Template matching technique is timely

Ohio Experiences Fuel Award-Winning Paper

Stephen Holtkamp probably would have been featured in these pages had he simply won the award for best student paper at an AAPG Annual Convention and Exhibition.

And he certainly would have had a story written about him if he was awarded the Matson Award for best oral paper presented by anyone at the ACE.

But Stephen Holtkamp, an AAPG member, won both at the same meeting, for the same paper.

Nobody’s ever done that before.

A story, at that moment, was a given.

“I am absolutely honored and excited to receive these awards,” says Holtkamp, a recent Miami of Ohio Ph.D., of the awards announced at the conclusion of this year’s ACE in Pittsburgh.

His paper, co-written with AAPG members Brian Currie and Michael R. Brudzinski, both professors at Miami, and titled “A More Complete Catalog of the 2011 Youngstown, Ohio, Earthquake Sequence From Template Matching Reveals a Strong Correlation to Pumping at a Wastewater Injection Well,” opened up more than just a few eyes.

And kept them open.

“In my presentation, I tried to showcase the advantages of the technique I’ve been working on,” Holtkamp said, “which is a multiple-station (or network) matched filter (or ‘template matching’) technique to consistently identify low level repeating signals in continuous seismic data.”

Traditional seismic techniques use these observations independently, usually by identifying P- and S- waves at each station, which requires that the signal be large enough at several stations to identify these phases.

Holtkamp’s actual waveform template provides information of the largest amplitude phases, and does so at multiple stations simultaneously.

“In stacking multiple observations of the same source signal, the signal-to-noise ratio increases and allows for more consistent detection and detection of smaller signals.”

Putting It to the Test

His lab was close to the action – literally. Like, right outside.

“We applied this technique to the Youngstown, Ohio, earthquake sequence, and did so using only regional seismic stations,” he said, “which are part of the backbone observational network in the United States and freely available online.”

He was able to turn what was a catalog of 11 located earthquakes into one of 282 earthquakes, a 25-fold increase.

What this did, he says, was allow him to test the hypothesis that injection, itself, was causing the earthquakes.

His findings:

  • Earthquake rate changes mirrored the injected volumes, including before and after injection.
  • There was a short delay (one to several days) in earthquake productivity from the time of injection to activity, indicating the amount of time it takes for the fluid pressure pulse to diffuse to the earthquake source region.
  • A progression of earthquake locations along the fault plane from directly beneath the injection occurred throughout the course of the sequence.

“The biggest advantage of this technique is that it allows us to achieve results without relying on expensive and scientifically focused ‘emergency deployments’ of local seismometers, which are only deployed after there is indication that seismicity may be being triggered.”

This is important because there are hundreds of thousands of active wells throughout the United States, and it would be impossible to equip each one with its own network of seismometers.

“We feel,” Holtkamp said, “this technique can provide an inexpensive solution to this problem.”

First, You Have to Drill

As to those eyes that were opened, though, he understands why.

Two topics: Hydraulic fracturing, and earthquakes.

“It is controversial, but the two issues are commonly confused with each other and I think they need to be viewed differently,” Holtkamp said.

“Hydraulic fracturing uses high pressure fluid to create networks of small fractures in shale formations, but once the fracture network is established, they plug up and move on to the next location, commonly adjacent in the same well,” he explained.

His procedure, by comparison, uses as small a volume of fluid as possible.

“And we attempt to recover as much fluid as possible before extracting the gas.”

And he says, simply, that what he is doing isn’t the culprit – if such a culprit even exists.

“Most of the recent cited cases of potentially induced seismicity are related to wastewater disposal or other long-term fluid injection operations (for example, deep water circulation for hydrothermal power plants).”

He says he understands the concern, but thinks it may be misplaced.

“My personal conclusion is that the size of the earthquake that can be triggered depends on the size of faults immediately available for those fluids to flow in to.”

And this is part of the problem – determining it.

“Unfortunately, we will never know the full extent of the fault/fracture network until we actually drill down into the formations.”

A Small Risk

He talks of his own experiment.

“In the Youngstown case, there was a buried fault in the underlying ancient bedrock that was at least a kilometer long, which allowed for larger earthquakes to occur (the largest was a M4.0 on Dec. 31, 2011),” he said. “The disposal well responsible for these events was drilled 200 feet into the bedrock and finished open hole, allowing injected fluids direct access to this fault zone.”

In contrast, he says, the shale formations that are subject to hydraulic fracturing typically only have networks of small joints, which generally don’t exceed a few meters in length each – usually somewhere in the magnitude -2 to -3 range.

“Or about the energy released when you drop a textbook off your desk,” he notes.

Wastewater injection also takes place over much longer time scales.

“It took one year of injection before the M4.0 earthquake, while most fracing phases only last a matter of days at most,” he observed.

“It’s obviously something we should monitor, but it seems like the earthquake risk associated with hydraulic fracturing operations is exceedingly small.”

His work may provide some of that monitoring.

“The techniques I presented in this paper could be used as a tool to investigate many more potential cases of induced seismicity in the country,” he said, “not just ones that have local seismometers monitoring nearby.”

Sounds like a good story idea.

Comments (0)

 

Emphasis: Geophysical Review