the processes
that may be involved; (3) to assess how these data might be efficiently and economically
obtained; (4) to determine what additional information could be requested of operators; (5) to
consider what regulatory changes the Railroad Commission could make to ensure that oil and gas
continues to be developed safely yet with minimal unwanted impact in Texas; and (6) to discuss
what future research and study is taking place that will enhance and further refine the scientific
findings for Azle and other areas in Texas.
All in attendance agreed that the researchers’ work was important and a good first step in
beginning to understand very complex models, assumptions and preliminary conclusions
regarding seismicity in Azle and across Texas.
Key Points
As it relates to the specifics of the study, there were some key findings based on explanations
from the research team:
1. The researchers stated that hydraulic fracturing was not a cause of these earthquakes, and
that this is a common misconception in the public and media that should be resolved.
2. The researchers made it clear that the vast majority of disposal wells do not have
associated seismicity. The researchers stated that disposal wells in close proximity to
critically stressed faults could change pore pressure and reduce effective stress, resulting
in seismicity if faults are oriented in a particular way.
3. The researchers stated that their work is ongoing and that better data could enhance their
modeling and research. As Dr. Ellsworth with the USGS stated, these are hard problems
and they’re not going to be answered quickly.
4. Figuring out how and if pressure was transmitted over 4 kilometers to the basement is a
key question. The origination depth of the earthquakes seems uncertain. The study
focused on pressure building up in the Ellenburger at a depth significantly shallower than
the depth of the recorded earthquakes in the basement. The researchers hypothesize that

  
 

the deeper earthquakes are due to downward pressure transfer within the fault system.
Enervest produced a document showing that the early sequence of events occurred in the
basement along the primary fault at around a depth of 20,000 feet. The Ellenburger depth
quakes were around 10,000 feet deep. The research model investigated pressure changes
along the antithetic fault at Ellenburger depth and did not investigate pressure changes in
the basement. The researchers said more study is needed to investigate the permeability
and porosity along faults to understand how pressure could have been transmitted to
basement depths, if in fact that is possible.
5. The study suggests that production of brine contributed to induced seismicity. However,
when this issue was discussed it was clear that more research is needed to determine
actual levels of brine production and what formation that water is being produced from.
Enervest believes that the majority of produced water in the area is from the Barnett
Shale and production of water from the Ellenburger is negligible. The statement in the
abstract that “a combination of brine production and wastewater injection near the fault
generated subsurface pressures sufficient to induce earthquakes,” seems arguable given
the discussion at the meeting, particularly if little to no Ellenberger water is being
produced (the model assumed all reported water production came from the Ellenburger).
6. The researchers said that Well #2 did not play a significant role in causing seismicity
(was not “very important”). The researchers also stated that because seismicity in the area
had decayed, there was not much more for the RRC to do vis-à-vis these injection wells.
They recommended keeping Well #1 at its current injection volumes and pressures and
stated that they would continue seismic monitoring.
Recommendations
The following types of data and activities were suggested as starting points to enhance our
understanding of seismicity in Texas and mitigate risks:
1. Look at GIS data to identify “shadow zones” to try to identify faults in Texas.
2. Assist researchers in collecting high-resolution seismic data in areas of known faults with
oil and gas activity.
3. Work with interested parties (researchers, industry and other governmental entities) to
develop detailed maps of fault stress orientations.
4. Work on ways to determine known fault characteristics in terms of permeability and
porosity.
5. Work on coordinating more detailed and accurate modeling, by including a model of
fluid transport and possible pressure changes along the fault from the injection depth to
the basement.
6. The researchers have requested additional water production information in a more timely
fashion. This request should be considered and we should explore if it is possible to
designate from which zones that water is produced.

  
 

7. Work with operators of disposal wells in close proximity to identified critically stressed
faults to place seismic monitors and develop pressure monitoring regimes.
8. Consider requiring more frequent pressure and volume reporting for disposal wells in
close proximity to known critically stressed faults
9. Consider requiring more frequent bottom hole and fall-off pressure testing for disposal
wells in close proximity to known critically stressed faults.
10. Consider trying to forge collaboration between industry and researchers including the
possibility of pressure monitoring wells for particular injection formations near critically
stressed faults.
11. Consider long-term stress monitoring (pressure over time) for injection formations near
critically stressed faults.
We need to determine how and if we can collect this information in an efficient and affordable
way. Much of this work will require coordination and collaboration between the RRC, industry
and researchers.
The researchers stated that planned additional seismic monitors, if funded, would provide them
the flexibility to deploy seismic monitors strategically across Texas and gather important seismic
data to help them better locate epicenters and depth information for seismic events. These
monitors would also help them study how and when earthquake activity builds and dissipates
over time.
The researchers also expressed a desire to have ongoing meetings similar to this one to continue
to exchange ideas and information as we seek to better understand seismic events in Texas.
Conclusion
This meeting was a good first step in bringing RRC technical staff, researchers and industry
together to discuss an important and complicated issue in Texas. The researchers were
incredibly generous with their time and the information they presented was very helpful – they
are to be commended for their outstanding efforts.
The RRC should do everything in its power to mitigate risks of induced seismicity in the very
small number of injection wells that may be in close proximity to critically stressed faults.
Mapping faults, determining stress orientations, gathering data, deploying monitors, analyzing
data and collaborating with researchers and industry will all take time and effort. We should act
deliberately to get this process moving and make sure that we have a sound scientific basis
before unnecessarily increasing regulatory burdens (e.g., mandating more frequent pressure
testing in areas with no known critically stressed faults would be a waste of time and money).
Universal rules to address seismicity concerns do not appear to be appropriate. Rather, the RRC
should focus on areas where possible seismic activity could occur.