Given the amount of seismic shot and number of wells drilled since the mid-1950s, the oil and gas industry has amassed tremendous amounts of data. Long before the concept of Big Data was popular, the E&P industry was searching for ways to tap into that data resource.

Now with all of the microseismic, geochemical, geomechanical and production information coming out of the wellbore, there is increased focus on getting the most value out of those data.

The complexity of unconventional development, for example, is much greater than conventional development. Along with that there is much more data from 3-D and 4-D seismic and microseismic technology. Computing capacity doubles every two years, adding to the complexity.

The data available to oil and gas companies are not limited to private datasets. Information is available from public and regulatory sources as well. But how much data can an operator actually use?

A roundtable on “Big Data Solutions” at the DUG Permian Conference May 20 in Fort Worth, Texas, with Luther Birdzell, CEO of OAG Analytics; Ben Shattuck, upstream analyst for Wood Mackenzie; Alan Lindsey, founder and CEO of PetroDE; and James Yockey, president of Oseberg; delved into how the industry can find value in Big Data.

E&P: What is the definition of Big Data?

Birdzell: Fundamentally, Big Data is about transforming data from a cost- into a revenue-generating asset. It can be measured as incremental barrels of oil per dollar deployed. I think Big Data would be broadly defined as data stored for data requirements that exceeded commercially available technology.

The industry can benefit most quickly and with the greatest magnitude by approaching this problem from the place
of maximizing value to the user by minimizing complexity.

Yockey: When we think about Big Data, we think about huge amounts of disparate data that have different formats
and shapes that are really messy and complex to deal with. But if you can get it right, you can get tremendous value.

Lindsey: The important thing to keep in mind is how the current view of Big Data differs from the way we looked at data in the past. In the past we dealt with a lot of wellstructured datasets that were pretty easy to analyze with the systems we have today. The tools of Big Data allow us to get a handle on much more complicated and less structured data so that we can make sense of information that used to be next to valueless.

Shattuck: The oil and gas industry has been thirsting for data from Day 1. Typically, on the technical side of thebusiness the engineers and geologists want the data. Later, for example with Facebook, there was this huge acceleration of the transformation of data into assets. That caused a particular acceleration of these technologies outside oil and gas. There’s a lot of value to be had by operators and financial services by harnessing the leading edge of that technology.

With the new style of oil and gas drilling, there are hundreds of thousands of wells drilled with associated data at each point along the way. It is wrapping your mind around what that dataset means, what are the components of those data and understanding―particularly as we roll this out to people who are not very technically minded―what the capabilities and the limitations of those datasets are.

Fundamentally, our company changed the way we look at things like forecasting. It is really due to the advent of Big Data. How do you take 4 million wells, separate the noise and distraction from what really matters and roll that up into a meaningful industry-level view?

E&P: How can an operator maximize data analysis?

Lindsey: What you need to do essentially is preprocess that old data to understand what that means. The ultimate is to take us from these expensive well-spacing tests where we’re drilling physical wells to determining that spacing from our microseismic, geochemical and geomechanical measurements so we can get very close to the correct answer right out of the starting gate. That will end up saving millions of dollars.

Most companies take the information on a well-by-well basis and analyze it, put it under a lot of intense scrutiny and then put it on a shelf. What is available now through the Cloud and data technologies is the ability to keep these data alive and on low-cost systems for instant access. Our industry is always worrying about how to optimize these completions, and someone will come up with a new technique.


E&P: What is the future for Big Data analysis?

Yockey: There are tons of great information that oil companies have been capturing for years. To put into context how this fits
in the industry, I was at an investment conference in Houston on May 19. One of the speakers was a managing director and head of the Houston office for McKinsey & Co.

He was talking about innovation and technology across the entire oil and gas industry. In a study that McKinsey published recently, its view was that data and data analytics are going to be the fastest growing segment of technology in the
oil and gas industry with the highest margins, which is pretty interesting.

The second datapoint was about a $1 billion contract that a major oil company signed this year with Palantir Technologies. This is a rumor and not confirmed. Palantir is known for helping the U.S. Department of Defense and intelligence agencies deal with insurgents and terrorists by aggregating a bunch of disparate information and helping them analyze it, normalize it and visualize patterns and trends, with all the data being collected from around the world. The contract was said to be for helping the oil company make sense of the data that it was sitting on.


Lindsey: There are mountains of public data out there, and each company carries with that their own mountains of more detailed information. Frankly, it is the integration of those pieces that brings a lot of added value. One of the things we do at PetroDE is allow companies to integrate all of their public information that they know from their competitors with their own operation-generated data so that they can get competitive analysis and understand where the various plays are going.

E&P: Given all of the data being collected by the oil and gas industry, the task of analyzing those data can be overwhelming. How much collaboration do you think will have to occur to maximize the value?

Birdzell: We’re able to materially reduce the uncertainty of key reservoir and completion practices using advanced statistical
methods from public data. We are able to do even more with proprietary data. One way to cost-effectivelyreduce the uncertainty is to contribute to pools at the same level of data that is required to be disclosed publicly in North Dakota. Operators that are willing to do that in those other plays will very quickly enjoy that value proposition. Ultimately, the most value comes from combining public and proprietary data with even more value for every company involved in pooling those data. Being able to make more decisions more quickly and accelerate beyond the primary process as part of getting our heads around optimized infill wells or wellbore density programs are some of the many areas that we are working on and (that are) contributing to add more value more quickly.

Shattuck: From publically available data sources we roll it into a dataset that [includes] essentially every well across the Lower 48. We found that if we wanted to do meaningful commercial analysis across the Lower 48, we needed to have granular, flow-level data. We found two things when we started to get into the data. First was that the data were overwhelming to begin with, and we needed to make the effort to manage them and pull out what was important. The second was that through the various reporting conventions the data have a lot of different pitfalls.

What we have to do is go back to the drawing board and spend a really tremendous amount of time cleaning up the data. When we talk about Big Data, we’ve got hundreds of data categories, standardizing and sourcing from numerous locations. We’ve got to make sure we have a good dataset and confidence that we can get down to the analysis that we find the most valuable.

Yockey: Frankly, it is the integration of those pieces that brings a lot of added value. There are mountains of public data out there, but each company carries its own mountain of more detailed information. There are lots of opportunities on both sides of that.

There are anomalies and outliers in the data. For example, last September Continental Resources announced its next big play in the Anadarko Basin and the South Central Oklahoma Oil Play. Except for the collapse of oil prices, this would be a play that everyone would be talking about now.

What we saw at Oseberg a year out in September 2013 was that the Springer would be its next big play in the Midcontinent. We could identify that this was a new formation that Continental was going to be talking about by mining the details of the formation-specific IPs that Continental and other operators were releasing through filings with the Oklahoma Corporation Commission.

We could have leveraged a bunch of other data like spacing and increased density from filings to not just have our attention drawn to the Springer but also to delineating what the fairway was and [what] Continental’s, Marathon’s and Newfield’s net positions [were] in the play months before this announcement. That’s worth a lot to geologists, engineers and business executives in the E&P industry.