URTeC: Shell’s Hackbarth says unconventional innovation begins with data

July 20, 2015
Operators should resist the impulse to pull back on investments in technology despite a down oil market, said Claudia J. Hackbarth, manager, unconventional gas and tight oil research and development, Shell International E&P Inc. during a July 20 luncheon speech at the Unconventional Resources Technology Conference (URTeC) in San Antonio. “Innovation is needed now more than ever.”

Operators should resist the impulse to pull back on investments in technology despite a down oil market, said Claudia J. Hackbarth, manager, unconventional gas and tight oil research and development, Shell International E&P Inc. during a July 20 luncheon speech at the Unconventional Resources Technology Conference (URTeC) in San Antonio. “Innovation is needed now more than ever.”

Hackbarth said energy demand is on pace to double by 2050. Even with increases in traditional and new forms of energy, it will be difficult to meet that demand. With 345 billion bbl of oil and more than 7,000 bcf of gas in technically recoverable resources, unconventional plays have the potential to help fill the shortfall. To realize that potential, she said, industry will have to do things smarter and better.

Hackbarth pointed out how much more we know now than just a decade ago. Operators no longer assess the engineering potential of every new play on how similar it is to the Barnett.

“We no longer believe that just because something worked or failed when we tried it in one place once, that it will happen when we try it—or don’t try it—somewhere else,” she said. “There is more variability that we are aware of now. We better appreciate the complexity of interactions between natural and engineered systems.”

The key to this understanding, as well as future innovation, says Hackbarth, is data.

Data treasure trove

Operators have a “treasure trove” of information and experience from a number of plays over many years, she said. For example, rock and fluid analyses, logs, and extensive production histories. There are multiple types of dynamic observation on a single well or pad. Additionally, we are finding clever ways to collect data and sophisticated means to analyze it.

“Data is the primary technical opportunity to focus on efficiency and cost reduction,” said Hackbarth.

The industry is finding that huge amounts of data can be sifted to find the most important information on any specific question. This leads to better design through better information.

While data mining always looks backwards, a direction for the future would be to see how the industry can learn from all past plays and start with a selection of variables to test at the outset rather than look at a wide swath of variables in retrospect. This would speed design, said Hackbarth.

Using field upsets as diagnostics would help as well, she said. As well spacing and frac spacing are reduced the incidence of well-to-well interference is likely to go up. This undermines well production.

New data, no cost

Short circuits between wells can be diagnosed from close examination of production data, said Hackbarth. An operator can shut in one well to see how it effects production in nearby wells. This matters because optimal spacing comes with an economic expectation. An operator may find out in a year that its wells were drilled too close together. The data to avoid this is readily available, she said, best of all its free.

Hackbarth offered other ideas for gathering new data at no cost. For example, full logging suites are rarely collected after initial appraisal of an area. But drilling data can provide rock strength pseudo-logs for free.

“Download continuously recorded time- and depth-based drilling data that are influenced by rock strength,” she said. “Calibrate once to laboratory rock properties as measured on core and produce pseudo-logs from the data.”

Integrating data across disciplines is critical, said Hackbarth. Today, data gathered by one discipline may be ignored by another. This can lead to different disciplines recommending optimization strategies that are contradictory to each other.

“We have an opportunity to find optimization that honors all the data,” she said. Operators should use available data as constraints, use the data consistently, and reach solid recommendations that are not counterproductive.

“We need to be less limited and more comprehensive,” said Hackbarth.

Contact Michael T. Slocum at [email protected].